Clinical Assessment of Malingering and Deception, Fourth Edition PDF - AZPDF.TIPS (2024)

CLINICAL ASSESSMENT OF MALINGERING ANDDECEPTION

Also Available

Conducting Insanity Evaluations, Second Edition Richard Rogers and Daniel W. Shuman

Clinical Assessment of Malingering and Deception F O U R T H

E D I T I O N

edited by

Richard Rogers Scott D. Bender

THE GUILFORD PRESS New York  London

Copyright © 2018 The Guilford Press A Division of Guilford Publications, Inc. 370 Seventh Avenue, Suite 1200, New York, NY 10001 www.guilford.com All rights reserved No part of this book may be reproduced, translated, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, microfilming, recording, orotherwise, without written permission from the publisher. Printed in the United States of America This book is printed on acid-free paper. Last digit is print number: 9 8 7 6 5 4 3 2 1 The authors have checked with sources believed to be reliable in their efforts to provide information that is complete and generally in accord with the standards of practice that are accepted at the time of publication. However, in view of the possibility of human error or changes in behavioral, mental health, or medical sciences, neither the authors, nor the editors and publisher, nor any other party who has been involved in the preparation or publication of this work warrants that the information contained herein is in every respect accurate or complete, and they are not responsible for any errors or omissions or the results obtained from the use of such information. Readers are encouraged to confirm the information contained in this book with other sources. Library of Congress Cataloging-in-Publication Data Names: Rogers, Richard, 1950– editor. | Bender, Scott D., editor. Title: Clinical assessment of malingering and deception / edited by Richard   Rogers, Scott D. Bender. Description: Fourth edition. | New York : The Guilford Press, [2018] |   Includes bibliographical references and index. Identifiers: LCCN 2018000638 | ISBN 9781462533497 (hardback) Subjects: LCSH: Malingering—Diagnosis. | Deception. | BISAC: PSYCHOLOGY /   Forensic Psychology. | MEDICAL / Psychiatry / General. | LAW / Mental   Health. | SOCIAL SCIENCE / Social Work. Classification: LCC RA1146 .C57 2018 | DDC 616.85/2—dc23   LC record available at https://lccn.loc.gov/2018000638

About the Editors

Richard Rogers, PhD, ABPP, is Regents Professor of Psychology at the University of North Texas. He is a recipient of the Guttmacher Award from the American Psychi­atric Association, the Distinguished Contributions to Forensic Psychology Award from the American Academy of Forensic Psychologists, and the Amicus Award from the American Academy of Psychiatry and Law. In addition, Dr. Rogers is only the fourth psychologist to receive Distinguished Professional Contributions awards for both Applied Research and Public Policy from the American Psychological Association. He is the principal author of the Structured Interview of Reported Symptoms (SIRS) and its second edition (SIRS-2), often considered the premier measure for feigned mental disorders. Scott D. Bender, PhD, ABPP-CN, is Associate Professor of Psychiatry and Neuro­ behavioral Science at the University of Virginia (UVA). His primary appointment is with the I­ nstitute of Law, Psychiatry, and Public Policy at UVA, where his duties include teaching, research, and conducting forensic neuropsychological evaluations. Dr. Bender has published extensively, and his research focuses on differential diagnosis of malingering and the effects of traumatic brain injury on neurocognitive and emotional functioning. He frequently testifies on these and related matters in both criminal and civil cases.

v

Contributors

ScottD.Bender, PhD, ABPP-CN, InstituteofLaw, Psychiatry, andPublicPolicy, andDepartment ofPsychiatry andNeurobehavioralScience, UniversityofVirginia, Charlottesville,Virginia DavidT.R.Berry, PhD, DepartmentofPsychology, UniversityofKentucky, Lexington,Kentucky MarcusT.Boccaccini, PhD, DepartmentofPsychology andPhilosophy, SamHoustonStateUniversity, Huntsville,Texas ChelseaM.Bosch, MS, DepartmentofPsychology, UniversityofKentucky, Lexington,Kentucky StaceyL.Brothers, BA, DepartmentofPsychology, UniversityofKentucky, Lexington,Kentucky AbbyP.Clark, MA, DepartmentofPsychology, TheUniversityofAlabama, Tuscaloosa,Alabama AmorA.Correa, PhD, FederalBureauofPrisons, FortWorth,Texas EricY.Drogin, JD, PhD, DepartmentofPsychiatry, HarvardMedicalSchool, Boston,Massachusetts MarcD.Feldman, MD, DepartmentofPsychiatry andBehavioralMedicine, TheUniversityofAlabama, Tuscaloosa,Alabama JamesR.Flens, PsyD, privatepractice, Valrico,Florida RichardFrederick, PhD, privatepractice, Springfield,Missouri NatashaE.Garcia-Willingham, MS, DepartmentofPsychology, UniversityofKentucky, Lexington,Kentucky NathanD.Gillard, PhD, FederalBureauofPrisons, OklahomaCity,Oklahoma JonathanW.Gould, PhD, privatepractice, Charlotte,NorthCarolina RobertP.Granacher, Jr., MD, MBA, LexingtonForensicNeuropsychiatry, Lexington,Kentucky vii

viii Contributors

JamesC.Hamilton, PhD, DepartmentofPsychology, TheUniversityofAlabama, Tuscaloosa,Alabama KimberlyS.Harrison, PhD, HarrisonPsychologicalServices, Austin,Texas NatalieHarrison, MA, DepartmentofPsychology, TheUniversityofAlabama, Tuscaloosa,Alabama JessicaR.Hart, MA, DepartmentofPsychology andPhilosophy, SamHouston StateUniversity, Huntsville,Texas AshleyC.Helle, MS, DepartmentofPsychology, OklahomaStateUniversity, Stillwater,Oklahoma SarahHenry,PhD, DepartmentofPsychology, UniversityofNorthTexas, Denton,Texas WilliamG.Iacono, PhD, DepartmentofPsychology, UniversityofMinnesota, Minneapolis,Minnesota RebeccaL.Jackson, PhD, FloridaCivilCommitmentCenter, Arcadia,Florida RichardA.A.Kanaan, MD, PhD, DepartmentofPsychiatry, UniversityofMelbourne, Heidelberg,Australia JamesL.Knoll, IV, MD, DivisionofForensicPsychiatry, StateUniversityofNewYork UpstateMedicalUniversity,Syracuse,NewYork FranzA.Kubak, PhD, DepartmentofPsychiatry, OregonStateHospital, Portland,Oregon ZinaLee, PhD, SchoolofCriminology andCriminalJustice, UniversityoftheFraserValley, Abbotsford, BritishColumbia,Canada JuliaLevashina, PhD, DepartmentofManagement andInformationSystems, CollegeofBusinessAdministration, KentStateUniversity, Kent,Ohio RichardJ.McNally, PhD, DepartmentofPsychology, HarvardUniversity, Cambridge,Massachusetts MazheruddinM.Mulla, MA, MPH, DepartmentofPsychology, TheUniversity ofAlabama, Tuscaloosa,Alabama DanielJ.Neller, PsyD, privatepractice, SouthernPines,NorthCarolina ChristopherJ.Patrick, PhD, DepartmentofPsychology, FloridaStateUniversity, Tallahassee,Florida SolR.Rappaport, PhD, privatepractice, Libertyville,Illinois PhillipJ.Resnick, MD, DepartmentofPsychiatry, CaseWesternReserveUniversity, Cleveland,Ohio RichardRogers, PhD, ABPP, DepartmentofPsychology, UniversityofNorthTexas, Denton,Texas RandallT.Salekin, PhD, DepartmentofPsychology andDisruptiveBehaviorClinic, TheUniversityofAlabama, Tuscaloosa,Alabama KennethW.Sewell, PhD, DivisionofResearch, OklahomaStateUniversity, Stillwater,Oklahoma GlennSmith, PhD, MentalHealth/BehavioralSciencesService, JamesA.Haley Veterans’Hospital, Tampa,Florida

Contributors ix

LyndaA.R.Stein, PhD, DepartmentofPsychology, UniversityofRhodeIsland, Kingston,RhodeIsland; DepartmentofBehavioral andSocialSciences, BrownUniversity SchoolofPublicHealth, Providence,RhodeIsland; DepartmentofChildren,YouthandFamilies, RhodeIslandTrainingSchool, Cranston,RhodeIsland MichaelJ.Vitacco, PhD, DepartmentofPsychiatry andHealthBehavior, AugustaUniversity, Augusta,Georgia BrittanyD.Walls, MS, DepartmentofPsychology, UniversityofKentucky, Lexington,Kentucky SaraG.West, MD, DepartmentofPsychiatry, CaseWesternReserveUniversity, Cleveland,Ohio CarolS.Williams, LLB, DepartmentofPsychology, AberystwythUniversity, Aberystwyth, Ceredigion,UnitedKingdom PhilipH.Witt, PhD, AssociatesinPsychologicalServices, Somerville,NewJersey ChelseaN.Wooley, PhD, FederalBureauofPrisons, Seagoville,Texas DustinB.Wygant, PhD, DepartmentofPsychology, EasternKentuckyUniversity, Richmond,Kentucky GregoryP.Yates, MA, InstituteofPsychiatry, Psychology andNeuroscience, King’sCollege, London,UnitedKingdom

Preface

This fourth edition of Clinical Assessment of Malingering and Deception represents an important advance for practitioners, researchers, and scholars invested in a sophisticated understanding of malingering and other response styles. On the 30th anniversary of the first edition, it pays homage to its rich past, while embracing the cutting-edge clinical and research methods of today. The fourth edition seeks to broaden its in-depth coverage of response styles with the addition of six valuable chapters in three broad domains. First, the conceptual framework is strengthened with chapters on neuropsychological models and cultural applications. Second, diagnostic issues are enhanced by chapters about deception as it relates to psycho­ pathy and conversion disorders. Third, and finally, two chapters on specialized applications address positive impression management in the context of contested custody evaluations and personnel selection. A major change with the fourth edition involves the addition of a coeditor, Scott D. Bender, who brings his expertise in neuropsychology to capably guide the relevant chapters. While I plan to play a leading role for years to come, Scott is committed to the continued success of Clinical Assessment of Malingering and Deception in future decades. Richard Rogers

xi

Contents

PART I.  CONCEPTUAL FRAMEWORK  1. An Introduction toResponse Styles 3 RichardRogers

 2. Detection Strategies forMalingering andDefensiveness 18 RichardRogers

 3. Neuropsychological Models ofFeigned Cognitive Deficits 42 ScottD.Bender andRichardFrederick

 4. Beyond Borders: Cultural andTransnational Perspectives ofFeigning 61

andOther ResponseStyles

AmorA.Correa

PART II.  DIAGNOSTIC ISSUES  5. Syndromes Associated withDeception 83 MichaelJ.Vitacco

 6. Malingered Psychosis 98 PhillipJ.Resnick andJamesL.Knoll,IV

 7. Malingered Traumatic Brain Injury 122 ScottD.Bender

 8. Denial andMisreporting ofSubstance Abuse 151 LyndaA.R.Stein, RichardRogers, andSarahHenry xiii

xiv Contents

 9. Psychopathy andDeception 174 NathanD.Gillard

10. The Malingering ofPosttraumatic Disorders 188 PhillipJ.Resnick, SaraG.West, andChelseaN.Wooley

11. Factitious Disorders inMedical andPsychiatric Practices 212 GregoryP.Yates, MazheruddinM.Mulla, JamesC.Hamilton, andMarcD.Feldman

12. Conversion Disorder andIllness Deception 236 RichardA.A.Kanaan

13. Feigned Medical Presentations 243 RobertP.Granacher,Jr., andDavidT.R.Berry

PART III.  PSYCHOMETRIC METHODS 14. Assessment ofMalingering andDefensiveness ontheMMPI-2 257

andMMPI‑2‑RF

DustinB.Wygant, BrittanyD.Walls, StaceyL.Brothers, andDavidT.R.Berry

15. Response Style onthePersonality Assessment Inventory 280

andOther Multiscale Inventories

MarcusT.Boccaccini andJessicaR.Hart

16. Dissimulation onProjective Measures: An Updated Appraisal 301

ofaVery OldQuestion

KennethW.Sewell andAshleyC.Helle

17. Feigned Amnesia andMemoryProblems 314 RichardFrederick

18. Assessment ofFeigned Cognitive Impairment Using Standard 329

Neuropsychological Tests

NatashaE.Garcia‑Willingham, ChelseaM.Bosch, BrittanyD.Walls, andDavidT.R.Berry

PART IV.  SPECIALIZED METHODS 19. Assessing Deception: Polygraph Techniques andIntegrity Testing 361 WilliamG.Iacono andChristopherJ.Patrick

20. Recovered Memories ofChildhood SexualAbuse 387 RichardJ.McNally

Contents xv

21. Detection ofDeception inSex Offenders 401 PhilipH.Witt andDanielJ.Neller

22. Structured Interviews andDissimulation 422 RichardRogers

23. Brief Measures fortheDetection ofFeigning 449

andImpression Management

GlennSmith

PART V.  SPECIALIZED APPLICATIONS 24. Deception inChildren andAdolescents 475 RandallT.Salekin, FranzA.Kubak, ZinaLee, NatalieHarrison, andAbbyP.Clark

25. Use ofPsychological Tests inChild Custody Evaluations: 497

Effects ofValidity Scale Scores onEvaluator Confidence inInterpreting Clinical Scales

JonathanW.Gould, SolR.Rappaport, andJamesR.Flens

26. Malingering: Considerations inReporting andTestifying 514

aboutAssessment Results

EricY.Drogin andCarolS.Williams

27. Evaluating Deceptive Impression Management inPersonnel Selection 530

andJob Performance

JuliaLevashina

28. Assessment ofLaw Enforcement Personnel: The Role ofResponse Styles 552 RebeccaL.Jackson andKimberlyS.Harrison

PART VI. SUMMARY 29. Current Status oftheClinical Assessment ofResponse Styles 571 RichardRogers

30. Researching Response Styles 592 RichardRogers

Author Index 615 Subject Index 638

PA R T I

CONCEPTUAL FRAMEWORK

CHAPTER 1

An Introduction toResponse Styles RichardRogers,PhD

Complete and accurate self-disclosure remains a rarity even in the uniquely supportive context of a psychotherapeutic relationship. Even the most involved clients may intentionally conceal and distort important data about themselves. Baumann and Hill (2016) found that outpatient clients sometimes did not divulge personal matters related to sexual experiences, substance abuse, and relationship experiences. Despite imagining positive gains from such personal disclosures, many clients elected not to be fully forthcoming about deeply personal issues. Deceptions in therapy are not relegated to undisclosed personal issues. In surveying 547 former or current therapy clients, Blanchard and Farber (2016) found that many minimized their distress (53.9%) and symptom severity (38.8%). Regarding their therapists, appreciable percentages resorted to deceit in pretending to like their comments/suggestions (29.4%), overstating the effectiveness of therapy (28.5%), and pretending to do homework or other actions (25.6%). Even more concerning was the frequency of these therapy-focused deceptions, which occurred moderate or greater amounts of time. To put these findings in context, therapists also vary considerably in their numbers and types of selfdisclosures (Levitt et al., 2016). Deceptions routinely occur in personal relationships, including intimate relationships, with relatively few (27%) espousing the belief that complete honesty is needed for a successful romantic

relationship (Boon, & McLeod, 2001). Interestingly, these authors found that most persons believe they are much better (Cohen’s d = 0.71) than their partners at “successful” (undetected) deceptions. Even in intimate relationships, willingness to self-disclose is variable and multidetermined (Laurenceau, Barrett, & Rovine, 2005). Romantic partners may have implicitly understood rules about what dishonesties may be allowed in their intimate relationships (Roggensack & Sillars, 2014). Beyond therapy and relationships, deceptions commonly occur in the workplace, including the concealments of mental disorders. Most of the 17 to 20% of employees affected by mental disorders annually elect not to disclose their conditions due to public stigma or more specific concerns about potential damage to their careers (De Lorenzo, 2013). A national survey of professionals and managers by Ellison, Russinova, MacDonald-Wilson, and Lyass (2003) has important implications for understanding individuals’ disclosures and deceptions regarding mental disorders. The majority of these employees had disclosed their psychiatric conditions to their supervisors and coworkers. However, many disclosures were not entirely voluntary (e.g., they were given in response to pressure to explain health-related absences), and about one-third regretted their decisions because of negative repercussions. Moreover, the degree of self-disclosure (e.g., diagnosis, symptoms, or im-

3

4

I .  C o n c e p t u a l F r a m e w o r k

pairment) and the timing of the disclosures were highly variable. Nondisclosing employees were typically motivated by fears of job security and concerns about stigma. What are the two key implications of the study by Ellison et al.? First, decisions about response styles (disclose or deceive) are often rational and multidetermined; this theme is explored later in the context of the adaptational model. Second, these decisions are often individualized responses to interpersonal variables (e.g., a good relationship with a coworker) or situational demands (e.g., explanation of poor performance). This model of complex, individualized decisions directly counters a popular misconception that response styles are inflexible trait-like characteristics of certain individuals. For example, malingerers are sometimes misconstrued as having an invariant response style, unmodified by circ*mstances and personal motivations.1 Decisions to deceive or disclose are part and parcel of relationships across a spectrum of social contexts. For instance, impression management plays a complex role in the workplace, especially with reference to what has been termed concealable stigmas. Jones and King (2014) provide a penetrating analysis of determinants for whether employees disclose, conceal, or signal (i.e., “testing the waters,” p.1471) about themselves (e.g., gender identity) and their own personal experiences (e.g., childhood traumas). Most individuals engage in a variety of response styles that reflect their personal goals in a particular setting. Certain behaviors, such as substance abuse, may be actively denied in one setting and openly expressed in another. Social desirability and impression management may prevail during the job application process but later be abandoned once hiring is completed. Clients in an evaluative context may experience internal and external influences on their selfreporting. Within a forensic context, for example, clients may respond to the adversarial effects of litigation—sometimes referred to as the lexogenic effects—in which their credibility is implicitly questioned (Rogers & Payne, 2006). As observed by Rogers and Bender (2003), these same clients may also be influenced internally by their diagnosis (e.g., borderline personality disorder), identity (e.g., avoidance of stigmatization), or intentional goals (e.g., malingering). By necessity, most chapters in this volume focus on one or more response style within a single domain (e.g., mental disorders, cognitive abilities, or medical complaints). In summary, all individuals fall short of full and accurate self-disclosure, irrespective of the social

context. To be fair, mental health professionals are often not fully forthcoming with clients about their assessment and treatment methods (Bersoff, 2008). In providing informed consent, how thoroughly do most practitioners describe therapeutic modalities, which they do not provide? This question is not intended to be provocative; it is simply a reminder that professionals and their clients alike may not fully embrace honesty at any cost. In the context of clinical assessments, mental health professionals may wish to consider what level of deception should be documented in their reports. One reasoned approach would be to record only consequential deceptions and distortions. For instance, Norton and Ryba (2010) asked coached simulators to feign incompetency on the Evaluation of Competency to Stand Trial—Revised (ECSTR; Rogers, Tillbrook, & Sewell, 2004). However, many simulators could be categorized as doublefailures; they failed to elude the ECST-R Atypical scales (i.e., screens for possible feigning) and also failed to produce anything more than normal to mild impairment (i.e., they appeared competent) on the ECST-R Competency scales. What should be done with such inconsequential distortions? In this specific case, the answer may be characterized as straightforward. Simply as screens, the ECST-R Atypical scales cast a wide net, so that few possible feigners are missed. As a result, no comment is needed, because substantial numbers of genuine responders score above the cutoff scores. The general issue of inconsequential deceptions should be considered carefully. Simply as a thought experiment, two extreme alternatives are presented: the taint hypothesis and the beyondreasonable-doubt standard. 1. Taint hypothesis: Any evidence of nongenuine responding is likely to signal a broader but presently undetected dissimulation. Therefore, practitioners have a professional responsibility to document any observed, even if isolated, deceptions. 2. Beyond-reasonable-doubt standard: Invoking the stringent standard of proof in criminal trials, only conclusive evidence of a response style, such as feigning, should be reported. Between the extremes, practitioners need to decide on a case-by-case basis how to balance the need to document sustained efforts regarding a particular response style with the sometimes very serious consequences of categorizing an examinee as a nongenuine responder.

1. An Introduction to Response Styles 5

In forensic practice, determinations of malingering are generally perceived as playing a decisive role in legal outcomes, because they fundamentally question the veracity and credibility of mental health claims. While it is likely that some genuinely disordered persons may attempt to malinger, the question remains unanswered2 whether fact finders simply dismiss all mental health issues as unsubstantiated. Mental health professionals must decide what evidence of response styles should be routinely included in clinical and forensic reports. Guided by professional and ethical considerations, their decisions are likely to be influenced by at least two dimensions: (1) accuracy versus completeness of their conclusion, and (2) use versus misuse of clinical findings by others. For example, a forensic psychologist may conclude that the examinee’s false denial of drug experimentation during his or her undergraduate years is difficult to establish and potentially prejudicial to a posttraumatic stress disorder (PTSD)-based personal injury case. As an introduction to response styles, this chapter has the primary goal of familiarizing practitioners and researchers with general concepts associated with malingering and deception. It operationalizes response styles and outlines common misconceptions associated with malingering and other forms of dissimulation. Conceptually, it distinguishes explanatory models from detection strategies. Because research designs affect the validity of clinical findings, a basic overview is provided. Finally, this chapter outlines the content and objectives of the subsequent chapters.

FUNDAMENTALS OFRESPONSESTYLES Basic Concepts andDefinitions Considerable progress continues to been made in the standardization of terms and operationalization of response styles. Such standardization is essential to any scientific endeavor for ensuring accuracy and replicability. This section is organized conceptually into four categories: nonspecific terms, overstated pathology, simulated adjustment, and other response styles. NonspecificTerms

Practitioners and researchers seek precision in the description of response styles. Why then begin the consideration of response styles with nonspecific terms? It is my hope that moving from general to

specific categories will limit decisional errors in the determination of response styles. As a consultant on malingering and related response styles, I find that a very common error appears to be the overspecification of response styles. For instance, criminal offenders are frequently miscategorized as malingerers simply because of their manipulative behavior, which may include asking for special treatment (e.g., overuse of medical call for minor complaints) or displaying inappropriate behavior (e.g., a relatively unimpaired inmate exposing his genitals). When disabled clients express ambivalence toward clinical or medical interventions, their less-than-wholehearted attitudes are sometimes misconstrued as prima facie evidence of secondary gain (see Rogers & Payne, 2006). The working assumption for errors in the overspecification of response styles is that practitioners approach this diagnostic classification by trying to determine which specific response style best fits the clinical data. Often, this approach results in the specification of a response style, even when the data are inconclusive, or even conflicting. As outlined in Box 1.1, a two-step approach is recommended. This approach asks practitioners to make an explicit decision between nonspecific or general descriptions and specific response styles. Clearly, conclusions about specific response styles are generally more helpful to clinical conclusions than simply nonspecific descriptions. Therefore, nonspecific descriptions should be considered first to reduce the understandable tendency of overreaching data when conclusions about specific response styles cannot be convincingly demonstrated. Nonspecific terms are presented in a bulleted format as an easily accessible reference. Terms are defined and often accompanied with a brief commentary: •• Unreliability is a very general term that raises questions about the accuracy of reported information. It makes no assumption about the individual’s intent or the reasons for inaccurate data. This

BOX 1.1.  Two-Step (General–Specific) Approach for Minimizing Overspecification 1. Do the clinical data support a nonspecific (e.g., “unreliable informant”) description? 2. If yes, are there ample data to determine a specific response style?

6

I .  C o n c e p t u a l F r a m e w o r k

term is especially useful when faced with conflicting clinical data. •• Nondisclosure simply describes a withholding of information (i.e., omission). Similar to unreliability, it makes no assumptions about intentionality. An individual may freely choose whether to divulge information, or alternatively, feel compelled by internal demands (e.g., command hallucinations) to withhold information. •• Self-disclosure refers to how much individuals reveal about themselves (Jourard, 1971). Persons are considered to have high self-disclosure when they evidence a high degree of openness. It is often considered an important construct within the context of reciprocal relationships (Hall, 2011). A lack of self-disclosure does not imply dishonesty but simply an unwillingness to share personal information. •• Deception is an all-encompassing term to describe any consequential attempts by individuals to distort or misrepresent their self-reporting. As operationalized, deception includes acts of deceit often accompanied by nondisclosure. Deception may be totally separate from the patient’s described psychological functioning (see dissimulation). •• Dissimulation is a general term to describe a wide range of deliberate distortions or misrepresentations of psychological symptoms. Practitioners find this term useful, because some clinical presentations are difficult to classify and clearly do not represent malingering, defensiveness, or any specific response style. OverstatedPathology

Important distinctions must be realized between malingering and other terms used to describe overstated pathology. For example, the determination of malingering requires the exclusion of factitious presentations (see Vitacco, Chapters 5, Yates, Mulla, Hamilton, & Feldman, Chapter 11, this volume). This subsection addresses three recommended terms: malingering, factitious presentations, and feigning. It also includes three quasi-constructs (secondary gain, overreporting, and suboptimal effort) that should be avoided in most clinical and forensic evaluations. Recommended terms to categorize overstated pathology: 1.  Malingering has been consistently defined by DSM nosology as “the intentional production of false or grossly exaggerated physical or psychologi-

cal symptoms, motivated by external incentives” (American Psychiatric Association, 2013, p.726). An important consideration is magnitude of the dissimulation; it must be the fabrication or gross exaggeration of multiple symptoms. The presence of minor exaggerations or isolated symptoms does not qualify as malingering. Its requirement of external incentives does not rule out the co-occurrence of internal motivations. 2.  Factitious presentations are characterized by the “intentional production or feigning” of symptoms that is motivated by the desire to assume a “sick role” (APA, 2000, p.517). However, the description of the motivation is no longer specified; DSM-5 (APA, 2013, p.324) offers only the following: “The deceptive behavior is evident even in the absence of obvious external rewards.” Thus, the diagnosis of factitious disorders does not preclude external incentives but rather requires some unspecified internal motivation. This nonexclusion of external motivations makes sense, since internal and external motivations can often cooccur (Rogers, Jackson, & Kaminski, 2004). 3.  Feigning is the deliberate fabrication or gross exaggeration of psychological or physical symptoms, without any assumptions about its goals (Rogers & Bender, 2003, 2013). This term was introduced because standardized measures of response styles (e.g., psychological tests) have not been validated to assess an individual’s specific motivations. Therefore, determinations can often be made for feigned presentations but not their underlying motivations. To underscore this point, psychological tests can be used to establish feigning but not malingering. Several terms that are common to clinical and forensic practice lack well-defined and validated descriptions. This absence stems from either the lack of clear inclusion criteria, or the presence of multiple and conflicting definitions. Three terms to be avoided in clinical and forensic practice are summarized: 1.  Suboptimal effort (also referred to as incomplete or submaximal effort) is sometimes misused as a proxy for malingering (Rogers & Neumann, 2003). However, this term lacks precision and may be applied to nearly any client or professional (see Rogers & Shuman, 2005). The “best” effort of any individual may be affected by a variety of internal (e.g., an Axis I disorder or fatigue) and external (e.g., client-perceived adversarial context) factors.

1. An Introduction to Response Styles 7

2.  Overreporting simply refers to an unexpectedly high level of item endorsem*nt, especially on multiscale inventories. It has also been called self-unfavorable reporting. Practitioners sometimes erroneously equate it with feigning. However, this descriptive term lacks clarity with respect to its content (i.e., socially undesirable characteristics, as well as psychopathology). Moreover, it has been used to describe both deliberate and unintentional acts (Greene, 2000). 3.  Secondary gain, unlike the other unacceptable terms, does have clear definitions. Its inherent problem for professional practice, however, stems from the presence of conflicting meanings (Rogers & Reinhardt, 1998). From a psychodynamic perspective, secondary gain is part of an unconscious process to protect the individual that is motivated by intrapsychic needs and defenses. From a behavioral medicine perspective, illness behaviors are perpetuated by the social context (e.g., health care providers), not by the individual. From a forensic perspective, individuals deliberately use their illness to gain special attention and material gains. Mental health professionals bear an important responsibility to use professional language that is clearly defined. Ambiguous terminology (e.g., suboptimal effort, overreporting, and secondary gain) adds unnecessary confusion to clinical and forensic assessments. Moreover, the misuse of professional language may lead to grievous errors in adjudicative settings, such as the courts. SimulatedAdjustment

Three closely related terms are used to describe specific response styles that are associated with simulated adjustment. Defensiveness is operationalized as the masking of psychological difficulties, whereas the other two terms apply more broadly the concealment of undesirable characteristics. 1.  Defensiveness is defined as the polar opposite of malingering (Rogers, 1984). Specifically, this term refers to the deliberate denial or gross minimization of physical and/or psychological symptoms. Defensiveness must be distinguished from ego defenses, which involve intrapsychic processes that distort perceptions. 2.  Social desirability is the pervasive tendency for certain individuals to “present themselves in the most favorable manner relative to social norms and mores” (King & Bruner, 2000, p.80). It in-

volves both the denial of negative characteristics and the attribution of positive qualities (Carsky, Selzer, Terkelsen, & Hurt, 1992). Not limited to psychological impairment, social desirability is a far more encompassing construct than defensiveness. Social desirability and its concomitant measurement should be carefully distinguished from defensiveness. 3.  Impression management refers to deliberate efforts to control others’ perceptions of an individual; its purposes may range from maximizing social outcomes to the portrayal of a desired identity (Leary & Kowalski, 1990). Impression management is often construed as more situationally driven than social desirability. It may often involve a specific set of circ*mstances, such as personnel selection (see Jackson & Harrison, Chapter 28, this volume). It can vary dramatically based on cultural expectations (Sandal et al., 2014). Although research studies often assume that impression management involves only a prosocial perspective, individuals may use this response style for a variety of purposes, such as hypercompetitiveness or “playing dumb” (Thornton, Lovley, Ryckman, & Gold, 2009). Preferred terms for simulated adjustment are likely to vary by the professional setting. Clinically, defensiveness is often the more precise term to describe an individual’s minimization of psychological difficulties. Importantly, this term applies to the concealment of psychological impairment rather than efforts to simulate a superior psychological adjustment (see Lanyon, 2001). At least theoretically, well-adjusted persons cannot engage in defensiveness. In many professional contexts that include clinical settings, efforts at self-presentation are likely to involve the concepts of social desirability and impression management. For research on social interactions, impression management is most versatile in describing different roles on a continuum from prosocial to antisocial. As a cautionary note, practitioners and researchers often need to examine the specific simulation instructions, because these terms are often used interchangeably as “fakegood” (Viswesvaran & Ones, 1999). Other ResponseStyles

Several additional response styles are not as well understood as malingering, defensiveness, and other approaches previously described. Four other response styles are outlined:

8

I .  C o n c e p t u a l F r a m e w o r k

1.  Irrelevant responding. This style refers to a response style in which the individual does not become psychologically engaged in the assessment process (Rogers, 1984). The given responses are not necessarily related to the content of the clinical inquiry. This process of disengagement may reflect intentional disinterest or simply carelessness. Occasionally, patterns emerge, such as the repetitive selection of the same option or an alternating response pattern (see commentary by Godinho, Kushnir, & Cunningham, 2016). 2.  Random responding. This style is a subset of irrelevant responding based entirely on chance factors. A likely example would be the completion of the Minnesota Multiphasic Personality Inventory–2 (MMPI-2) in less than 5 minutes. In this instance, the individual has probably read only a few of its 567 items and completed the remainder without any consideration of their content. 3.  Acquiescent responding. This style is commonly referred to as “yea-saying,” which is rarely experienced in its pure form (i.e., indiscriminately agreeing). Rogers, Sewell, Drogin, and Fiduccia (2012) examined acquiescent responding among pretrial detainees. Only 3% showed even a moderate level of acquiescence, but it did not occur most of the time. As an important distinction, acquiescence is clearly distinguishable from social desirability (Gudjonsson & Young, 2011). 4.  Disacquiescent responding. As the opposite of acquiescence, this style is characterized as “naysaying.” When used on scales focused on psychopathology, such as the MMPI-2, disacquiescence essentially eliminates elevations on feigning indicators (Burchett et al., 2016) and presumably for clinical scales. The reason appears to stem from the comparatively few inverted items (i.e., false responses signifying psychopathology). 5.  Role assumption. Individuals may occasionally assume the role or character of another person in responding to psychological measures. For example, Kroger and Turnbell (1975) asked undergraduates to simulate the role of commissioned officers in the air force. This response style is poorly understood but potentially important. 6.  Hybrid responding. This style describes an individual’s use of more than one response style in a particular situation (Rogers, 1984). For example, clients may evidence honest responding about most facets of their lives but engage in defensiveness with respect to substance abuse. Hybrid responding underscores the importance of consid-

ering response styles as adaptable and potentially transitory.

Domains ofDissimulation Response styles are almost never pervasive. For example, malingerers do not feign everything from viral infections to intellectual disabilities. A convenient framework for understanding and assessing response styles is the concept of domains. As I describe in detail in Chapter 2, this volume, three broad domains encompass most attempts at dissimulation: (1) feigned mental disorders, (2) feigned cognitive abilities, and (3) feigned medical complaints/symptoms. These domains are essential to assessment of response styles, because detection strategies are rarely effective across these three domains.

Common Misconceptions aboutMalingering Malingering is unique among response styles in its number of associated myths and misconceptions. Rogers (1998; Rogers & Bender, 2013) outlined common fallacies about malingering held by both practitioners and the public. Common misconceptions are summarized: •• Malingering is rare. Some clinicians simply ignore the possibility of malingering, perhaps erroneously equating infrequency with inconsequentiality. Large-scale surveys of more than 500 forensic experts (Rogers, Duncan, & Sewell, 1994; Rogers, Salekin, Sewell, Goldstein, & Leonard, 1998) suggest that malingering is not rare either in forensic or clinical settings.3 When the outcome of an evaluation has important consequences, malingering should be systematically evaluated. Its professional neglect is a serious omission. •• Malingering is a static response style. Some practitioners use—at least implicitly—the flawed logic, “Once a malingerer, always a malingerer.” On the contrary, most efforts at malingering appear to be related to specific objectives in a particular context. For example, descriptive data by Walters (1988) suggest that inmates rarely feign except when hoping to achieve a highly desired goal (e.g., a single cell based on psychological reasons); among those applying for parole, many inmates understandably manifest the opposite response style (i.e., defensiveness). As a corollary to static response style, researchers have sought to establish personality characteristics linked to ma-

1. An Introduction to Response Styles 9

lingering (e.g., antisocial features; see Kucharski, Falkenbach, Egan, & Duncan, 2006). •• Malingering is an antisocial act by an antisocial person. This common misperception is perpetuated by DSM-5, which attempts to use the presence of antisocial personality disorder (ASPD) as a screening indicator for malingering. As I detail in Chapter 2 (this volume; see the section “Conceptual Issues”), this serious error arises from confusing common characteristics (e.g., criminality in criminal settings) with discriminating characteristics, which consistently differentiate malingerers from nonmalingerers. •• Deception is evidence of malingering. This fallacy is apparently based on the erroneous and illogical notion that “malingerers lie; therefore, liars malinger.” Egregious cases have been observed in which the clients’ marked minimization of symptoms (i.e., defensiveness) was misreported by a practitioner as evidence of malingering. More commonly, deceptions by manipulative inpatients or treatment-seeking inmates are mistakenly equated with malingering (Vitacco & Rogers, 2010). •• Malingering is similar to the iceberg phenomenon. Like the taint hypothesis, this misconception appears to be based on the theory that any evidence of malingering is sufficient for its classification. The erroneous assumption appears to be that any observable feigning, similar to the visible tip of an iceberg, represents a pervasive pattern of malingering. •• Malingering precludes genuine disorders. An implicit assumption is that malingering and genuine disorders are mutually exclusive. This common misconception can sometimes be detected by a careful record review. The typical two-step sequence begins with description of all symptoms as genuine. After the determination of malingering, all symptoms are dismissed as bogus. A more nuanced approach is to doubt, if not discount, all genuine impairment once any feigning has been observed; this negative bias has been observed with the previously noted taint hypothesis and performance validity (see Rogers, Chapter 2, this volume). •• Syndrome-specific feigning scales measure syndrome-specific malingering. Intuitively, mental health professionals—like all persons—would like to assume that names of psychometric scales accurately reflect their descriptions. As a straightforward example, research participants asked to feign somatic problems score high on the MMPI-

2-Restructured Form (MMPI-2-RF) Fs (Infrequent Somatic Responses). Wouldn’t that indicate Fs measures feigned somatic complaints? When using Fs cutoff scores, a very different picture emerges; it is much more likely to identify feigned mental disorders than feigned somatic complaints (Sharf, Rogers, Williams, & Henry, 2017). Clearly, syndrome-specific feigning scales must be able differentiate designated syndrome-specific feigning from generic feigning. •• Malingering has stable base rates. As reported by Rogers et al. (1998), marked variations are observed in the base rates (i.e., SD = 14.4%) for malingering across forensic settings. Even within the same setting, marked variations are likely to occur depending on the referral question and individual circ*mstances. Within the forensic context, the motivation to malinger is dramatically lower for a child custody determination than for an insanity case. Moreover, the assessment process itself also affects applicable base rates. When malingering measures are used with all referrals, the base rate is likely to be relatively low (e.g., 10–30%) even in forensic settings. However, when validated screens (e.g., the Miller Forensic Assessment of Symptoms Test [M-FAST]) are used to identify possible malingerers, the base rate is likely to exceed 50%. Finally, efforts to “correct” base rates in malingering studies often make unbuttressed assumptions, such as the stability of sensitivity.4 The additive effects of multiple misconceptions may fundamentally damage clinicians’ abilities to evaluate malingering and render sound judgments. The effects of inadequate evaluations can be profound for misclassified malingerers and other affected parties. When untested hunches supersede science, then the professional standing of mental health specialties is called into question.

CLINICAL ANDRESEARCHMODELS Motivational Basis ofResponseStyles This section introduces a clinical framework for understanding response styles, such as malingering. Because most response styles are conceptualized as deliberate efforts, individual motivations become a central concern. The motivational basis for response styles, sometimes referred to as explanatory models, has far-reaching implications for clinical and forensic practice. As summarized in the subsequent paragraphs, decisions to dis-

10

I.  Concep t ua l Fr a me work

simulate, such as acting in socially desirable ways or feigning medical complaints, can be viewed in terms of their predicted utility. Often, selection of a particular response style is based on the options available and the desired outcome. The general category of simulated adjustment is likely the most common constellation of response styles, and it encompasses defensiveness, impression management, and social desirability. For example, the minimization of suicidal ideation may serve twin goals, each with its own predicted utility: the maintenance of a positive image and the minimization of social sanctions (e.g., civil commitment). Predicted utilities may focus on others or be predominantly self-focused. As an example of the latter, a male executive may not want to acknowledge his depression, because to do so would be a personal sign of weakness. While it is possible that such defensiveness is unconscious (see, e.g., the Self-Deceptive Enhancement scale; Paulhus, 1998), data suggest that individuals can deliberately modify their “self-deceptive” responses to achieve a desired goal (see Rogers & Bender, 2003). Within the general category of overstated pathology, conceptual and empirical work has focused primarily on malingering. Again, the prevailing model relies on expected utility. Described as the adaptational model, malingerers attempt to engage in a cost–benefit analysis in choosing to feign psychological impairment. In an analysis of malingering cases from 220 forensic experts, the cost–benefit analyses within adversarial contexts were prototypical of malingerers (Rogers et al., 1998). Two other explanatory models have been posited for malingering: pathogenic and criminological (DSM-5). Influenced by psychodynamic thinking, the pathogenic model conceptualizes an underlying disorder as motivating the malingered presentation (Rogers, Sewell, & Goldstein, 1994). The malingerers, in an ineffectual effort to control their genuine impairment, voluntarily produce symptoms. As their condition deteriorates, they presumably become less able to control the feigned disorders. A distinctive feature of the pathogenic model is this prediction of further deterioration. While immediate recovery following litigation is uncommon (i.e., accident neurosis; see Resnick, West, and Payne (2008), research does not support this “further deterioration” hypothesis. Prototypical analysis (Rogers et al., 1998) of the pathogenic model indicated that it is not representative of most malingerers, especially those found in a forensic context.

DSM classifications (American Psychiatric Association, 1980, 1987, 1994, 2000, 2013) have adopted the criminological model to explain the primary motivation for malingering. Its underlying logic is that malingering is typically an antisocial act that is likely to be committed by antisocial persons. Whether this logic is persuasive, empirical data (Rogers, 1990) strongly questioned whether its current operationalization in DSM-5 as four indicators (i.e., forensic context, antisocial background, uncooperativeness, and discrepancies with objective findings) is useful. When DSM indices are evaluated in criminal forensic settings, they are wrong approximately four out of five times. According to Rogers and Shuman (2005), the DSM indicators should not be used even as a screen for potential malingering, because they produce an unacceptable error rate. The fundamental problem with the criminological model is that it relies on common rather than distinguishing characteristics of malingering (see Rogers, Chapter 2, this volume). Most malingerers in criminal forensic settings have antisocial backgrounds and are participating in a forensic consultation. However, the same conclusion is true for many nonmalingering individuals with genuine disorders. Therefore, the criminological model is not useful with criminal forensic and correctional settings. It has yet to be tested with other populations, where it may be less common yet still not distinguish characteristics of malingerers. Returning to predominant predicted-utility model, Lanyon and Cunningham (2005) provide an elegant example of how this model can apply across both domains and response styles. Simulators may attempt to maximize the predicted utility of their efforts by using both overstated pathology (e.g., malingering psychiatric symptoms and health problems) and simulated adjustment (e.g., exaggerating their personal virtues). The latter response style may serve two-related goals: (1) enhance the credibility of the disability claim (e.g., good citizens do not file false insurance reports) and (2) emphasize the magnitude of the purported loss (e.g., the avoidable suffering of an upstanding citizen).

Overview ofResearchDesigns Many skilled practitioners and experienced researchers benefit from a quick overview of research designs as they related to response styles. This brief section highlights key differences in designs and their relevance to clinical practice; for a more

1. An Introduction to Response Styles 11

extensive treatment of response styles, see Rogers (Chapter 30, this volume). Together with Rogers (Chapter 2, this volume), this summary should facilitate the sophisticated use of response style measures presented in subsequent chapters. Four basic research designs are used in most studies of response styles (see Table 1.1). Two basic designs complement each other with their respective strengths: Simulation designs can provide unparalleled control over internal validity, whereas known-group comparisons are unequalled in their consideration of external validity (Rogers & Gillard, 2011). Because of the challenges in establishing the independent categorization required for known-group comparisons, two other designs have been introduced. These designs differ markedly in methodological rigor, from patently simplistic (i.e., differential prevalence design) to potentially sophisticated (partial criterion). The following paragraphs describe these four basic designs and provide salient examples of how each may be misused by clinicians. SimulationDesign

Most research on response styles relies on simulation designs that use an analogue design, which may be augmented by additional samples (see Rogers, Chapter 29, this volume). As noted in Table 1.1, this research often has excellent internal validity, using standardized methods and relying partly on an experimental design, with the random assignment of participants to different experimental conditions. In most malingering studies, for example, community participants are randomly assigned to feigning and control (honest) conditions. To address the critical issue (genuine vs. feigned disorders), the feigning group is typically compared to a nonrandom clinical sample of convenience. The inclusion of clinical comparison groups can become more challenging for research on simulated adjustment. For example, Stein and Rogers (2008; Stein, Rogers, & Henry, Chapter 8, this volume) found that face valid screens may appear to be highly effective when administered to selfdisclosing substance abusers, but understandably, fail utterly when completed by denying substance abusers. For parents in child custody disputes, the key issue in establishing clinical comparison groups is how to distinguish “normal” parents presenting with social desirability from psychologically impaired parents engaging in defensiveness. The lack of an operationalized, clinical com-

parison sample represents a fundamental flaw in simulation research. This fundamental flaw is summarized in Box 1.2, which illustrates how simulation research can be confounded by the absence of relevant clinical samples. For feigning, MMPI-2 research has clearly demonstrated that patients with genuine PTSD can demonstrate extreme elevations on the F scale when responding honestly (Rogers, Sewell, Martin, & Vitacco, 2003). For denied psychopathy, offenders can easily suppress their psychopathy scores on self-report measures below the levels found in undergraduates (Kelsey, Rogers, & Robinson, 2015). In both examples, the failures to include relevant clinical comparison groups represent fundamental oversights in methodology. Known-GroupsComparisons

This design has been increasingly implemented, spurred by rigorously validated measures for feigned mental disorders (e.g., Structured Interview of Reported Symptoms [SIRS and SIRS-2]) and a very stringent detection strategy for feigned cognitive impairment. Regarding the latter, the detection strategy of significantly below chance performance (SBCP; see Chapter 2) can provide compelling data on feigning. To minimize misclassifications, it is critically important to remove an indeterminant group, which in this case includes protocols from slightly below to slightly above chance performance. Performance in this indeterminate range may reflect severe impairment, disengagement (e.g., filling in responses without reference to the test items), or feigning. As noted in Table 1.1, known-groups comparisons should strive for high classification rates (≥ 90%) in order to earn the designation of “known groups.” In doing so, the removal of too-close-tocall cases is essential to minimize both measurement and classification errors (see Rogers, Chapter 2, this volume). It is also imperative to completely mask researchers administering the target measures from any data about known groups. Otherwise, the study fails because of criterion contamination. Differential PrevalenceDesign

Because of challenges in establishing knowngroups comparisons, this design attempts to substitute an expedient proxy, such as referral status for well-established criteria. As a common example, researchers might lump all clients with litigation

12

I.  Concep t ua l Fr a me work TABLE 1.1.  Researching Response Styles: AnOverview of Basic Designs

1.  Simulation research a.  Description. Analogue research randomly assigns participants to different experimental conditions. Results are typically compared to relevant clinical groups. b.  Internal validity. Strong: Procedures include standardized instructions, operationalized conditions, incentives, and manipulation checks. c.  External validity. Weak: Participants do not face the often grave circ*mstances and consequences of succeeding or failing at a particular response style. d.  Classification. Effectively tested: With cross-validation, the accuracy of classification can be evaluated against the experimental condition for specific response styles. 2.  Known-groups comparison a.  Description. The objective is the establishment of highly accurate (≥ 90%) independent classifications of known groups in clinical or other professional settings. Initially, experts using the state-of-the-art methods were used to establish known groups. More recently, rigorous measures of response styles have been implemented with one important caveat: An indeterminant group must be excluded, so that a rigorous standard (≥ 90%) for classification can be achieved (Rogers & Gillard, 2011). b.  Internal validity. Comparatively weak: Researchers have no control over experimental assignment or investment in the investigation (e.g., manipulation checks). However, standardized procedures with aprioristic decision rules can provide systematic data. c.  External validity: Exceptionally strong: The participants, settings, issues and incentives are consistent real-world considerations. d.  Classification. Effectively tested: With cross-validation, the accuracy of classification can be evaluated for specific response styles, often by using rigorous measures and excluding an indeterminate group. 3.  Differential prevalence design a.  Description. Based on assumed incentives, greater numbers of a broadly defined group (e.g., litigation) are presumed to have a specific response style when compared to a second group (e.g., nonlitigation). b.  Internal validity. Weak: Researchers have no control over experimental assignment or other standardized procedures. c.  External validity. Weak to moderate: Participants are often involved in real-world consultations facing important consequences. These consequences could possibly influence the decision to engage in a specific response style. However, the complete lack of any independent classification of response styles stymies the ability to test its effectiveness. When tested empirically, differential prevalence design has produced unacceptably weak effect sizes (e.g., MMPI-2 meta-analysis on feigning yielded a mean d of merely 0.43; Rogers, Sewell, Martin, & Vitacco, 2003). d.  Classification. Untestable: Without knowing group membership, the accuracy of classification is impossible to establish. 4.  Partial criterion design a a.  Description. By using multiple scales or indicators, researchers seek to increase the likelihood of an accurate classification. The goal is to achieve a moderate level of classification, perhaps ≥ 75%. As a partial criterion, it sacrifices accuracy for more expedient research. b.  Internal validity. Weak: Researchers have no control over experimental assignment or other standardized procedures. c.  External validity. Moderately strong when conducted with the appropriate clinical samples. The participants, settings, issues and incentives fit real-world considerations. d.  Classification. Variable: The greatest risk is false positives, because an unknown percentage of classified dissimulators (e.g., deniers of substance abuse) do not warrant this classification.

aThe

partial criterion design was previously described as a “bootstrapping comparison” (Rogers, 2008).

1. An Introduction to Response Styles 13

BOX 1.2. Examples of Flawed Simulation Designs 1. Feigning studies without clinical comparison samples: Researchers do not know whether elevations whether feigners’ scores are any different from genuine responders with severe disorders. 2. Studies of psychopathy on self-report mea‑ sures without a clinical comparison group of defensive psychopaths. Researchers do not know whether their confidentiality-protected responses have any practical relevance to psychopaths practicing general deception or goal-oriented defensiveness.

into a “suspected feigning” group and all nonlitigating clients into a “genuine” group. Such simplism should not be tolerated in clinical research, although it may play a marginal role in advancing theory.5 The fundamental and fatal weaknesses of differential prevalence design can be convincingly illustrated with respect to interpersonal violence. Research (e.g., Whiting, Simmons, Havens, Smith, & Oka, 2009) has clearly supported the intergenerational influences on violence. But, put bluntly, would any self-respecting professional use childhood victimization with violence as an expedient proxy for categorizing all childhood victims as violent persons? However, this use of an expedient proxy is still occasionally applied to feigning; that is, all litigation equals faking, and all nonlitigation equals honest responding. Why should the differential prevalence design be categorically excluded from the classification of response styles? Even when base rates and results appear to be favorable, the fatal weakness of this design prevents its clinical use. For example, using a high estimate of malingering for forensic referrals (32%)6 does not help. It is unlikely but possible that 0.0% of malingerers were identified (i.e., all high scores are false positives); it is also as possible but even less likely that 100% of malingerers were identified. On average, we would expect that about two-thirds (100% – 32% = 68%) of the socalled “malingerers” would be wrongly classified. PartialCriterion

Researchers often provide an external criterion that is limited in its accuracy. Clearly, such research should not be equated with a known-groups comparison, simply because the accuracy of the

classification is not known. Formerly, this design was termed “bootstrapping comparison” (using one measure to improve another measure; see Rogers, 1997). More recently, Rogers and Bender (2013) recommended a more descriptive name: partial criterion design. As noted in Table 1.1, the external measure should have moderately good classification abilities, perhaps ≥ 75%. Rather than simply using the term external criterion for all levels of accuracy, researchers are provided with two designations: known-groups (high accuracy in group membership) and partial criterion (perhaps ≥ 75% accuracy in group membership). Because of its limited accuracy, the partial criterion design should not be used to evaluate utility estimates. Some readers may wonder whether both terms are really needed. As a brief illustration of the issue, Tarescavage and Glassmire (2016) described their design as a “criterion groups” comparison in examining sensitivities between the SIRS and SIRS-2. However, their primary “criterion” measure consisted of a brief feigning screen, specifically, the M-FAST (Miller, 2001). Given the purpose of the M-FAST as a screen, the designation “partial criterion” design would have accurately described this study. Determinations of response styles represent a complex, multifaceted process that includes domains, detection strategies, and measures. A critical first step in mastering assessment methods is the accurate identification of the four basic designs for dissimulation research. Knowledge of these designs allows practitioners to develop a sophisticated appreciation of empirical findings and their clinical relevance. In addition to understanding their respective strengths, mental health professionals must also be able to recognize faulty designs for clinical classification (i.e., the differential prevalence design) and flawed applications to dissimulation research.

LOOKINGFORWARD This book is organized into six major sections that provide a logical progression in the examination of malingering and other forms of deception. Although chapters vary substantially in their content and scope, a unifying theme is the integration of research, theory, and clinical practice. As will become evident, chapters vary in their success at achieving this integration. This variability accurately reflects the strengths and weaknesses in our

14

I.  Concep t ua l Fr a me work

knowledge of response styles. For example, hundreds of studies have examined feigned mental disorders. In contrast, denial of medical complaints is a vast but largely uncharted territory. Understandably, the integration of research and clinical practice will be substantively different between well-established (e.g., feigned mental disorders) and recently considered (e.g., denial of medical complaints) areas of dissimulation research. The overriding goal of most chapters is the provision of clear, usable information that impacts directly on professional practice and clinical research. Whenever possible, specific guidelines are provided regarding the clinical applications of particular measures, scales, and detection strategies. Some dissimulation scales are especially useful for the ruling in (i.e., identification and classification) specific response styles. Other scales may serve an important purpose for the ruling out one or more response styles. When accomplished efficiently, such measures are very useful as screens. Despite our positive focus on advances in the assessment of response styles, we also consider common missteps and inaccuracies that may lead to grievous errors in the determination of dissimulation. Part I, Conceptual Framework, comprises the first four chapters, which operationalize response style terms and provide a conceptual basis for remaining chapters. The centerpiece of Chapter 2 is the description of detection strategies that are organized by response styles and domains. This examination of detection strategies constitutes the essential template for the remaining chapters. As evidence of its growing importance, Chapter 3 delves more closely into different neuropsychological models of feigning. Finally, Chapter 4 recognizes transnational growth in dissimulation research, examining issues of language and culture, and their effects on the assessment of response style. Part II, Diagnostic Issues, comprises nine chapters that address a range of disorders and syndromes for which dissimulation can become a central concern. Chapter 5 provides a broad and valuable overview of specific syndromes and clinical conditions that are frequently associated with dissimulation. Chapters 6 through 13 examine specifically diagnostic categories in which response styles are often considered, especially when consultations have significant financial or forensic relevance. Feigned psychosis (Chapter 6) and denied psychopathy (Chapter 9) represent critically important issues, especially in forensic assessments. Chapters 7 and 10 address very different aspects of traumatic events that may pro-

foundly affect neurocognitive functioning as well as produce psychological reactions, such as PTSD. Chapter 8 is essential to most professional practices, given the nearly endemic substance abuse and its widespread denial. Chapter 11 focuses on nearneighbor comparisons in distinguishing factitious presentations from the closely related construct of malingering. Finally, Chapters 12 and 13 broaden the scope of response styles to consider conversion disorders and deceptive medical presentations. Part III, Psychometric Methods comprises five chapters. Given the breadth and sophistication of dissimulation research, multiscale inventories and feigned cognitive impairment are covered in multiple chapters. In particular, each area is subdivided into two chapters: MMPI-2 and MMPI2-RF (Chapter 14), and the Personality Assessment Inventory and other inventories (Chapter 15). Likewise, cognitive feigning is organized into two chapters: memory and amnesia (Chapter 17), and neuropsychological measures (Chapter 18). Finally, Chapter 16 covers the controversies and clinical data concerning response styles and the use of projective methods. Part IV, Specialized Methods, also comprises five chapters. The usefulness of physiological and other standardized measures is considered in relationship to lie detection (Chapter 19) and sexual deviation (Chapter 21). With continued controversies, Chapter 20 discusses the usefulness and limitations of clinical methods used for the recovery of early memories. Finally, structured interviews (Chapter 22) and brief measures (Chapter 23) make substantial contributions to the assessment of response styles. In Part V, Specialized Applications, chapters are devoted to specific populations and applications. Youth (Chapter 24) and custody and family issues (Chapter 25) are discussed in relationship to response styles. Chapter 26 examines how law professionals learn and are sometimes misled with respect to malingering. Regarding deception and the workplace, Chapter 27 examines this broad and challenging topic, whereas Chapter 28 deals specifically with law enforcement. Part VI, Summary, has an integrative goal of bringing together common and diverse findings across the considerable array of chapters. Chapter 29 summarizes the key conclusions and provides useful guidelines for conducting evaluations of response styles. Chapter 30, presents detailed guidelines—when empirically warranted—on recommended practices for researching malingering and deception. Importantly, it seeks to improve

1.  An Introduction to Response Styles 15

our research methods to more effectively study the complex issues surrounding dissimulation.

NOTES 1.  As an implicit example, a report of malingering during adolescence was used as “evidence” decades later to corroborate the current classification of malingering. 2.  Kirkley (2008) represents a rare attempt to examine the effects of a malingering classification within the context of a disability case. She found that testimony on malingering strongly affected the damage awards but not the decision itself. 3.  The two surveys of mostly forensic psychologists yielded similar data for forensic (Ms of 15.7 and 17.4%) and nonforensic (Ms of 7.2 and 7.4%) referrals. However, the percentages for nonforensic cases may be skewed higher, because forensic practitioners often consult on nonforensic issues that are still highly consequential to clients (e.g., insurance disability claims). 4. These efforts implicitly assume that sensitivity is a stable estimate, whereas positive predictive power (PPP) is not. Although PPP does vary in relationship to base rates, sensitivity also evidences nonsystematic variability). 5.  More precisely, this design would be best used to discount a hypothesized relationship if predicted findings are not observed. 6.  Rogers et al. (1998) used estimates from 221 highly experienced forensic experts. For forensic referrals, the 32% prevalence assumes a rate that is approximately one standard deviation above the Rogers et al. average (M = 17.44%, SD = 14.44%).

REFERENCES American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disorders (3rd ed.). Washington, DC: Author. American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd ed., rev.). Washington, DC: Author. American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: Author. American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: Author. Baumann, E. C., & Hill, C. E. (2016). Client concealment and disclosure of secrets in outpatient psychotherapy. Counselling Psychology Quarterly, 29(1), 53–75.

Bersoff, D. N. (2008). Ethical conflicts in psychology (4th ed.). Washington, DC: American Psychological Association. Blanchard, M., & Farber, B. A. (2016). Lying in psychotherapy: Why and what clients don’t tell their therapist about therapy and their relationship. Counselling Psychology Quarterly, 29(1), 90–112. Boon, S. D., & McLeod, B. A. (2001). Deception in romantic relationships: Subjective estimates of success at deceiving and attitudes toward deception. Journal of Social and Personal Relationships, 18(4), 463–476. Burchett, D., Dragon, W. R., Smith-Holbert, A. M., Tarescavage, A. M., Mattson, C. A., Handel, R. W., et al. (2016). “False feigners”: Examining the impact of non-content-based invalid responding on the Minnesota Multiphasic Personality Inventory–2 Restructured Form content-based invalid responding indicators. Psychological Assessment, 28(5), 458–470. Carsky, M., Selzer, M. A., Terkelsen, K. G., & Hurt, S. W. (1992). The PEH: A questionnaire to assess acknowledgment of psychiatric illness. Journal of Nervous and Mental Disease, 180, 458–464. De Lorenzo, M. S. (2013). Employee mental illness: Managing the hidden epidemic. Employee Responsibilities and Rights Journal, 25(4), 219–238. Ellison, M. L., Russinova, Z., MacDonald-Wilson, K. L., & Lyass, A. (2003). Patterns and correlated of workplace disclosure among professionals and managers with psychiatric conditions. Journal of Vocational Rehabilitation, 18, 3–13. Godinho, A., Kushnir, V., & Cunningham, J. A. (2016). Unfaithful findings: Identifying careless responding in addictions research. Addiction, 111(6), 955–956. Greene, R. L. (2000). The MMPI-2: An interpretive manual (2nd ed.). Needham Heights, MA: Allyn & Bacon. Gudjonsson, G. H., & Young, S. (2011). Personality and deception: Are suggestibility, compliance and acquiescence related to socially desirable responding? Personality and Individual Differences, 50(2), 192–195. Hall, J. A. (2011). Sex differences in friendship expectations: A meta-analysis. Journal of Social and Personal Relationships, 28(6), 723–747. Jourard, S. M. (1971). Self-disclosure: An experimental analysis of the transparent self. New York: Wiley-Interscience. Kelsey, K. R., Rogers, R., & Robinson, E. V. (2015). Selfreport measures of psychopathy: What is their role in forensic assessments? Journal of Psychopathology and Behavioral Assessment, 37, 380–391. King, M. F., & Bruner, G. C. (2000). Social desirability bias: A neglected aspect of validity testing. Psychology and Marketing, 17, 79–103. Kirkley, S. M. (2008). The impact of neuropsychological testimony: Malingering, cognitive impairment, and language complexity. Unpublished doctoral dissertation, University of Alabama, Tuscaloosa, AL. Kroger, R. O., & Turnbull, W. (1975). Invalidity of valid-

16

I.  Concep t ua l Fr a me work

ity scales: The case of the MMPI. Journal of Consulting and Clinical Psychology, 43(1), 48–55. Kucharski, L. T., Falkenbach, D. M., Egan, S. S., & Duncan, S. (2006). Antisocial personality disorder and the malingering of psychiatric disorder: A study of criminal defendants. International Journal of Forensic Mental Health, 5(2), 195–204. Jones, K. P., & King, E. B. (2014). Managing concealable stigmas at work: A review and multilevel model. Journal of Management, 40(5), 1466–1494. Lanyon, R. I. (2001). Dimensions of self-serving misrepresentation in forensic assessment. Journal of Personality Assessment, 76(1), 169–179. Lanyon, R. I., & Cunningham, K. S. (2005). Construct validity of the misrepresentation scales of the Psychological Screening Inventory. Journal of Personality Assessment, 85, 197–206. Laurenceau, J.-P., Barrett, L. F., & Rovine, M. J. (2005). The interpersonal process model of intimacy in marriage: A daily-diary and multilevel modeling approach. Journal of Family Psychology, 19, 314–323. Leary, M. R., & Kowalski, R. M. (1990). Impression management: A literature review and two component model. Psychological Bulletin, 107, 34–47. Levitt, H. M., Minami, T., Greenspan, S. B., Puckett, J. A., Henretty, J. R., Reich, C. M., et al. (2016). How therapist self-disclosure relates to alliance and outcomes: A naturalistic study. Counselling Psychology Quarterly, 29(1), 7–28. Miller, H. A. (2001). MFAST: Miller Forensic Assessment of Symptoms Test professional manual. Lutz, FL: Psychological Assessment Resources. Norton, K. A., & Ryba, N. L. (2010). An investigation of the ECST-R as a measure of competence and feigning. Journal of Forensic Psychology Practice, 10(2), 91–106. Paulhus, D. L. (1998). Paulhus Deception Scales (PDS): The Balanced Inventory of Desirable Responding–7. North Tonawanda, NY: Multi-Health Systems. Resnick, P. J., West, S., & Payne, J. W. (2008). Malingering of posttraumatic disorders. In R. Rogers (Ed.), Clinical assessment of malingering and deception (3rd ed., pp.109–127). New York: Guilford Press. Rogers, R. (1984). Towards an empirical model of malingering and deception. Behavioral Sciences and the Law, 2, 93–112. Rogers, R. (1997). Researching dissimulation. In R. Rogers (Ed.), Clinical assessment of malingering and deception (2nd ed., pp.398–426). New York: Guilford Press. Rogers, R. (1998). Assessment of malingering on psychological measures: A synopsis. In G. P. Koocher, J. C., Norcross, & S. S. Hill, III (Eds.), Psychologist’s desk reference (pp.53–57). New York: Oxford University Press. Rogers, R. (1990). Models of feigned mental illness. Professional Psychology, 21, 182–188. Rogers, R. (2008). Researching response styles. In R. Rogers (Ed.), Clinical assessment of malingering and

deception (3rd ed., pp.411–434). New York Guilford Press. Rogers, R., & Bender, S. D. (2003). Evaluation of malingering and deception. In A. Goldstein & I. B. Weiner (Eds.), Handbook of psychology (Vol. 11, pp.109–129). New Jersey: Wiley Rogers, R., & Bender, S. D. (2013). Evaluation of malingering and related response styles. In R. K. Otto, I. B. Weiner, R. K. Otto, & I. B. Weiner (Eds.), Handbook of psychology: Forensic psychology (pp.517–540). Hoboken, NJ: Wiley. Rogers, R., Duncan, J. C., & Sewell, K. W. (1994). Prototypical analysis of antisocial personality disorder: DSM-IV and beyond. Law and Human Behavior, 18, 471–484. Rogers, R., & Gillard, N. D. (2011). Research methods for the assessment of malingering. In B. Rosenfeld, S. D. Penrod, B. Rosenfeld, & S. D. Penrod (Eds.), Research methods in forensic psychology (pp.174–188). Hoboken, NJ: Wiley. Rogers, R., Jackson, R. L., & Kaminski, P. L. (2004). Factitious psychological disorders: The overlooked response style in forensic evaluations. Journal of Forensic Psychology Practice, 3, 115–129. Rogers, R., & Neumann, C. S. (2003). Conceptual issues and explanatory models of malingering. In P. W. Halligan, C. Bass, & D. A. Oakley (Eds.), Malingering and illness deception: Clinical and theoretical perspectives (pp.71–82). Oxford, UK: Oxford University Press. Rogers, R., & Payne, J. W. (2006). Damages and rewards: Assessment of malingered disorders in compensationcases. Behavioral Sciences and the Law, 24, 645–658. Rogers, R., & Reinhardt, V. (1998). Conceptualization and assessment of secondary gain. In G. P. Koocher, J. C. Norcross, & S. S. Hill (Eds.), Psychologist’s desk reference (pp.57–62). New York: Oxford University Press. Rogers, R., Salekin, R. T., Sewell, K. W., Goldstein, A., & Leonard, K. (1998). A comparison of forensic and nonforensic malingerers: A prototypical analysis of explanatory models. Law and Human Behavior, 22, 353–367. Rogers, R., Sewell, K. W., Drogin, E. Y., & Fiduccia, C. E. (2012). Standardized Assessment of Miranda Abilities (SAMA) professional manual. Lutz, FL: Psychological Assessment Resources. Rogers, R., Sewell, K. W., & Goldstein, A. (1994). Explanatory models of malingering: A prototypical analysis. Law and Human Behavior, 18, 543–552. Rogers, R., Sewell, K. W., Martin, M. A., & Vitacco, M. J. (2003). Detection of feigned mental disorders: A meta-analysis of the MMPI-2 and malingering. Assessment, 10(2), 160–177. Rogers, R., & Shuman, D. W. (2005). Fundamentals of forensic practice: Mental health and criminal law. New York: Springer. Rogers, R., Tillbrook, C. E., & Sewell, K. W. (2004).

1.  An Introduction to Response Styles 17

Evaluation of Competency to Stand Trial—Revised (ECST-R). Lutz, FL: Psychological Assessment Resources. Roggensack, K. E., & Sillars, A. (2014). Agreement and understanding about honesty and deception rules in romantic relationships. Journal of Social and Personal Relationships, 31(2), 178–199. Sandal, G. M., van de Vijver, F., Bye, H. H., Sam, D. L., Amponsah, B., Cakar, N., et al. (2014). Intended selfpresentation tactics in job interviews: A 10-country study. Journal of Cross-Cultural Psychology, 45(6), 939–958. Sharf, A. J., Rogers, R., Williams, M. M., & Henry, S. A. (2017). The effectiveness of the MMPI-2-RF in detecting feigned mental disorders and cognitive deficits: A meta-analysis. Journal of Psychopathology and Behavioral Assessment, 39(3), 441–455. Stein, L. R., & Rogers, R. (2008). Denial and misreporting of substance abuse. In R. Rogers (Ed.), Clinical assessment of malingering and deception (3rd ed., pp. 87–108). New York: Guilford Press. Tarescavage, A. M., & Glassmire, D. M. (2016). Differences between Structured Interview of Reported Symptoms (SIRS) and SIRS-2 sensitivity estimates

among forensic inpatients: A criterion groups comparison. Law and Human Behavior, 40(5), 488–502. Thornton, B., Lovley, A., Ryckman, R. M., & Gold, J. A. (2009). Playing dumb and knowing it all: Competitive orientation and impression management strategies. Individual Differences Research, 7(4), 265–271. Viswesvaran, C., & Ones, D. S. (1999). Meta-analysis of fakability estimates: Implications for personality measurement. Educational and Psychological Measurement, 59, 197–210. Vitacco, M. J., & Rogers, R. (2010). Assessment of malingering in correctional settings. In C. L. Scott & J. B. Gerbasi (Eds.), Handbook of correctional mental health (2nd ed., pp.255–276). Washington, DC: American Psychiatric Publishing. Walters, G. D. (1988). Assessing dissimulation and denial on the MMPI in a sample of maximum security, male inmates. Journal of Personality Assessment, 52(3), 465–474. Whiting, J. B., Simmons, L. A., Havens, J. R., Smith, D. B., & Oka, M. (2009). Intergenerational transmission of violence: The influence of self-appraisals, mental disorders and substance abuse. Journal of Family Violence, 24(8), 639–648.

CHAPTER 2

Detection Strategies forMalingering andDefensiveness RichardRogers,PhD

This chapter introduces detection strategies and provides a conceptual framework for understanding their development and validation. In this context, five essential criteria of detection strategies are examined. The second major section provides an overview of detection strategies as they are applied to specific response styles. This latter section is intended as a template for understanding the specific contributions provided in subsequent chapters. It briefly addresses the assumptions of posttest probabilities. The approach is briefly introduced in the this chapter and comprehensively examined by Bender and Frederick (Chapter 3, this volume).

CONCEPTUALISSUES The modern era for the systematic assessment of response styles was heralded by the empirical development of the Minnesota Multiphasic Personality Inventory (MMPI; Hathaway & McKinley, 1940). Seminal efforts on the MMPI relied on discriminating items that were uncharacteristic of normative populations. Simplest in construction was the F scale that merely relies on MMPI items infrequently endorsed by the Minnesota normative samples. Unlike its current use in determina 18

tions of feigned mental disorders, the F scale was originally intended as a measure of “carelessness and misunderstanding” (Meehl, 1946, p.517). Lacking any conceptual underpinnings, interpretations of F scale elevations cover the gamut, from attentional difficulties and poor reading comprehension to psychotic interference, hostile noncompliance, and deliberate feigning (Dalstrom, Welsh, & Dahlstrom, 1972). Early MMPI versions went beyond carelessness to evaluate core aspects of simulated adjustment, including social desirability and defensiveness (see Rogers, Chapter 1, this volume). The L, or Lie, scale was originally intended to “identify deliberate or intentional efforts to avoid answering the test frankly and honestly” (Dalstrom et al., 1972, p.108). However, a closer inspection of its item content reveals principally a measure of social desirability. Modeled after Hartshorne and May (1928), Hathaway and McKinley (1940) constructed 15 items that involved the denial of personal faults and foibles. As evidence that these faults are widely observed, normative samples typically endorse two-thirds of them as “true.” In contrast to the L scale, the K scale evaluates defensiveness (i.e., the denial of psychological impairment). When McKinley, Hathaway, and Meehl (1948, p.20) observed “normal profiles” in clinically diag-

2.  Detection Strategies for Malingering and Defensiveness 19

nosed samples, they assumed that it was “suggestive of a defensive attitude in the patient’s responses.” What lessons can be learned from the MMPI validity scales? First, the initial conceptualization of the detection strategy is paramount to its subsequent interpretation. The clarity of the K scale development and interpretation is easily contrasted with the interpretational challenges faced by the F scale, which lacks sufficient conceptual underpinnings. Second, the selection of criterion groups dictates the precision of subsequent interpretations. Using the F scale as an example, scales developed simply on normative samples are fundamentally confounded as measures of malingering. Without clinical samples, test developers do not know whether the elevations could result from either feigned or genuine psychopathology. Despite its limitations, the MMPI represents the first critical stage in the development of empirically based detection strategies for response styles (Rogers & Gillard, 2011). Prior to the MMPI, most assessments of malingering and other response styles relied on unproven methods based on case studies. Two major pitfalls can easily arise from relying solely on case studies. First, without systematic investigations, clinicians can unwittingly engage in a tautological exercise: “salient” characteristics of malingering are identified with reference to suspected malingerers, who are thus identified on the basis of these “salient” characteristics (i.e., a classic example of confirmatory bias; Borum, Otto, & Golding, 1991). Second, common features of malingering can be mistaken for discriminating characteristics. Despite the development of empirically validated detection strategies, described later, common features versus discriminating characteristics remain a fundamental problem, especially for the assessment of malingering. DSM screening indices for malingering,1 first established more than 25 years ago (American Psychiatric Association, 1980), continue to misuse two common features of malingering as if they are discriminating characteristics (American Psychiatric Association, 2013). Although malingerers are commonly found in forensic evaluations and often have antisocial backgrounds, the facile use of these two common indices can produce disastrous inaccuracies. In criminal forensic evaluations, all individuals are involved in a forensic evaluation, with many also warranting the diagnosis of antisocial personality disorder (ASPD). For the purpose of this discussion, let us make two simple assumptions: (1) The

prevalence of malingering is about 20% based on extensive surveys (see Rogers, Chapter 1, this volume), and (2) the prevalence of ASPD is 50%. In conducting 1,000 forensic evaluations, 100 of 200 malingerers would be correctly identified. However, 400 of the 800 genuine responders would be wrongly classified. Being wrong four out of five times is simply catastrophic. This hypothetical example appears similar to what has been found empirically (see Rogers, 1990). This simple analysis demonstrates the devastating consequences of confusing common features with discriminating characteristics. The key distinction between common and distinguishing characteristics is illustrated by the following descriptions: • Common features, sometimes described as “clinical correlates,” are often observed in individuals with particular response styles. Even when frequencies exceed 50%, common clinical correlates do not assist in accurate classification. • Distinguishing characteristics refer to specific clinical features that reliably differentiate between relevant groups. They can facilitate accurate classification. The earlier example regarding ASPD and malingering in criminal forensic evaluations clearly illustrates why common characteristics should not be used for accurate classifications. Moreover, most clinical correlates are relatively modest (e.g., rs < .40), which markedly diminishes their practical value. Even when correlates are high (e.g., rs > .70), they do not assist in accurate classification, because other clinical conditions or other response styles may have correlations of a similar magnitude. Using the Millon Clinical Multiaxial Inventory–III (MCMI-III; Millon, 1994) as an example, Scale Z (Debasem*nt Index or faking bad) correlates at .60 or above with six clinical scales; these same scales also correlate with other clinical scales and with Scale X (Disclosure Index or willingness to disclose problems) at the same magnitude.2 An important outgrowth of this focus on discriminating characteristics resulted in the careful formulation of detection strategies. The next major section describes empirically validated detection strategies for common response styles, including a critical examination of their applicability to professional practice.

20

I.  Concep t ua l Fr a me work

Description ofDetectionStrategies Discriminating characteristics are typically specific to a particular scale and cannot—without systematic research—be generalized to other assessment methods. For example, early research on the MMPI L scale suggested that the denial of personal shortcomings and foibles may be useful in the assessment of defensiveness. Does this finding qualify as a detection strategy? As a conditional answer, this approach only becomes a detection strategy when it has been clearly conceptualized, operationalized in terms of specific items, and rigorously tested with multiple measures across multiple settings. Building on earlier conceptualizations (e.g., Rogers, Harrell, & Liff, 1993; Rogers & Bender, 2013), the general definition of detection strategies is presented in Box 2.1. This definition includes five critical criteria, specifically (1) standardized method, (2) conceptual basis, (3) empirical validation, (4) systematic differentiation, and (5) a specific response style. Each component is briefly examined as follows: 1.  Standardized methods are essential to all scientific endeavors. Detection strategies must be operationalized to provide tailored items plus systematic scoring and administration, so that their results can be rigorously tested and cross-validated. 2. A conceptual basis must be described in order to test the underlying rationale for a specific detection strategy and evaluate competing hypotheses. Without a well-defined construct, research may be squandered on an atheoretical approach with difficult-to-interpret results. 3.  Empirical validation focuses on the use of proven methodology to establish the validity of a specific detection strategy. As I summarized in Chapter 1, the empirical validation of detection strategies optimally includes both simulation designs and known-group comparisons. It avoids flawed methodology (e.g., differential prevalence rates) and unsuitable comparisons (e.g., contrasting feigned vs. unimpaired protocols). Other BOX 2.1.  Definition of Detection Strategies A detection strategy is a conceptually based, empirically validated standardized method for systematically differentiating a specific response style (e.g., malingering or defensiveness) from other response styles (e.g., honest responding and irrelevant responding).

important methodological considerations are summarized in Rogers and Gillard (2011) and in Rogers (Chapter 29, this volume). 4.  Systematic differentiation is centered on estimations of accuracy. By itself, statistical significance constitutes an inadequate proxy for accuracy. Many studies of response styles yield results with a high probability of statistical significance but have very little practical utility in professional practice. Instead, the magnitude of difference is the critical issue (Wilkinson & the Task Force on Statistical Inference, 1999). Because of its clarity, this book uses Cohen’s d as the standard measure of effect sizes.3 Beyond effect sizes, the sine qua non of clinical accuracy is level of individual classification. Utility estimates are used to calculate the probabilities that cutoff scores can correctly identify a specific response style (see Streiner, 2003). 5. The delineation of a specific response style (SRS) is essential to the accurate interpretation of results. For example, some research on malingered cognitive impairment attempts to substitute suboptimal effort for malingering. This construct drift (i.e., broadening the conceptualization of malingering to embrace any manifestation of inadequate motivation) results in imprecise and likely misleading results. Researchers and clinicians must verify that criterion groups and instructional sets correspond to the specific response styles under consideration. What is a good detection strategy? In meeting the five previously mentioned criteria, the detection strategy should be cross-validated with different measures and consistently produce large effect sizes and accurate classifications (Rogers & Bender, 2013). If not effective across different measures, then the detection strategy is not sufficiently established and may be capitalizing on idiosyncratic features of one particular scale. If the effect sizes are only moderate, then the detection strategy is comparatively ineffective and should be avoided, unless it can be accurately applied for a circ*mscribed goal, such as ruling out a response style. Importantly, detection strategies for response styles must be targeted, focusing on a specific response style within a particular domain. Indeed, more recently, researchers have attempted to focus more specifically on the faking of specific diagnoses. For example, an MMPI-2 scale was developed to specifically evaluate feigned posttraumatic stress disorder (PTSD; i.e., FPTSD scale; Elhai, Ruggiero, Frueh, Beckham, & Gold, 2002).

2.  Detection Strategies for Malingering and Defensiveness 21

Focused Nature ofDetectionStrategies A fundamental principle is that detection strategies are not universal but must be considered within specific response styles and well-defined domains. Unquestionably, different detection strategies are needed for the evaluation of dissimilar response styles. For example, the assessment of malingering on the MMPI-2 (Rogers, Sewell, Martin, & Vitacco, 2003) uses very different detection strategies than those for the evaluation of defensiveness (Baer & Miller, 2002). Although an inverse relationship between malingering and defensiveness may occur (see, e.g., the bipolarity hypothesis; Greene, 2011), detection strategies focused on a specific response style have proven to be the most effective. To illustrate this point with the MMPI-2, Rogers et al. (2003) found large to very large effect sizes for validity scales based on detection strategies for feigning. In stark contrast, the absence of defensiveness (i.e., low scores on scales using its detection strategies) generally produced only small to moderate effect sizes for feigned responding. Detection strategies must also take into account the broad domains in which specific response styles commonly occur. Three broad domains include mental disorders, cognitive abilities, and medical presentations (Rogers & Bender, 2013). Consider malingering. Individuals feigning a schizophrenic disorder are faced with a very different task than those feigning an intellectual disability. With feigned schizophrenia, malingerers must create believable sets of symptoms and associated features. To be sophisticated, feigners must also decide on the course of the current episode, its concomitant impairment, and their insight into their disorder (e.g., awareness that psychotic behaviors are symptoms). In contrast, persons feigning an intellectual disability must put forth a convincing effort while failing on intellectual and cognitive measures. To be sophisticated, these feigners must also decide not only how badly to fail but on which measures, and how such failures should affect their day-today functioning. Because the tasks of malingerers are dissimilar, different detection strategies are needed. The medical domain is far more complex than either mental disorders or cognitive abilities. With medical malingering, patients can specialize in one debilitating symptom (e.g., pain), portray a constellation of common but distressing ailments (e.g., headaches, fatigue, and gastrointestinal difficulties), or specialize in complex syndromes (e.g.,

fibromyalgia). Researchers sometimes attempt to apply measures for genuine medical impairment to feigned cognitive impairment. For example, Waddell’s signs (i.e., indicators of nonorganic neurological findings) have been extensively misused as evidence of feigning, an application clearly not supported by the empirical literature (Fishbain, Cutler, Rosomoff, & Rosomoff, 2004). In light of the complexities of medical presentations, detection strategies for the medical domain face formidable challenges in their development and validation. As noted, two parameters are necessary in evaluating the usefulness of detection strategies: effect sizes and utility estimates. Effect sizes provide a standardized method for evaluating the comparative value of different detection strategies in distinguishing between relevant criterion groups. Utility estimates examine the effectiveness of particular cutoff scores for individual and group classification of response styles. These parameters are considered in the next two sections.

Effect Sizes andDetectionStrategies Cohen’s (1988) seminal work on effect sizes was designed to consider relatively small differences as relevant to psychological research; for example, he recommended that an effect size of .80 be considered “large,” even though differences were substantially less than one pooled standard deviation. More rigorous standards are needed for professional practice, especially when the presence of a response style may serve to invalidate an individual’s clinical presentation (Ferguson, 2009). For the assessment of malingering, Rogers et al. (2003) proposed more rigorous standards for Cohen’s ds: “Moderate” ≥ .75; “Large” ≥ 1.25; and “Very Large” ≥ 1.75. Based on the meta-analysis of defensiveness by Baer and Miller (2002), the recommended standard for a “Very Large” effect size (≥ 1.50) is slightly lower than that for malingering. As a ready reference, categories for effect sizes (Cohen’s d) are presented in Box 2.2.

BOX 2.2.  Categorization of Effect Sizes Basedon Cohen’s d for the Classification ofResponse Styles • Moderate ≥ .75 • Large ≥ 1.25 • Very Large ≥ 1.50

22

I.  Concep t ua l Fr a me work

Cutoff Scores fortheAccurate Assessment ofResponseStyles This important section includes key issues that should be considered equally by practitioners and researchers. Each paragraph is briefly captioned to facilitate its use as an easily accessible resource. •• Imprecision in psychological assessment. Mental health professionals need to know the accuracy of cutoff scores and the more complex decision rules applied in clinical determinations, such as a specific response style. Even our very best psychometric measures are often imprecise and are sometimes simply inaccurate. As evidence of imprecision, consider for the moment the Wechsler Adult Intelligence Scale–IV (WAIS-IV; Wechsler, 2008), a superb psychometric measure. When taking into account its standard error of measurement (i.e., SEM = 2.16), we can say with 95% certainty that a tested WAIS-IV Full-Scale IQ (FSIQ) of 100 (the 50th percentile) falls somewhere between the 39th and 61st percentiles (see Wechsler, 2008, Appendix A.7). Imprecision is also evident on measures of psychopathology, such as the MMPI-2. Taking into account the SEM on standard MMPI-2 clinical scales (conservatively 6T points; see Rogers & Sewell, 2006), a marginal elevation at 65T has a 95% likelihood of falling between 53T (normal—no elevation) to 77T (moderate elevation). My point here is that some clinicians are overly confident in the precision of their results. Especially when making consequential decisions about response styles, clinicians should take great care to be prudent in their conclusions by taking into account the imprecision of measurement. •• Perils of single-point cutoff score. Rogers and Bender (2013) strongly recommended that psychologists and other practitioners take into account the imprecision of single-point cutoff scores (e.g., for the Personality Assessment Inventory [PAI], a 70T score represents a clinical elevation, whereas a 69T does not). They press this point by suggesting that such exacting discriminations implicitly assume the “laser accuracy myth of cut scores” (Rogers, Gillard, Wooley, & Ross, 2012, p.79; emphasis in the original). Moreover, Rogers et al. demonstrated empirically that PAI scores, which are too close to the cutoff score (± 5T), had an error rate for classifications exceeding 50%. When measurement errors were also considered, the combined error rate was about 75%.

•• Well-defined cutoff scores. The laser accuracy myth can be mostly avoided by simply creating an indeterminate band of scores that is “too close to classify without substantial errors” (Rogers & Bender, 2013, p.522; emphasis in the original). Well-defined cutoff scores—eliminating the narrow band of indeterminate scores (e.g., ± 5T or 1 SEM)—can substantially improve the accuracy of classification. Using Rogers et al. (2012) as an example, the single-point Negative Impression Management (NIM) score ≥ 70T for feigning (i.e., < 70T for genuine responding) becomes a well-defined cut score by removing ± 5T (i.e., increasing the cut score to ≥ 75T for feigning and lowering it to < 65T for genuine responding). In addition, well-defined categorizations may also be achieved via multiple scales constituting a decision model (e.g., the Structured Interview of Reported Symptoms–2 [SIRS-2]; Rogers, Sewell, & Gillard, 2010). •• Advantages of well-defined cutoff scores. The professional use of well-defined cutoff scores decrease both measurement and classification errors. Measurement errors occur whenever measuring complex psychological and medical constructs.4 Scores within one SEM (± 1) are especially vulnerable to measurement error. As researchers on response styles can easily attest, clearly defined bimodal distribution of scores between specific response styles simply do not occur in professional settings. As a result, the establishment of cutoff scores includes some arbitrariness that is reflected in classification errors. Thus, classification accuracy is improved by excluding a narrow range of too-close-to-classify cases. Practitioners and researchers often fail to consider the comparative advantages of single-point versus well-defined cutoff scores. For ready reference, they are delineated in Box 2.3.

BOX 2.3.  Single-Point versus Well-Defined Cutoff Scores • Single-point cutoff scores classify all exam‑ inees as being above or below a designated point; they are prone to both measurement errors and classification errors, which, com‑ bined, may exceed 50%. • Well-defined cutoff scores remove an indeter‑ minate group (e.g., ± 1 SEM) that is too close to classify as a systematic method for improv‑ ing utility estimates.

2.  Detection Strategies for Malingering and Defensiveness 23

Overview ofUtilityEstimates The accuracy of classifications for SRSs should be formally evaluated via utility estimates. Least helpful is the overall hit rate or overall correct classification (OCC), which may obscure important weaknesses. As an extreme example, a cutoff score could miss every single person with a factitious disorder but still achieve a 90% hit rate because of the very low prevalence for factitious disorders. For accuracy of classification, two utility estimates should be considered at a particular cutoff score: • Sensitivity is the proportion of persons with the SRS correctly identified by the cutoff score. If 18 of 20 malingerers are identified by a particular cutoff score (e.g., 65T on the MMPI-2 F scale), then the sensitivity is .90. • Positive predictive power (PPP) is the likelihood that persons meeting a particular cutoff score will be correctly identified as having the SRS. If the same cutoff score correctly identifies 18 of 20 malingerers but misclassifies 60 genuine patients, then the PPP (18/78) is only .23. This example clearly illustrates the importance of considering both sensitivity and PPP in evaluating the accuracy of particular cutoff scores. Extremely high sensitivity can be achieved at the expense of PPP. However, error rates (e.g., false positives) are very important and have been recognized by the Supreme Court as a critical component for the admissibility of expert testimony (see Daubert v. Merrell Dow Pharmaceuticals, Inc., 1993). Historically, utility estimates have overlooked genuine responding as a clinically relevant SRS. Instead, intentional distortions have represented the focal point. This neglect of genuine responding disadvantages many forensic examinees by overlooking potentially “good” news about their efforts at accurate self-representation (i.e., genuine responding). As a more balanced approach, clinicians should weigh—on the basis of the empirical evidence—intentional distortions, unintentional distortions, and genuine responding when evaluating response styles. Use of utility estimates is discussed more extensively in subsequent chapters (see also Streiner, 2003). Of critical importance, cutoff scores should take into account the professional goal for the clinical classification. A crucial distinction must be articulated between screens and clinical determinations. Screens are often effective when used to “screen out” genuine responders. With malinger-

ing as a salient example, I have elaborated on the work of Rogers, Robinson, and Gillard (2014) with the following distinction: 1. Screen-outs try to eliminate the majority of clearly genuine responders from further consideration. To ensure that only genuine responders are removed from further consideration, specificity (likelihood of genuine responding) is intentionally maximized at the expense of sensitivity (likelihood of feigned responding). Thus, by design, those not screened out typically include a substantial number of genuine responders. 2. Clinical determinations are based on comprehensive assessment of response styles. To minimize false positives, PPP is frequently emphasized at some cost to negative predictive power (NPP). Effective clinical determinations set stringent cutoff scores to avoid the misclassification of genuine responders. What are the professional perils? In seeking greater efficiency, some clinicians continue to substitute brief screens for time-consuming clinical determinations. Besides eschewing a more comprehensive multimethod approach, these professionals are substantially increasing the likelihood of false positives, with potentially devastating consequences.

DETECTION STRATEGIES FORSPECIFIC RESPONSESTYLES Detection strategies form two general categories that capitalize on either unlikely presentations or amplified presentations. For unlikely presentations, detection strategies emphasize the presence of unusual and atypical characteristics that are not generally observed in genuine populations. For amplified presentations, detection strategies evaluate the frequency and intensity of characteristics commonly found in genuine populations. This categorization has been tested with feigned mental disorders (Rogers, Jackson, Sewell, & Salekin, 2005) and presented as a useful heuristic for feigned cognitive impairment (Rogers & Bender, 2013). At present, unlikely and amplified presentations provide us with a useful conceptual framework for evaluating the comprehensiveness of detection strategies. Research on detection strategies has focused intensively on two domains: feigned mental dis-

24

I .  C o n c e p t u a l F r a m e w o r k

orders and feigned cognitive abilities. For feigned mental disorders, the emphasis has been divided between malingering and defensiveness. For cognitive abilities, the focus has been solely on malingering, with concerns about defensiveness being largely neglected. Within the domain of cognitive abilities, mental health professionals have assumed that clients’ performance cannot be better than their actual abilities. Although mainly true, some individuals are able to conceal their cognitive deficits while performing work-related responsibilities. Concealed deficits may be either chronic (e.g., cognitive decline) or temporary (e.g., hangover effects for airline pilots; Bates, 2002). Others may not be able to conceal cognitive problems, such as memory failures, but engage in “damage control” explanations that lessen the “severity” of perceived deficits (Erber & Prager, 2000). Thus, concealed cognitive deficits remains an important but understudied area with response styles. In light of the current research, the following three major subsections examine (1) feigning and mental disorders,5 (2) defensiveness and mental disorders, and (3) feigning and cognitive abilities. Despite fewer studies of detection strategies, the fourth and final section summarizes the current data on feigning and medical presentations.

Feigning andMentalDisorders Rogers (1984) provided the original analysis of detection strategies, which combined empirical and heuristic models of malingering. From this earliest analysis, detection strategies for feigning have gradually evolved and continue to be subjected to rigorous examination (Rogers, 2008; Rogers & Bender, 2013). As summarized in Table 2.1, 10 detection strategies for malingered mental disorders6 have been validated. We begin with an examination of “unlikely presentation” detection strategies; for simplicity, this is referred to as unlikely detection strategies. Unlikely DetectionStrategies

As noted, unlikely detection strategies focus on unusual or atypical clinical features that are rarely observed in genuine clinical populations (see Table 2.1). Rare symptoms best exemplify the category of unlikely detection strategies. The strategy is most effective when employing a stringent criterion for identifying rare symptoms (i.e., < 5% of patients with genuine responding) derived from

clinically complex samples (i.e., typically patients warranting multiple diagnoses of serious mental disorders plus personality disorders [PDs]). This strategy tends to produce large to very large effect sizes for specialized feigning measures, such as the SIRS-2 Rare Symptoms (RS) scale (Rogers et al., 2010), and multiscale inventories (e.g., the MMPI2 F-psychiatric or Fp scale; see Rogers et al., 2003). Quasi-rare symptoms represent a much weaker detection strategy than do rare symptoms. As a thought experiment, a naive researcher could devise a subset of quasi-rare symptoms from the Beck Depression Inventory (BDI), based on a well-adjusted community sample, only to find that it failed utterly with depressed inpatients. As illustrated with the BDI, this method of scale development is fundamentally faulty, because selected items may either reflect genuine disorders or feigned disorders. As a case in point, only 25.0% of the items on the F scale (a quasi-rare strategy) were also uncommon in clinical populations.7 Quasi-­ rare symptoms, while producing large to very large effect sizes, are difficult to interpret. They may also result in unacceptable levels of false positives (i.e., the misclassification of a genuine patient as malingering). The third detection strategy, devoted to individual symptoms and features, is improbable symptoms. Virtually by definition, endorsem*nts of improbable symptoms cannot be veridical. However, improbable symptoms represent a trade-off. On the one hand, their fantastic quality increases the likelihood that endorsem*nts are feigned. On the other hand, the high face validity of these items (i.e., recognizable as bogus symptoms) may decrease their effectiveness, especially with sophisticated malingerers. As observed in Table 2.1, two unlikely strategies move beyond individual items to examine infrequent pairs (symptom combinations) and scale configurations (spurious patterns of psychopathology). Given the complexities of establishing “spurious patterns,” this strategy has only been applied successfully to the PAI. However, the strategy of “symptom combinations” is very versatile and comparatively easy to implement. It represents a primary detection strategy on the SIRS-2 and was recently applied to the Structured Inventory of Malingered Symptomatology (SIMS; Widows & Smith, 2005). For the SIMS, Rogers, Robinson, and Gillard (2014) created a Symptoms Combination (SC) scale based on uncorrelated or negatively correlated item-pairs in genuine patients that were often endorsed by feigners.

2.  Detection Strategies for Malingering and Defensiveness 25

TABLE 2.1.  Detection Strategies for Feigned Mental Disorders Unlikely presentation detection strategies Rare symptoms 1. Description: This strategy capitalizes on symptoms or features, which are very infrequently reported (e.g., < 5.0%) by genuine clinical populations. Malingerers are often detected because they overreport these infrequent psychological symptoms and features. 2. Strengths: This detection strategy has been widely applied to different psychological measures; it tends to yield large to very large effect sizes. 3. Limitations: None are noted when this strategy is tested with diverse clinical populations using stringent criteria (e.g., < 5.0%). Problems may occur when it is tested on more hom*ogenous samples (e.g., TSI-2 or with a lax criterion; e.g., < 20%). 4. Examples: (1) SIRS-2 RS (Rare Symptoms) scale, (2) MMPI-2 Fp (F-psychiatric) scale, (3), PAI NIM (Negative Impression Management) scale, and (4) PAI NDS (Negative Distortion Scale). Quasi-rare symptoms 1. Description: This strategy uses symptoms and features that are infrequently found in normative and nonclinical (e.g., community) samples. It is considered a “quasi” strategy, because infrequent items may reflect either genuine disorders or feigned disorders. 2. Strength: This detection strategy produces large to very large effect sizes. 3. Limitations: Because scale items represent infrequent problems for nonclinical groups, many are more frequently reported by patients with genuine mental disorders. Thus, the interpretation of quasi-rare items is often confounded. For example, clients with schizophrenia or posttraumatic stress disorder (PTSD) routinely have marked elevations (e.g., M scores ≥ 80T) on the MMPI-2 F and Fb scales (see Rogers et al., 2003). This confound also contributes to a wide array of cutoff scores. 4. Examples: MMPI-2 F and Fb (F-back) scales. Improbable symptoms 1. Description: This strategy represents an extreme variant of rare symptoms. It utilizes symptoms or features that have a fantastic or preposterous quality. 2. Strength: Because of their fantastic nature, most of its items could not possibly be true. Therefore, substantial endorsem*nt of improbable symptoms typically does not reflect genuine clinical characteristics. 3. Limitation: The extremeness of improbable symptoms may limit its usefulness with sophisticated malingerers, who may be able identify their items as bogus symptoms. 4. Examples: SIRS-2 IA (Improbable and Absurd Symptoms) scale and MCMI-IV Validity Index (VI).

Symptom combinations 1. Description: This strategy utilizes symptoms and features that are common to clinical populations but rarely occur together. Malingerers often endorse a substantial number of infrequent pairs (e.g., grandiosity and increased sleep) rarely observed together in genuine clinical populations. 2. Strengths: This strategy is sophisticated and should be resistant to coaching and other forms of preparation. It also produces large effect sizes. Moreover, it is easily adapted to structured interviews and multiscale inventories, because unlikely pairs can be identified by negligible or even negative correlations in genuine clinical samples. 3. Limitation: At present, it has been tested primarily with feigning measures but could be adapted to multiscale inventories. 4. Examples: SIRS-2 SC (Symptom Combinations) scale, M-FAST RC (Rare Combinations) scale, and SIMS SC (Symptom Combinations) scale. Spurious patterns of psychopathology 1. Description: This strategy is an extensive elaboration of symptom combinations. It relies on certain scale configurations that are very uncommon in clinical populations but sometimes found with feigners. 2. Strength: Its complexity minimizes the possibility that malingerers could prepare and foil its detection strategy. It functions best when using both conceptual and empirical methods for deriving indicators (see MAL below). 3. Limitation: Because of its complexity, the strategy, spurious patterns of psychopathology, requires extensive cross-validation with diverse clinical samples to ensure that its results are not capitalizing on chance variance. Care must also be taken against overinterpretation (e.g., drawing conclusions in the absence of clinical elevations). 4. Examples: PAI MAL (Malingering) index and PAI RDF (Rogers Discriminant Function) index. Amplified detection strategies Indiscriminant symptom endorsem*nt 1. Description: This strategy relies on the finding that some malingerers, unlike most genuine clients, tend to endorse a large proportion across a spectrum symptoms. 2. Strength: The overall proportion of endorsed symptoms is easy to calculate and may be applied to all psychological measures that cover a broad array of clinical characteristics. 3. Limitations: It has been tested mostly with structured interviews. Care must be taken that measures cover a broad array of symptoms; otherwise, its use may lead to false positives. 4. Examples: SIRS SEL (Symptom Selectivity) scale, SADS SEL (Symptom Selectivity) scale, and PAI MFI (Multiscale Feigning Index).

(continued)

26

I.  Concep t ua l Fr a me work

TABLE 2.1. (continued) Symptom severity 1. Description: This strategy capitalizes on the finding that even severely impaired patients experience only a discrete number of symptoms described as “unbearable” or “extreme” in intensity. Malingerers often endorse a wide array of psychological problems with extreme intensity. 2. Strengths: This strategy is easily adaptable to a wide range of structured interviews and clinical scales. It produces large effect sizes. 3. Limitation: At present, symptom severity is usually considered across elevations on multiple scales. However, it may be more effective to evaluate at an item level, such a PAI frequency count of the most severe (e.g., a score of 3, or “very true” for psychotic symptoms) pathological responses. Further research may improve its effectiveness by identifying which psychological problems are almost never characterized as “extreme” in clinical populations. 4. Examples: SIRS-2 SEV (Symptom Severity) scale; MMPI-2 LW (Lachar–Wrobel Critical Items); and MFAST ES (Extreme Symptomatology) scale. Obvious symptoms 1. Description: This strategy relies on the idea that malingerers are likely to report or endorse prominent symptoms that are clearly indicative of serious mental disorders. Obvious symptoms are either considered alone or in relation to subtle symptoms (i.e., perceived by nonprofessionals as “everyday” problems). 2. Strength: This strategy produces large to very large effect sizes. 3. Limitation: Researchers debate whether obvious symptoms should be considered alone or in relation to subtle symptoms. In the latter case, both obvious and subtle symptoms work best if converted to standard scores (e.g., MMPI-2 T scores). 4. Examples: SIRS BL (Blatant Symptoms) scale and MMPI-2 O­S (Obvious–Subtle) scales. Reported versus observed symptoms 1. Description. This strategy uses a pattern of marked discrepancies between the person’s markedly overstated account of his or her clearly visible symptoms, and a professional’s clinical observations. Malingerers can sometimes be identified by this specific pattern of discrepancies (i.e., reporting salient symptoms/features that cannot be observed). 2. Strength: With standardized observations, this strategy provides independent verification of unsubstantiated symptoms and features.

3. Limitation: Because many genuine patients lack insight about their psychopathology, standardization is essential for accurate discrimination. In addition, discrepancies must form clear pattern of markedly overstated clinical features. 4. Examples: SIRS-2 RO (Reported vs. Observed) scale and the M-FAST RO (Reported vs. Observed) scale. Erroneous stereotypes 1. Description: This strategy capitalizes on the finding that many persons, including mental health professionals, have misconceptions about which clinical characteristics are commonly associated with mental disorders. Malingerers are often identifiable by their major overendorsem*nt of erroneous stereotypes. 2. Strength: This strategy appears resistant to preparation because even mental health professionals have difficulty detecting erroneous stereotypes. 3. Limitation: It has been mostly tested with the MMPI-2. 4. Examples: the MMPI-2 Ds (Dissimulation) scale and the PSI EPS (Erroneous Psychiatric Stereotype) scale. Requiring further validation: Close approximations to genuine symptoms 1. Description: This strategy uses apparently bogus symptoms that parallel genuine symptoms except for some important detail.

2. Strength: None is noted.

3. Limitation: Genuine patients may respond to the gist of the item and be misclassified. It has only been tested with one measure; its item content is considered proprietary. 4. Example: MPS MAL (Malingering) scale. Requiring further validation: Overly specified symptoms 1. Description: This strategy assumes that malingerers may be willing to endorse symptoms with an unrealistic level of precision that would distinguish them from patients presenting with genuine-only disorders 2. Strength: None is noted. 3. Limitation: It appears to be a measure of inattention to detail rather than a feigning scale per se. Similar to “close approximations,” genuine patients may respond to the gist or the item. 4. Example: SIRS-2 OS (Overly Specified) symptoms.

Note. The full names of measures listed by abbreviations are presented alphabetically: MCMI, Millon Clinical Multiaxial Inventory; M-FAST, Miller Forensic Assessment of Symptoms Test; MMPI, Minnesota Multiphasic Personality Inventory; MPS, Malingering Probability Scale; PAI, Personality Assessment Inventory; PSI, Psychological Screening Inventory; SADS, Schedule of Affective Disorders and Schizophrenia; SIMS, Structured Inventory of Malingered Symptomatology; SIRS, Structured Interview of Reported Symptoms; TSI, Trauma Symptom Inventory.

2.  Detection Strategies for Malingering and Defensiveness 27

Amplified DetectionStrategies

In contrast to unlikely strategies, amplified detection strategies focus on the excessive degree (e.g., frequency and intensity) of purported symptoms. The two clearest examples are indiscriminant symptom endorsem*nt and symptom severity. In the first instance, malingerers can be identified simply by the sheer number of reported symptoms. In the second instance, malingerers report a high proportion of symptoms as having extreme severity. In both instances, the marked amplification of the clinical presentation becomes the decisive factor, without any consideration of the specific item content. The remaining three strategies focus on content that may appear plausible to malingerers. Because obvious symptoms are easy to recognize and appear to be clinically significant, malingerers may endorse them in greater numbers than their genuine counterparts. While each symptom by itself is plausible, the detection strategy capitalizes on sheer number of endorsed obvious symptoms. Interestingly, malingerers do not necessarily endorse more obvious than subtle symptoms; 8 the detection strategy relies primarily on the proportion of obvious symptoms. Reported versus observed symptoms is a detection strategy that incorporates the clinician’s standardized perceptions of the patient’s self-described symptoms. Although the reported symptoms may be plausible, the detection strategy uses patterned discrepancies with much greater reported symptoms than observed psychopathology. The final detection strategy, erroneous stereotypes, deserves an extended comment. Especially on multiscale inventories (e.g., the MMPI-2), with their complex array of clinical and content scales, persons may wrongly assume that certain clinical characteristics are common among patient populations. When these misassumptions are widespread, erroneous stereotypes can be used to detect likely malingerers. The MMPI-2 Dissimulation scale represents a singular good example, because even mental health professionals make misassumptions about erroneous stereotypes (see Rogers et al., 2003). Similarly, Erroneous Psychiatric Stereotype (EPS; Lanyon, 1993) for the Psychological Screening Inventory (PSI; Lanyon, 1970) uses common misassumptions about persons with mental disorders to evaluate feigning. Although community participants and inpatients have comparable scores on the EPS (d = 0.18), simulators do not generally recognize these erroneous stereotypes. Their overendorsem*nts produce very large

effect sizes (ds of 2.33 and 2.44, respectively). This detection strategy deserves further investigation because of its amplified presentation and possibility for excellent discriminability. Strategies withLimitedValidation

Two detection strategies lack sufficient validation, which limits their use in clinical practice. The use of close approximations to genuine symptoms needs to be cross-validated by other investigators and tested on different measures. A second strategy without sufficient validation is overly specified symptoms, which has a relatively weak conceptual basis: High scores may reflect a general inattention to detail rather than feigning per se. For both strategies, a common risk involves genuine patients responding to the gist of the item and being tripped up by either the close approximation or unrealistic detail. In summary, clinicians have a wealth of detection strategies relying on both unlikely and amplified presentations. These strategies generally produce large to very large effect sizes, which are critical to accurate classifications. As noted in subsequent chapters in this volume, many scales designed to implement these detection strategies remain effective, even when feigners are coached or otherwise prepared.

Malingering andCognitiveAbilities Rogers et al. (1993) provided the first systematic review of detection strategies for feigned cognitive abilities. These strategies continue to be refined and tested with diverse clinical populations. However, two problematic trends have emerged. First, a few detection strategies (e.g., the floor effect) have gained unwarranted popularity at the expense of other sound detection strategies. Second, many researchers have concentrated their efforts with detection strategies on short-term learning and consequently have neglected other important facets of cognitive functioning. Awareness of these trends is important to practitioners so that they (1) select detection strategies based on effectiveness rather than popularity and (2) utilize methods appropriate to the purported deficits. Detection strategies of feigned cognitive abilities, similar to the domain of feigned mental disorders, may also be conceptualized as either unlikely or amplified. Rogers (2008; Rogers & Bender, 2013) originally used the term excessive impairment to describe amplified detection strategies. In this

28

I.  Concep t ua l Fr a me work

edition, we opted for amplified as a clear parallel to the domain of feigned mental disorders. As before, detection strategies with unlikely presentations are referred to as unlikely detection strategies. They focus on unusual and atypical response pattern, infrequently observed in patients with genuine neuropsychological impairment. In contrast, amplified detection strategies emphasize the magnitude of the purported deficits. With respect to organization, the next two subsections address unlikely and amplified detection strategies, respectively (see Table 2.2). These subsections omit neuroimaging indicators, which have yet to be extensively validated (Kingery & Schretlen, 2007). Unlikely DetectionStrategies

Two unlikely detection strategies (magnitude of error and performance curve) focus on response patterns, which are very uncharacteristic of genuine patients. Rogers et al. (1993) first described magnitude of error as unexpected patterns of incorrect responses among feigners that ranges from close misses (approximate answers) to patently wrong answers. This strategy shows definite promise, because most feigners do not take into consideration which wrong responses are much less plausible than another. It has been utilized with existing neuropsychological measures (Liff, 2004) and in developing the Test of Cognitive Abilities (TOCA; see Bender & Rogers, 2004). The performance curve is an unlikely detection strategy that was first recognized by Goldstein (1945); it examines comparative success, taking item difficulty into account. This strategy is simple yet sophisticated, because malingerers are unlikely to be aware of item difficulty when deciding on which items to fail. Frederick (1997, 2003) has produced the most sophisticated application of the performance curve in developing the Validity Indicator Profile (VIP). Violation of learning principles differs from the two other unlikely detection strategies in its conceptual complexity. While representing a general construct, this strategy contains a constellation of well-established learning concepts. The most common learning principle for feigned cognitive impairment involves the comparative advantage of recognition versus recall. Feigners often violate this principle by obtaining comparable scores, despite the relative ease of simple recognition. For example, the Word Memory Test (WMT; Green, 2003) allows for delayed recognition (approxi-

mately 30 minutes) to be compared to immediate recall, delayed recall, and long delayed recall.9 As expected, effect sizes vary by specific comparisons. For 315 compensation cases evidencing good effort, a comparison of delayed recall to delayed recognition yielded a very large effect size (d = 3.09; see Green, Astner, & Allen, 1997). Clearly, more feigning research is needed to examine the effectiveness of this detection strategy beyond the WMT. AmplifiedStrategies

Rogers (2008; see also Rogers & Correa, 2008) described three detection strategies involving claims of excessive impairment (i.e., amplified detection strategies). Because forced-choice testing (FCT) often lacks a clear conceptual basis, it has been moved in this edition to a new category that is designated as “Nonstrategy Methods” in a later section. Of amplified detection strategies, the floor effect has been adapted to dozens of feigning measures. Simply put, malingerers sometimes claim impairment on simple cognitive tasks that are successfully completed by most cognitively compromised populations. The Test of Memory Malingering (TOMM; Tombaugh, 1996) represents a well-validated example of the floor effect strategy that has been successfully applied across various clinical populations (see Frederick, Chapter 17, this volume). However, it is still important to rule out dementias10 and other severe conditions prior to the classification of feigned cognitive impairment on the TOMM. In general, a major drawback of the floor effect, especially in stand-alone measures, is that feigners can easily be educated about how defeat it. Comorbidity may also be an important consideration. For example, persons with severe depression or dementia may lack the motivation and attentional abilities required to complete more time-intensive floor effect measures; even a modest decrement in functioning (e.g., 10–15% errors) may meet the feigning criterion for the floor effect strategy (e.g., the TOMM Retention trial). The strategy of significantly below chance performance (SBCP) was previously termed symptom validity testing (SVT; Rogers, 2008). As originally described (Pankratz, Fausti, & Peed, 1975), SVT clearly referred to SBCP. In the last decade, however, the term and its abbreviation have been used more generally to describe a variety of feigning measures and detection strategies (e.g., Dandachi-

TABLE 2.2.  Detection Strategies for Feigned Cognitive Impairment Unlikely presentation detection strategies Magnitude of error 1. Description: This strategy relies on data indicating that genuine patients often make predictable errors. Most malingerers do not focus on which incorrect answers are common; they are frequently detectable by choosing very wrong responses that are unlikely among genuine patients. 2. Strength: It is less transparent than most cognitive detection strategies and less vulnerable to coaching (Bender & Rogers, 2004). It produces large effect sizes. This strategy could easily be adapted to current forcedchoice formats of standard tests, such as the WAIS-IV Matrix Reasoning subtest. 3. Weakness: None is noted. 4. Examples: “d errors” on the “b Test” and the TOCA Magnitude of Error (MOE) scale. Performance curve 1. Description: This strategy is based on the general finding that genuine patients produce a predictable pattern: fewer successes with increased item difficulty. When plotted, this “rate of decay” forms a characteristic “performance curve.” Malingerers, unaware of this pattern, typically produce much less discrimination between easy and difficult items. 2. Strength: It is a sophisticated strategy that may prove to be resistant to coaching. 3. Weakness: It may be challenging to implement this strategy on existing measures, because it requires a broad range of item difficulty. 4. Examples: VIP Performance Curve and the TOCA Performance Curve (PC). Violation of learning principles 1. Description: This strategy is a specialized application of performance curve; some malingerers are unaware of underlying learning principles. 2. Strength: It is conceptually strong because it is based on rigorously evaluated learning principles. Malingerers may not take into consideration that different performances are expected based on learning principles: (a) recognition vs. recall, (b) cued recall vs. free recall, (c) immediate vs. delayed recall, (d) simple recall vs. cognitive transformation (e.g., “Tell us in your own words”) and (e) priming effect (Haines & Norris, 1995). 3. Weakness: Some violations of learning principles produce only modest group differences. Therefore, this detection strategy needs to be rigorously tested to minimize false positives. 4. Examples: RAVLT (see Sullivan et al., 2002), WMT Immediate Recognition versus Delayed Recognition, and WMT Delayed Recall versus Delayed Recognition. Amplified detection strategies Floor effect 1. Description: This strategy capitalizes on the finding that some malingerers do not recognize that very easy

cognitive tasks (i.e., “too simple to fail”) can be successfully completed by most impaired persons. 2. Strength: It is easily adaptable to many cognitive measures. 3. Limitation: When the strategy is used in a stand-alone measure, malingerers can easily be coached (e.g., “just succeed”). 4. Examples: Rey-15, TOMM, WMT, and LMT. Significantly below chance performance (SBCP) 1. Description: This strategy uses a forced-choice paradigm to test whether an individual’s failure rate is significantly below probability based on a chance performance. When given two equiprobable choices, even the most impaired individuals should succeed approximately 50% of the time (i.e., chance levels). 2. Strength: Failures significantly below chance provide definitive evidence of feigning. 3. Limitation: Most malingerers do not need to fail at such an unlikely level to achieve their objectives. Therefore, this strategy is typically successful in less than 25% of feigned cases. 4. Examples: PDRT, CARB, VSVT, WCT, and TOMM. Requiring further validation: Consistency across comparable items 1. Description: Genuine patients with stable mental status tend to perform consistently across items of comparable difficulties. Some malingerers are much more variable in their performance and can be identified by their marked inconsistencies. 2. Strength. With rigorous testing, discrepancies can be effective in distinguishing between criterion groups (feigners vs. genuine patients). At present, only the VIP has been rigorously tested for consistency across comparable items. 3. Weaknesses: Research sometimes addresses only group differences, without looking at intraindividual performances on comparable items. It is increasingly evident that appreciable intertest scatter is the norm, not the exception. Because of unknown effects of comorbidity, the presence of serious mental disorders or personality disorders represents a potential confound. 4. Example: VIP Equivalent Item Pairs.a Requiring further validation: Symptom frequency 1. Description: This strategy is based on the idea that some malingerers may report symptoms associated with cognitive impairment at a much higher rate than genuine populations. 2. Strength: With cross-validation, this strategy extends beyond cognitive performance to examine patients’ reported symptoms and their potential interference in day-to-day functioning. 3. Weaknesses: This approach has only been systematically evaluated with a single measure. A major concern is whether its results will be confounded by severe comorbidity. 4. Example: NSI (Gelder et al., 2002) total score; most feigners endorse more than 25% the total possible score.

Note. The full names of measures listed by abbreviations are presented alphabetically: CARB, Computerized Assessment of Response Bias; LMT, Letter Memory Test; NSI, Neuropsychological Symptom Inventory; PDRT, Portland Digit Recognition Test; RAVLT, Rey Auditory Verbal Learning Test; TOMM, Test of Memory Malingering; VIP, Validity Indicator Profile; VSVT, Victoria Symptom Validity Test; WAIS-IV, Wechsler Adult Intelligence Scale–IV; WMT, Word Memory Test. aThey are described in Frederick (1997) but not in the revised manual (Frederick, 2003).

29

30

I.  Concep t ua l Fr a me work

FitzGerald, Ponds, & Merten, 2013). Although unwieldy, a precise term will be employed: SBCP. Unique among detection strategies, SBCP can be used to accurately calculate the likelihood of false positives. In two-choice paradigms with equiprobable alternatives, binomial probabilities can be calculated to approximate the likelihood that a particular below-chance performance could occur in a nonfeigning individual. At extremely low scores (e.g., ≤ 25 on the 72-item Portland Digit Recognition Test; Binder, 1993), the probability of feigning exceeds 99%, with less than 1% of genuine responders being misclassified. The strength of SBCP (i.e., virtual certainty at extremely low scores) is also its inherent limitation. Feigning catastrophic failures on psychometric measures are not required in most clinical and forensic cases. Therefore, malingerers frequently feign severe but not extreme cognitive problems, thus eluding detection on SBCP measures. Nonetheless, practitioners continue to use SBCP (1) because of the unparalleled certainty that can be achieved, albeit in a small proportion of cases, and (2) because it can be integrated into measures employing other more sensitive (but less specific) strategies. Strategies withLimitedValidation

The potential strategy, consistency across comparable items, makes conceptual sense with higher functioning examinees but is subject to error for patients with genuine cognitive impairment who exhibit variable performances (see Bender & Frederick, Chapter 3, this volume). However, significant numbers—because of attentional and motivational issues—evidence a substantial decline in cognitive functioning as their testing progresses. A pattern of decline or other variability in cognitive performance can confound this potential strategy. On this point, Frederick (1997) described Equivalent Item Pairs for the VIP, but did not feature them in the revised manual (Frederick, 2003). Detection strategies for feigned cognitive abilities—both unlikely and amplified—focus predominantly on examinees’ test performance on cognitive measures. As an alternative, Gelder, Titus, and Dean (2002) utilized the Neuropsychological Symptom Inventory (NSI) to evaluate the potential feigning of symptoms and clinical characteristics associated with the reported cognitive deficits. They devised a potential detection strategy, symptom frequency, to evaluate how often an array of neuropsychological symptoms is reported in feigners compared to those with genuine cogni-

tive deficits. Fitting within the amplified domain, an obvious concern is whether severely impaired genuine examinees might be misclassified as feigning. NonstrategyMethods

Two clinical indicators—FCT and atypical presentation (ATP)—generally lack the sound conceptual basis required for detection strategies (see Table 2.3). FCT was intended to relax the standards for SBCP, so that more feigners would be

TABLE 2.3.  Nonstrategy Methods for Feigned Cognitive Impairment

These methods do not utilize conceptually based detection strategies. Instead, they are based on performance that is lower or otherwise different than that found in samples of presumably genuine patients. Forced-choice testing (FCT) 1.  Description: This method is based on the observation that some malingerers evidence lower performance than typically found for genuine patients with cognitive impairment. 2.  Strength: None is noted. 3.  Limitations: FCT lacks a conceptually based detection strategy. It relies simply on poor performance without specifying how that performance differs from genuine efforts. Other limitations for most FCT measures include not being tested with a full range of clinical conditions, and not being tested with serious mental disorders that may confound results. 4.  Examples: CARB and PDRT. Atypical test pattern (ATP) 1.  Description: This method is based on the observation that some malingerers perform differently on certain scales than samples of genuinely impaired persons. In its clearest form, these patterns are identified statistically without any consideration of their theoretical bases. 2.  Strength: None is noted. 3.  Weaknesses: It is a conceptually weak approach, lacking the conceptual clarity of detection strategies. Use of discriminant analysis often capitalizes on chance variation; thus, this approach requires extensive cross-validation. 4.  Examples: Approximate answers for the Ganser syndrome.

Note. The full names of measures listed by abbreviations are presented alphabetically: CARB, Computerized Assessment of Response Bias; PDRT, Portland Digit Recognition Test.

2.  Detection Strategies for Malingering and Defensiveness 31

classified. It differs from the floor effect in that its items are not typically evaluated for their simplicity. Rather than insisting on “below chance” performance, FCT simply requires “below expected” performance. How is below expected performance assessed? Typically using clinical samples of convenience, the lower range of scores for genuine patients is used to establish the cutoff scores for malingering. This expediency (simply below expected performance), characteristic of FCT, lacks the sound conceptual basis underlying most established detection strategies. Without extensive normative data including major mental disorders and cognitively compromised diagnostic groups, the rate of false positives cannot be established. Similar to FCT, ATPs typically are not informed by a specific detection strategy. Scholars have sought to capitalize on an innovative idea or a striking observation. As a salient historical example, the response pattern of “approximate answers” was apparently observed in three prisoners by Sigbert Ganser in 1897 and continues to be sporadically reported in case studies (e.g., Dwyer & Reid, 2004), despite the lack of empirical research linking it to factitious disorders. As a more recent example, Mittenberg, Theroux-Fichera, Zielinski, and Heilbronner (1995, p.492) posited that brain-injured patients would “show similar levels of performance” on the WAIS-R Vocabulary and Digit Span subtests; therefore, greater decrements in Digit Span might be evidence of malingering. PosttestProbabilities

Frederick’s (2015) penetrating analysis shows that focusing only on counting feigning indicators to produce posttest probabilities can produce utterly spurious results. According to Larrabee (2008), three or more “positive scores” for feigning increased the odds for feigning to virtually a certainty (99.0 to 99.5%). As noted by Frederick (2015), “negative scores” (i.e., not feigning) must also be considered, which may remarkably reduce the likelihood of feigning. Moreover, applying three or more “positive scores” creates a numerical nightmare. Even if practitioners and researchers were to artificially limit their analysis to 30 feigning indicators, the possibilities are truly staggering, with a factorial of 265,252,859,812,191,058,636,308,480,000,000. Larrabee’s use of 3+ positive scores as highly indicative of malingering relies on at least four implicit assumptions that are summarized in Box 2.4.

BOX 2.4.  Implicit Assumptions of Posttest Probabilities 1. Interchangeability. It assumes that any combi‑ nation of feigning indicators is equal to any other combination of the same number (i.e., assuming that any four indicators produce identical results to any other four indicators). 2. Independence. It assumes uncorrelated indicators; otherwise, the summing of related items would be inappropriate (e.g., adding highly correlated items would be tantamount to “double-counting”). 3. Equal weighting. It assumes that the predic‑ tive power of each indicator is identical; otherwise, items would need to be weighted. 4. Universal applicability. Within the realm of feigned cognitive abilities, indicators are assumed to be equally applicable to any feigned presentation (e.g., malingered intellectual disability and feigned learning disability).

When looking across feigning indicators and specialized populations, practitioners are confronted with very heterogeneous results (see, e.g., Boone, 2007) based on cutoff scores, measures, and clinical presentations. Despite positive results on a small number of two-way and three-way combinations (e.g., Larrabee, 2003), the crucial assumptions of the posttest probabilities remain virtually untested. Therefore, the use of combined probabilities for feigning indicators—in light of these unknowns—may be inaccurate and misleading.

Simulated Adjustment andMentalDisorders As described by Rogers (Chapter 1, this volume), simulated adjustment is a broad term used to describe an overly favorable self-presentation. For detection strategies, clinical research has focused predominantly on defensiveness (denial or minimization of especially psychological symptoms) and social desirability (simulation of a much more positive self-image). Although these constructs are clearly distinguishable, they are often conflated in both the development of detection strategies and the interpretation of test findings. The prevalence of defensiveness among mentally disordered samples is unknown, but it likely exceeds malingering and other forms of dissimulation. Baer and Miller (2002) estimated the base rate of defensiveness at 30% in job applicant and

32

I.  Concep t ua l Fr a me work

child custody referrals. Applying the three most effective scales in Baer and Miller to Greene’s (2000) analysis of Caldwell’s dataset, estimates of defensiveness range from 16 to 33% of clinical referrals. In the community—even with the promise of anonymity—nonclinical participants may engage in defensiveness and fail to disclose subthreshold psychotic experiences (DeVylder & Hilimire, 2015). Despite its importance clinically and nonclinically, the development of detection

strategies for defensiveness is less advanced when compared to the assessment of malingering. Defensiveness is operationalized through two amplified detection strategies (see Table 2.4). The MMPI-2 K scale best represents the denial of patient characteristics in its empirical development of test items common in clinical populations. Using empirical keying, its items tend to be denied by clients with an array of mental disorders, but who produced unelevated MMPI clinical scales.

TABLE 2.4.  Detection Strategies forSimulated Adjustment: DefensivenessandSocialDesirability

Defensiveness Denial of patient characteristics 1.  Description: This strategy capitalizes on research demonstrating that certain attributes are commonly endorsed by clinical populations. 2.  Strength: It is designed specifically to evaluate patients who do not acknowledge their psychological problems. Its items may be less transparent than those rationally based on idealized attributes. 3.  Limitation: Scales produce moderate effect sizes and are vulnerable to coaching. 4.  Examples: MMPI-2 K scale, and MMPI-2 Edwards Social Desirability (Esd) scale.a Spurious patterns of psychological adjustmentb 1.  Description: This strategy relies on certain scale configurations are characteristic of defensiveness but are very uncommon in clinical and community populations. 2.  Strength: Its complexity minimizes the extent to which defensive responders can foil this detection strategy. 3.  Limitation: Because of its complexity, the spurious patterns of simulated adjustment strategy requires extensive cross-validation to ensure that its results are not capitalizing on chance variance. Care must also be taken against overinterpretation (e.g., drawing conclusions when clinical elevations are present). 4.  Examples: PAI Defensiveness Index (DEF) and PAI Cashel Discriminant Function (CDF). Social desirability

2.  Strength: None is noted. 3.  Limitations: The complete denial, as contrasted with comparative statements (e.g., “better than most”), is unnecessary for clients denying maladjustment. It tends to produce moderate effect sizes and appears vulnerable to coaching. 4.  Example: MMPI-2 Lie (L) scale. Blended strategy with affirmation of virtuous behavior and denial of personal faults 1.  Description: This strategy combines the affirmation of overly positive attributes with the denial of common foibles. 2.  Strength: It produces moderate to large effect sizes. 3.  Limitation: As a blended strategy, it is difficult to know which component (i.e., affirmation or denial) is more effective. 4.  Examples: Marlowe–Crowne, PDS Impression Management (IM), and MMPI-2 Superlative (S) scale. Social desirability 1.  Description: This strategy attempts to identify persons striving to create a very favorable selfpresentation to others. 2.  Strengths: It produces larger effect sizes than most other defensive strategies. It also appears to be comparatively effective even when persons are coached about the strategy. 3.  Limitation: The term social desirability has been defined and operationalized in several different ways. This definition has been examined via the use of a single scale (Wsd). 4.  Example: MMPI-2 Wsd.

Denial of personal faults 1.  Description: This strategy is based on the idea that persons minimizing maladjustment take this to the extreme and deny any shortcomings or nonvirtuous behaviors. Note. Eds, Edwards Social Desirability scale; PDS, Paulhus Deception Scales; Wsd, Wiggins Social Desirability scale. aDespite its name, the Esd focuses on the denial of common psychological problems. bIn the 2008 edition of this volume, this strategy was referred to as simulated adjustment instead of psychological adjustment. The new term better captures the strategy’s focus on defensiveness.

2.  Detection Strategies for Malingering and Defensiveness 33

Despite its focus on denied emotional and social issues, K scale effect sizes remain comparatively modest for defensive patients and do not appear to be effective when simulators are coached (Baer & Miller, 2002). Morey’s (1996) PAI research on the Defensiveness Index (DEF) is a sophisticated attempt to identify defensive clients via the detection strategy referred to as spurious patterns of psychological adjustment. The DEF comprises two general indicators (positive impression and unwillingness to consider treatment) plus six pairwise comparisons. As an example of the latter, defensive clients tend to be more willing to endorse mildly grandiose characteristics than irritability on the PAI Mania scale. This difference is generally not observed among clients who genuinely disclose their psychological issues. Overall, this detection strategy appears to be highly effective on the DEF (Morey, 2007). Given its complexity, it appears surprisingly vulnerable to coaching for both the PAI DEF and the Cashel Discriminant Function (CDF) (Baer & Wetter, 1997). Three amplified detection strategies that have been developed for social desirability are also routinely applied to defensiveness. The simplest approach involves the general denial of faults and foibles. It is best exemplified by the MMPI-2 L (Lie) scale. As the name denotes, it was intended to measure general deception with regards to “exemplary” personal conduct. Multifaceted scales (i.e., blends of “Affirmation of Virtuous Behavior” and “Denial of Nonvirtuous Behavior”) appear to be more effective than denial alone. However, further research is needed to understand what elements of each contribute to this effectiveness. For instance, the S scale has three affirmation and two denial factors (Butcher & Han, 1995). In general, multifaceted scales appear to be moderately effective at assessing both social desirability and defensiveness. Wiggins (1959) developed a social desirability scale that bears his name (i.e., Wiggins’s Social Desirability scale [Wsd]), based conceptually on students’ ratings of individual items regarding favorable impressions. Both socially favorable (scored as true) and unfavorable (scored as false) impressions were included. This bidirectional approach is much more effective than positive-only detection strategies for simulated adjustment. With the Wsd, MMPI-2 research has demonstrated its superior effect sizes, even when individuals are coached regarding this strategy (Baer & Miller, 2002). Before leaving this section, I should note that all the detection strategies for simulated adjust-

ment fit solidly in the amplified category. This raises an important question: Could unlikely detection strategies be developed for defensiveness and social desirability? Quite possibly, the answer is “yes.” Analogous to symptom combinations, unlikely pairs of positive attributes could be identified that are very infrequent in general and clinical populations. In this vein, logical inconsistencies could be explored. For instance, thoughtfulness and decisive action can easily be construed as positive characteristics, yet they would constitute an unlikely pair of positive attributes. In summary, detection strategies for simulated adjustment lack some of the breadth and sophistication found with several malingering strategies. Still, many detection strategies have proven effective with uncoached simulators. Work on the Wsd with bidirectional—favorable and unfavorable items—demonstrate excellent results, even in the presence of coaching.

Response Styles andSpecific ClinicalDomains The development and validation of detection strategies has largely been limited to three domains (i.e., feigned mental disorders, feigned cognitive impairment, and simulated adjustment with mental disorders) described in previous sections. This section provides brief summaries of detection strategies for three other domains: (1) defensiveness and cognitive impairment, (2) malingering and medical presentations, and (3) defensiveness and medical presentations. These descriptions are very brief, because detection strategies are in their early stage of development. At present, most work on detection strategies is more conceptual than empirical. Defensiveness andCognitiveAbilities

Traditionally, practitioners have assumed that clinical populations cannot mask their cognitive weaknesses, because testing measures optimal functioning. Simply put, patients cannot do better than their best. Because of this widespread assumption, research has largely neglected whether the concealment of cognitive problems is achievable via preparation and coaching. Is this assumption accurate? Research has generally indicated that performance on standardized aptitude tests may be substantially enhanced by practice and preparation (Kulik, Bangert-Drowns, & Kulik, 1984), especially when identical tests are used. As evidence for the “benefits” of practice effects, WAIS-IV Processing Speed Index increases an av-

34

I.  Concep t ua l Fr a me work

erage of 9 points after a single readministration, even after a lengthy interval (i.e., 3 or 6 months; Estevis, Basso, & Combs, 2012). For job selection (i.e., law enforcement positions), readministrations after 12 months significantly enhanced performances on tests of both cognitive abilities and oral communication (Hausknecht, Trevor, & Farr, 2002). With the ease and availability of Webbased information, individuals can practice and prepare for enhanced performances on the cognitive measures. The other critical question remains unaddressed: “Can preparation successfully mask cognitive deficits and their concomitant impairment?” Cognitive defensiveness may involve either the masking of cognitive deficits or the false portrayal of cognitive strengths. In the former case, successful individuals may wish to conceal even minor decrements in cognitive abilities. For instance, commercial pilots may wish to mask even the slightest declines in cognitive abilities. Rebok, Li, Baker, Grabowski, and Willoughgy (2002) surveyed 1,310 airline pilots and found that almost none rated their own cognitive skills as diminished; on the contrary, a common pattern was to claim enhanced cognitive abilities. In the latter case, the false presentation of cognitive strengths may be viewed as an “asset” in securing a highly competitive position. Examples include selections for executive training, highly sought promotions, and acceptances into graduate school. In recent PsycINFO searches, no empirically validated detection strategies have been located for defensiveness and cognitive abilities. A potential detection strategy for defensiveness and cognitive abilities would be “practice-effect gains” (i.e., the lack of any improvement could be an indirect indicator of previous preparation). However, this potential strategy is very concerning, because genuine responders with marked cognitive deficits are likely to demonstrate the same pattern. Still, readministrations of brief measures should produce predictable patterns: substantial improvements on some scales and negligible differences on others. For example, Bird, Papadopoulou, Ricciardelli, Rossor, and Cipolotti (2004) found that several very brief scales evidence substantial improvement (e.g., 11.2% for verbal fluency in generating words beginning with s) whereas one subscale (i.e., Digit Symbol) evidence a slight decrement in performance (–3.2%). Conceptually based strategies, such as practice-effect gains, would need to be carefully developed and rigorously validated. The effectiveness of detection strategies must be evaluated on two dimensions: the intentional conceal-

ment of cognitive deficits and the false presentation of cognitive strengths. Malingering andMedicalPresentations

Illness behavior is far more complex than malingering per se (Halligan, Bass, & Oakley, 2003). Beyond malingered and factitious presentations, patients with chronic medical complaints can adopt one of several maladaptive responses to their illnesses. According to Radley and Green (1987), these maladaptive patterns may include accommodation and resignation. With accommodation, the illness becomes incorporated into the patient’s identity thereby complicating assessment and treatment. With resignation, patients become overwhelmed by their diseases and may passively accept their illness status. Such maladaptive responses may be mistaken for deliberate efforts by patients to malinger by prolonging their medical conditions and thwarting treatment efforts (Rogers & Payne, 2006). The importance of empirically validated detection strategies for feigned medical presentations is underscored by recent investigations of Waddell’s classic signs for nonorganic pain (see Waddell, McCulloch, Kummel, & Venner, 1980). The presence of these signs was interpreted as either malingering or psychological stress (Kiester & Duke, 1999). However, Fishbain et al. (2003) conducted a comprehensive review of Waddell’s signs and the validity of chronic pain. Despite early claims, they found that Waddell’s signs generally did not discriminate between (1) organic and nonorganic pain, and (2) genuine presentation and secondary gain.11 In a further analysis, Fishbain et al. (2004) found that Waddell’s signs do not provide credible evidence of malingering or secondary gain. Their impressive work strongly questions the use of Waddell’s signs to assess “sincerity of effort,” as touted by other investigators (Lechner, Bradbury, & Bradley, 1998). As a further complication, recent research (Lloyd, Findlay, Roberts, & Nurmikko, 2014)—contrary to the “sincerity of effort” hypothesis—has uncovered significant differences in brain circuitry via functional magnetic resonance imaging for pain patients with high versus low Waddell’s signs. Tearnan and Lewandowski’s (1997) original work on feigned medical complaints resulted in the development of the Life Assessment Questionnaire (LAQ), which was intended to evaluate preliminary detection strategies. Tearnan and Ross (2012) refined and tested a large number of detection strategies, including (1) rare symptoms,

2.  Detection Strategies for Malingering and Defensiveness 35

although the criterion appears too low (i.e., < 25% of genuine patients); (2) improbable symptoms referred to as “nonsensical symptoms”; (3) symptom combinations; and (4) physician selection of items suggesting possible malingering.12 Interestingly, these four detection strategies evidenced very similar effect sizes (ds from 0.94 to 1.03) that fell consistently in the moderate level. Although addi-

tional research is needed, this seminal work is used as a template for examining preliminary detection strategies for feigned medical complaints. Detection strategies for malingered medical presentations, similar to other domains, may be conceptualized in two general categories: unlikely and amplified detection strategies (see Table 2.5). Unlikely detection strategies, as described

TABLE 2.5.  Initial Detection Strategies for Malingered Medical Presentations

Rare medical complaints

Intensity of medical complaints

1.  Description: This strategy capitalizes on reported symptoms and ailments that are infrequently described by genuine populations. 2.  Potential strengths: Within the general category of unlikely presentations, it has received the most attention in the initial development of two specialized measures and the recent adaptation of a standardized test, the MMPI-2. The initial empirical data are promising. 3. Examples: MMPI-2 Fs (Infrequent Somatic Complaints) scale and LAQ Infrequent Symptoms.

1.  Description: This strategy relies on observations that persons malingering medical problems are likely to overstate the frequency, duration, and severity of their physical complaints. 2.  Potential strengths: It combines several parameters (e.g., frequency and severity) to create a composite strategy. Some results have produced very large effect sizes. 3.  Examples: BHI-2 Self-Disclosure scale and NSI.

Improbable medical complaints 1.  Description: This strategy is an extreme variant of rare medical complaints. It utilizes symptoms or features that have fantastic or preposterous quality. 2.  Potential strength: If established, its items call into question the genuineness of the reported complaints. 3.  Example: LAQ Nonsensical Symptoms. Symptom combinations 1.  Description: This strategy relies on complaints and symptoms that are common to medical populations but rarely occur together. Malingerers are unlikely to be aware of their low co-occurrence. 2.  Potential strength: With extensive validation, this strategy is a sophisticated approach to malingered medical presentations. 3.  Example: LAQ Unusual Symptom Combinations. Indiscriminant endorsem*nt of health problems 1.  Description: This strategy is based on the finding that some malingerers report a broad array of physical symptoms and complaints, when provided with extensive checklists. 2.  Potential strength: If systemic diseases (e.g., lupus) can be ruled out, the breadth of health-related complaints may provide excellent discrimination between genuine and malingered medical presentations. 3.  Examples: LAQ Physical Symptoms and PSI HPO (Health Problem Overstatement) scale.

Reported versus observed symptoms 1.  Description: This strategy uses marked discrepancies between the person’s own account of his or her medical complaints and corresponding observations. Malingerers can often be identified by systematic discrepancies (i.e., medical complaints unsupported by clinical observations). 2.  Potential strength: With standardized observations, this strategy provides independent verification of reported symptoms. 3.  Potential example: Comparisons of the PRS and PBC. Dependency on medical complaints 1.  Description: This strategy is based on the idea that malingerers may be willing to acknowledge positive attributes of their physical condition or disability status. 2.  Potential strength: None is noted. The willingness to acknowledge any potential motivation to malinger appears counterintuitive. 3.  Example: BHI-2 Symptom Dependency scale. Endorsem*nt of excessive virtue 1.  Description: This strategy relies on the finding that some malingerers attempt to obfuscate response style issues (e.g., malingering) by falsely claiming overly positive attributes. 2.  Potential strength: In combination with other detection strategies, it may augment the discrimination between genuine and feigned medical complaints. 3.  Example: PSI EEV (Endorsem*nt of Excessive Virtue) scale.

Note. The limitations for these initial detection strategies are not listed separately. All strategies require extensive validation and cross-validation. BHI-2, Battery for Health Improvement–2; LAQ, Life Assessment Questionnaire; NSI, Neuropsychological Symptom Inventory; PBC, Pain Behavior Checklist; PRS, Pain Rating Scale; PSI, Psychological Screening Inventory.

36

I.  Concep t ua l Fr a me work

with the LAQ, include three detection strategies. Of these, the greatest attention has been paid to rare symptoms. Empirically, Wygant, Ben-Porath, Berry, and Arbisi (2006) identified relatively uncommon symptoms on MMPI-2 protocols for more than 55,000 medical and chronic pain patients. Like Tearnan and Ross (2012), Wygant et al. (2006) adopted a lenient criterion of infrequency (i.e., < 25% of genuine protocols) for item inclusion; still, their Infrequent Somatic Complaints (Fs) scale produced large to very large effect sizes. Simulators feigning health problems as part of disability claims produced very high Fs scores, although they are also elevated on scales typically associated with feigned mental disorders (Sellbom, Wygant, & Bagby, 2012). In summary, the strategy of rare medical complaints shows strong promise for feigned medical presentations. Two other strategies for unlikely presentations (i.e., improbable medical complaints and symptom combinations) have received comparatively less attention than rare medical complaints. At present, Tearnan and Ross (2012) have demonstrated moderate effectiveness and good convergent validity. More broadly, they should be considered preliminary detection strategies. Five detection strategies for malingered medical presentations rely on amplified presentations for which the classification is based on the magnitude rather than the presence of specific indicators. The strategy, the indiscriminant endorsem*nt of health problems, is the best researched. In an early study, Furnham and Henderson (1983) found that persons feigning medical illnesses endorsed a broad range of somatic and psychological symptoms. More recently, McGuire and Shores (2001) compared simulators to chronic pain patients on the Symptom Checklist 90—Revised (SCL-90-R; Derogatis, 1992). They found that simulators were indiscriminant in their reporting of somatic and psychological symptoms, with marked elevations on each of the clinical scales.13 As a specialized scale on the PSI, Lanyon (2003) developed the Health Problem Overstatement (HPO) scale, which assesses examinees who “overstate their health problems in general” (p.2). The majority of items involve physical complaints, fatigue, and overall poor health. As strong evidence of discriminant validity, persons simulating severe physical problems on the HPO had a much higher endorsem*nt level than medical inpatients (Cohen’s d = 2.10). A second established strategy for amplified presentation is the intensity of medical complaints. This strategy is operationalized in terms of time (e.g., frequency and duration) and severity, which

is often focused on distress and impairment. For example, an adaptation of the NSI concentrates on the frequency of medical and psychological symptoms. Gelder et al. (2002) found that persons feigning neurological conditions reported frequent symptoms across a broad spectrum. Other potential detection strategies that could be tested, such as symptom combinations (e.g., ringing in the ears and decreased appetite; changes in vision and reading problems) could also be tested for their co-occurrences. This strategy could also be operationalized in terms of severity and distress. In comparing likely malingerers to patients with genuine pain, Larrabee (2003) found that the intensity of pain by itself did not effectively differentiate the groups, largely due to ceiling effects for both groups. Instead, the severity of somatic and autonomic perceptions on the Modified Somatic Perception Questionnaire (MSPQ; Main, 1983) produced very large effect sizes. However, because the MSPQ items were not developed for feigning, it is unlikely that a well-defined detection strategy will emerge. The strategy of reported versus observed symptoms parallels the detection of malingered mental disorders. With operationally defined characteristics and systematic methods, correspondence between medical complaints and health care observations can be standardized. For example, Dirks, Wunder, Kinsman, McElhinny, and Jones (1993) compared pain ratings by patients with parallel ratings by health care professionals. Patients who were deliberately exaggerating pain had very frequent discrepancies (64.6%) that were very different from those of genuine pain patients (14.2%). The chief consideration with this strategy is that there be sufficient groundwork to establish normative data on what is expected from heterogeneous medical populations that vary in pain location, frequency, and intensity. Two strategies have yielded promising results, although their conceptual basis is less precise than other strategies. The strategy, dependency on medical complaints, expects that malingerers would acknowledge the undeserved benefits of their feigning. Conceptually, this strategy appears counterintuitive. The second strategy, endorsem*nt of excessive virtue (Lanyon, 2003), does have merit, because some malingerers want to strengthen their cases for achieving their desired goals (e.g., compensation or a favorable outcome in the criminal justice system). The concern is its nonspecific nature. This strategy is potentially confounded by defensiveness (i.e., affirmation of virtuous behavior) or narcissistic personality traits. As noted in

2.  Detection Strategies for Malingering and Defensiveness 37

Table 2.5, the strategy, endorsem*nt of excessive virtue, might best be conceptualized as an ancillary strategy that could augment the discriminability of other detection strategies. In summary, the current literature provides an excellent conceptual framework for the further study of detection strategies in the medical feigning domain. Within unlikely presentations, work on rare medical complaints has demonstrated the feasibility of this detection strategy. More research is needed on improbable medical complaints and symptom combinations to augment strategies based on unlikely presentations. Within amplified presentations, considerable progress has been made with two detection strategies: indiscriminant endorsem*nt of health problems and the intensity of medical complaints. In addition, reported, as opposed to observed, symptoms show good potential, whereas the final two strategies (i.e., dependency on medical complaints and endorsem*nt of excessive virtue) may require refinement on their conceptual basis. Overall, malingered medical presentations represent a critically important domain of response styles that is poised for further scale development and empirical validation.

Defensiveness andMedicalPresentations The denial and gross minimization of medical complaints is rampant in North America and represents an immense public health problem (Kortte & Wegener, 2004). For example, the leading cause of death is cardiovascular illness (Hoyert, Heron, Murphy, & Hsiang-Ching, 2006), which is often treatable in early stages of the disease. Nonetheless, defensiveness is common in medical patients, even at the end stages of heart disease (Williams et al., 2000). Defensiveness plays a similar role with other common diseases, including cancer, diabetes, and substance abuse. It contributes to treatment noncompliance, estimated to be between 35 and 50% for chronic medical conditions; poor outcomes from untreated conditions add astronomically to health costs (Sokol, McGuigan, Verbrugge, & Epstein, 2005). The public dissemination of medical information has been an important step in increasing awareness of medical conditions and health-risk behaviors (e.g., smoking and unprotected sexual practices). However, defensiveness plays an important role in how media campaigns are processed. For example, smokers are likely to minimize or simply reject media presentations against smoking, while maintaining an “illusion of personal immunity” (Freeman, Hennessy, & Marzullo, 2001,

p.425). Health care professionals and their consultants assess many patients that actively hide their medical symptoms (e.g., failing to disclose angina) or minimize their investment in treatment (Fowers, 1992). According to Bullard (2003), 95% of patients concealed relevant information from medical staff, including symptoms and unhealthy practices (e.g., poor diet or no exercise). Most deceptions in the medical context involve concealments and equivocations rather than direct lying (Burgoon, Callister, & Hunsaker, 1994). The critical issue is whether researchers can develop detection strategies to identify those patients who actively mask their medical symptoms. Early research has demonstrated the obvious with respect to medical defensiveness. Furnham and Henderson (1983) found that defensive individuals simply did not report prominent medical symptoms; they admitted to less than half the symptoms acknowledged by a presumably healthy community sample of young adults. This finding raises an interesting question: Are some physical symptoms so common that their absence could be used to identify medical defensiveness? Lees-Haley and Brown (1993) found that patients in a group family practice often report (i.e., > 50%) headaches, fatigue, nervousness, and sleeping problems. More extensive research might uncover a predictable pattern of common physical symptoms that could be used as a potential detection strategy for medical defensiveness. A second potential approach would be the systematic evaluation of health attitudes, which may provide an unobtrusive measure of medical defensiveness. Rather than query patients about their symptoms directly, questions could focus on their attitudes toward physicians, health, and illness. For instance, does a fatalistic approach to illness predict medical defensiveness? In addition, a normative approach (e.g., “What do most people think?”) to health-related attitudes may be even less obtrusive. Extrapolating from Rogers, Vitacco, Cruise, Sewell, and Neumann, (2002), expressed beliefs about how most people view antisocial behavior were useful for identifying psychopathic youth. When youth attempted to deny their own psychopathy, their expressed attitudes about others made them more identifiable. How might a normative approach be applied to medical defensiveness? Items about general health attitudes could be tested for their discriminability between genuine disclosures and medical defensiveness: “Most persons see illness as a sign of personal weakness.” Bruns, Disorbio, and Copeland-Disorbio (2003) developed the Defensiveness scale for the Battery

38

I.  Concep t ua l Fr a me work

for Health Improvement–2 (BHI-2) to assess physically injured patients who are unwilling to disclose medical complaints and personal problems. Its detection strategy relies on general complaints and psychological problems, and does not appear to be specific to medical symptoms and physical ailments. Discriminating between genuine patients and those faking good lacks the conceptual clarity needed for distinguishing between medical defensiveness and underreporting of psychological disorders. In summary, clinical researchers have neglected medical defensiveness in their studies of response styles. Several potential strategies appear conceptually sound and deserve empirical validation. Critical to the validation is the identification of criterion groups that either deny medical symptoms or conceal their seriousness. As with other domains, detection strategies for medical defensiveness require the operationalization and systematic testing of potential methods.

SUMMARY Detection strategies provide the structural framework for the systematic assessment of response styles. Three domains (feigned mental disorders, feigned cognitive impairment, and defensiveness for mental disorders) have been subjected to intensive investigations. As a result, the developed detection strategies and concomitant scales are conceptually sound and empirically validated. Three additional domains (defensiveness and cognitive impairment, feigned medical presentations, and defensiveness and medical presentations) are conceptually based and await the intensive investigations found with the first three domains.

viations. In most response-style research, Cohen’s d should not be affected by a restricted range (Li, 2015). 4.  In contrast to psychological constructs, comparatively simple physiological measurements (e.g., pulse and blood pressure) can evidence high accuracy in terms of instrumentation but marked variability (i.e., measurement error) across time. 5.  As a related matter, a small percentage of individuals can produce markedly “enhanced performances” on cognitive measures via coaching (see, e.g., Powers & Rock, 1999). 6.  The term mental disorders is used as shorthand for the broad category of feigning psychological and emotional difficulties that may include mental disorders and syndromes. 7. Specifically, 15 of the 60 F-scale items were retained on the Fp scale, which represents a true rare symptom strategy. 8. For example, T-score transformations—but not raw scores—produce the highly significant differences on the MMPI-2 O-S scale. 9. As an important caution, some patient groups may lack the focus and attentional abilities to complete tasks over an extended period. Focusing on patients with schizophrenia, 15.2% failed the WMT without any apparent motivation to feign (Strauss, Morra, Sullivan, & Gold, 2015). 10.  Tombaugh (1996, Table 3.7, p.14) found a falsepositive rate of 27.0% for dementias, even when nonstandardized procedures (e.g., additional cuing) were used. 11. As an important clarification, Waddell never recommended the extension of his work to response styles. 12. Physician ratings of suspect symptoms do not represent a well-defined detection strategy. 13.  Simulators may have reasoned that psychological issues are related to health issues. It would be very valuable to specifically instruct simulators to feign medical complaints only, as a way of rigorously evaluating the discriminability of the Fs scale.

NOTES

REFERENCES

1. Please note that, according to the DSM-5, the presence of these two indices alone is considered sufficient cause to suspect malingering (American Psychiatric Association, 2013). 2.  The observant reader will note that only the first edition of the MCMI-III manual (Millon, 1994) was cited. Inexplicably, this valuable correlational matrix was omitted from subsequent editions (e.g., Millon, Davis, & Millon, 1997; Millon, Millon, Davis, & Grossman, 2009). 3. Cohen’s d examines the difference between two criterion groups (e.g., feigning and genuinely disordered) in standardized units based on the pooled standard de-

American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disorders (3rd ed.). Washington, DC: Author. American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: Author. Baer, R. A., & Miller, J. (2002). Underreporting of psychopathology on the MMPI-2: A meta-analytic review. Psychological Assessment, 14, 16–26. Baer, R. A., & Wetter, M. W. (1997). Effects of information about validity scales on underreporting of symptoms on the Personality Assessment Inventory. Journal of Personality Assessment, 68(2), 402–413.

2.  Detection Strategies for Malingering and Defensiveness 39

Bates, J. E. W. (2002). An examination of hangover effects on pilot performance. Dissertation Abstracts International B: The Sciences and Engineering, 62(9), 4257. Bender, S. D., & Rogers, R. (2004). Detection of neurocognitive feigning: Development of a multi-strategy assessment. Archives of Clinical Neuropsychology, 19(1), 49–60. Binder, L. M. (1993). Assessment of malingering after mild head trauma with the Portland Digit Recognition Test. Journal of Clinical and Experimental Neuropsychology, 15, 170–182. Bird, C. M., Papadopoulou, K., Ricciardelli, P., Rossor, M. N., & Cipolotti, L. (2004). Monitoring cognitive changes: Psychometric properties of six cognitive tests. British Journal of Clinical Psychology, 43, 197–210. Boone, K. B. (2007). Assessment of feigned cognitive impairment: A neuropsychological perspective. New York: Guilford Press. Borum, R., Otto, R. K., & Golding, S. (1991). Improving clinical judgment and decision making in forensic evaluation. Journal of Psychiatry and Law, 21, 35–76. Bruns, D., Disorbio, J. M., & Copeland-Disorbio, J. (2003). Manual for the Battery for Health Improvement 2. Minnetonka, MN: Pearson Assessments. Bullard, P. L. (2003). Patients’ intentional deceptions of medical professionals: Emphasizing concealment and under-reporting by purportedly healthy individuals. Dissertation Abstracts International: Section B: The Sciences and Engineering, Vol. 63(10-B), 4891. Burgoon, M., Callister, M., & Hunsaker, F. G. (1994). Patients who deceive: An empirical investigation of patient-physician communication. Journal of Language and Social Psychology, 13(4), 443–468. Butcher, J. N., & Han, K. (1995). Development of an MMPI-2 scale to assess the presentation of self in a superlative manner: The S scale. In J. N. Butcher & C. D. Spielberger (Eds.), Advances in personality assessment (Vol. 10, pp.25–50). Hillsdale, NJ: Erlbaum. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum. Dahlstrom, W. G., Welsh, G. S., & Dahlstrom, L. E. (1972). An MMPI handbook: Vol. I. Clinical interpretation (rev. ed.). Minneapolis: University of Minnesota Press. Dandachi-FitzGerald, B., Ponds, R. M., & Merten, T. (2013). Symptom validity and neuropsychological assessment: A survey of practices and beliefs of neuropsychologists in six European countries. Archives of Clinical Neuropsychology, 28(8), 771–783. Daubert v. Merrell Dow Pharmaceuticals. 509 U.S. 579, 113 S. Ct. 2786 (1993). Derogatis, L. R. (1992). SCL-90-R: Administration, scoring, and procedures manual II for the revised version. Towson, MD: Clinical Psychometric Research. DeVylder, J. E., & Hilimire, M. R. (2015). Screening for psychotic experiences: Social desirability biases in a non-clinical sample. Early Intervention in Psychiatry, 9(4), 331–334. Dirks, J. F., Wunder, J., Kinsman, R., McElhinny, J., &

Jones, N. F. (1993). A Pain Rating Scale and a Pain Behavior Checklist for clinical use: Development, norms, and the consistency score. Psychotherapy and Psychosomatics, 59, 41–49. Dwyer, J., & Reid, S. (2004). Ganser’s syndrome. Lancet, 364(9432), 471–473. Elhai, J. D., Ruggiero, K. J., Frueh, B. C., Beckham, J. C., & Gold, P. B. (2002). The Infrequency-Posttraumatic Stress Disorder Scale (Fptsd) for the MMPI-2: Development and initial validation with veterans presenting with combat-related PTSD. Journal of Personality Assessment, 79(3), 531–549. Erber, J. T., & Prager, I. G. (2000). Age and excuses for forgetting: Self-handicapping versus damage-control strategies. International Journal of Aging and Human Development, 50(3), 201–214. Estevis, E., Basso, M. R., & Combs, D. (2012). Effects of practice on the Wechsler Adult Intelligence Scale– IV across 3- and 6-month intervals. Clinical Neuropsychologist, 26(2), 239–254. Ferguson, C. J. (2009). An effect size primer: A guide for clinicians and researchers. Professional Psychology: Research and Practice, 40(5), 532–538. Fishbain, D. A., Cole, B., Cutler, R. B., Lewis, J., Rosomoff, H. L., & Rosomoff, R. S. (2003). A structured evidence-based review on the meaning of nonorganic physical signs: Waddell signs. American Academy of Pain Medicine, 4, 141–181. Fishbain, D. A., Cutler, R. B., Rosomoff, H. L., & Rosomoff, R. S. (2004). Is there a relationship between nonorganic physical findings (Waddell Signs) and secondary gain/malingering? Clinical Journal of Pain, 20(6), 399–408. Fowers, B. J. (1992). The cardiac denial of impact scale: A brief, self-report research measure. Journal of Psychosomatic Research, 36(5), 469–475. Frederick, R. I. (1997). The Validity Indicator Profile. Minneapolis, MN: National Computer Systems. Frederick, R. I. (2003). The Validity Indicator Profile (2nd ed.). Minneapolis, MN: National Computer Systems. Frederick, R. (2015, March). Too much information: Problems when using multiple malingering tests [Invited lecture]. American Psychology-Law Conference, San Diego, CA. Freeman, M. A., Hennessy, E. V., & Marzullo, D. M. (2001). Defensive evaluation of antismoking messages among college-age smokers: The role of possible selves. Health Psychology, 20, 424–433. Furnham, A., & Henderson, M. (1983). Response bias in self-report measures of general health. Personality and Individual Differences, 4, 519–525. Gelder, B. C., Titus, J. B., & Dean, R. S. (2002). The efficacy of Neuropsychological Symptom Inventory in the differential diagnosis of medical, psychiatric, and malingering patients. International Journal of Neuroscience, 112(11), 1377–1394. Goldstein, H. (1945). A malingering key for mental tests. Psychological Bulletin, 42, 104–118. Green, P. (2003). The Word Memory Test. Seattle, WA: Green’s.

40

I .  C o n c e p t u a l F r a m e w o r k

Green, P., Astner, K., & Allen, L. M., III. (1997, November). Validation of the Word Memory Test in 400 consecutive compensation claimants, including 196 cases of head injury. Paper presented at the National Academy of Neuropsychology, Las Vegas, NV. Greene, R. L. (2000). The MMPI-2: An interpretive manual. Boston: Allyn & Bacon. Greene, R. L. (2011). The MMPI-2/MMPI-2-RF: An interpretive manual. Boston: Allyn & Bacon. Haines, M. E., & Norris, M. P. (1995). Detecting the malingering of cognitive deficits: An update. Neuropsychology Review, 5(2), 125–148. Halligan, P. W., Bass, C., & Oakley, D. A. (2003). Malingering and illness deception. Oxford, UK: Oxford University Press. Hartshorne, H., & May, M. A. (1928). Studies in deceit. New York: Macmillan. Hathaway, S. R., & McKinley, J. C. (1940). A multiphasic personality schedule (Minnesota): I. Construction of the schedule. Journal of Psychology, 10, 249–254. Hausknecht, J. P., Trevor, C. O., & Farr, J. L. (2002). Retaking ability tests in a selection setting: Implications for practice effects, training performance, and turnover. Journal of Applied Psychology, 87, 243–254. Hoyert, D. L., Heron, M. P., Murphy, S. L., & HsiangChing, K. (2006). Deaths: Final data for 2003. National Vital Statistics Reports, 54, 1–120. Kiester, P. D., & Duke, A. D. (1999). Is it malingering, or is it “real”?: Eight signs that point to nonorganic back pain. Postgraduate Medicine, 106, 77–80, 83–84. Kingery, L. R., & Schretlen, D. J. (2007). Functional neuroimaging of deception and malingering. In K. B. Boone (Ed.), Assessment of feigned cognitive impairment: A neuropsychological perspective (pp.13–25). New York: Guilford Press. Kortte, K. B., & Wegener, S. T. (2004). Denial of illness in medical rehabilitation populations: Theory, research, and definition. Rehabilitation Psychology, 49, 187–199. Kulik, J. A., Bangert-Drowns, R. L., & Kulik, C. C. (1984). Effective of coaching on aptitude tests. Psychological Bulletin, 95, 179–188. Lanyon, R. I. (1970). Development and validation of a Psychological Screening Inventory. Journal of Consulting and Clinical Psychology, 35(1, Pt.2), 1–24. Lanyon, R. I. (1993). Development of scales to assess specific deception strategies on the Psychological Screening Inventory. Psychological Assessment, 5, 324–329. Lanyon, R. I. (2003). Assessing the misrepresentation of health problems. Journal of Personality Assessment, 81, 1–10. Larrabee, G. J. (2003). Detection of malingering using atypical performance patterns on standard neuropsychological tests. Clinical Neuropsychologist, 17, 410–425. Larrabee, G. J. (2008). Aggregation across multiple indicators improves the detection of malingering: Relationship to likelihood ratios. Clinical Neuropsychologist, 22(4), 666–679.

Lechner, D. E., Bradbury, S. F., & Bradley, L. A. (1998). Detecting sincerity of effort: A summary of methods and approaches. Physical Therapy, 78, 867–888. Lees-Haley, P. R., & Brown, R. S. (1993). Neuropsychological complaint base rates of 170 personal injury claimants. Archives of Clinical Neuropsychology, 8, 203–209. Li, J. C. (2015). Cohen’s d corrected for CASE IV range restriction: A more accurate procedure for evaluating subgroup differences in organizational research. Personnel Psychology, 68(4), 899–927. Liff, C. A. (2004). The detection of neuropsychological malingering. Dissertation Abstracts International, 64, 4623. Lloyd, D. M., Findlay, G., Roberts, N., & Nurmikko, T. (2014). Illness behavior in patients with chronic low back pain and activation of the affective circuitry of the brain. Psychosomatic Medicine, 76(6), 413–421. Main, C. J. (1983). Modified Somatic Perception Questionnaire. Journal of Psychosomatic Research, 27, 503–514. McGuire, B. E., & Shores, E. A. (2001). Simulated pain on the Symptom Checklist 90-Revised. Journal of Clinical Psychology, 57, 1589–1596. McKinley, J. C., Hathaway, S. R., & Meehl, P. E. (1948). The Minnesota Multiphasic Personality Inventory: VI. The K Scale. Journal of Consulting Psychology, 12(1), 20–31. Meehl, P. E. (1946). Profile analysis of the Minnesota Multiphasic Personality Inventory in differential ­diagnosis. Journal of Applied Psychology, 30(5), 517– 524. Millon, T. (1994). The Millon Clinical Multiaxial Inventory-III manual. Minneapolis, MN: National Computer Systems. Millon, T., Davis, R., & Millon, C. (1997). The Millon Clinical Multiaxial Inventory–III manual (2nd ed.). Minneapolis, MN: National Computer Systems. Millon, T., Millon, C., Davis, R., & Grossman, S. (2009). The Millon Clinical Multiaxial Inventory–III manual (4th ed.). Bloomington, MN: Pearson. Mittenberg, W., Theroux-Fichera, S., Zielinski, R., & Heilbronner, R. L. (1995). Identification of malingered head injury on the Wechsler Adult Intelligence Scale—Revised. Professional Psychology: Research and Practice, 26(5), 491–498. Morey, L. C. (1996). An interpretive guide to the Personality Assessment Inventory. Lutz, FL: Psychological Assessment Resources. Morey, L. C. (2007). The Personality Assessment Inventory professional manual (2nd ed.). Lutz, FL: Psychological Assessment Resources. Pankratz, L., Fausti, S. A., & Peed, S. (1975). A forcedchoice technique to evaluate deafness in the hysterical or malingering patient. Journal of Consulting and Clinical Psychology, 43(3), 421–422. Powers, D. E., & Rock, D. A. (1999). Effects of coaching on SAT I: Reasoning Test scores. Journal of Educational Measurement, 36(2), 93–118. Radley, A., & Green, R. (1987). Illness as adjustment: A

2.  Detection Strategies for Malingering and Defensiveness 41

methodology and conceptual framework. Sociology of Health and Illness, 9(2), 179–207. Rebok, G. W., Li, G., Baker, S. P., Grabowski, J. G., & Willoughgy, S. (2002). Self-rated changes in cognition and piloting skills: A comparison of younger and older airline pilots. Aviation, Space, and Environmental Medicine, 73, 466–471. Rogers, R. (1984). Towards an empirical model of malingering and deception. Behavioral Sciences and the Law, 2, 93–112. Rogers, R. (1990). Development of a new classificatory model of malingering. American Academy of Psychiatry and Law, 18, 323–333. Rogers, R. (2008). Detection strategies for malingering and defensiveness. In R. Rogers (Ed.), Clinical assessment of malingering and deception (3rd ed., pp.14–35). New York: Guilford Press. Rogers, R., & Bender, S. D. (2013). Evaluation of malingering and related response styles. In R. K. Otto, I. B. Weiner Handbook of psychology: Vol. 11. Forensic psychology (2nd ed., pp.517–540). Hoboken, NJ: Wiley. Rogers, R., & Correa, A. A. (2008). Determinations of malingering: Evolution from case-based methods to detection strategies. Psychiatry, Psychology and Law, 15(2), 213–223. Rogers, R., & Gillard, N. D. (2011). Research methods for the assessment of malingering. In B. Rosenfeld & S. D. Penrod (Eds.), Research methods in forensic psychology (pp.174–188). New York: Wiley. Rogers, R., Gillard, N. D., Wooley, C. N., & Ross, C. A. (2012). The detection of feigned disabilities: The effectiveness of the Personality Assessment Inventory in a traumatized inpatient sample. Assessment, 19(1), 77–88. Rogers, R., Harrell, E. H., & Liff, C. D. (1993). Feigning neuropsychological impairment: A critical review of methodological and clinical considerations. Clinical Psychology Review, 13(3), 255–274. Rogers, R., Jackson, R. L., Sewell, K. W., & Salekin, K. L. (2005). Detection strategies for malingering: A confirmatory factor analysis of the SIRS. Criminal Justice and Behavior, 32(5), 511–525. Rogers, R., & Payne, J. W. (2006). Damages and rewards: Assessment of malingered disorders in compensation cases. Behavioral Sciences and the Law, 24, 645–658. Rogers, R., Robinson, E. V., & Gillard, N. D. (2014). The SIMS screen for feigned mental disorders: The development of detection-based scales. Behavioral Sciences and the Law, 32, 455–466. Rogers, R., & Sewell, K. W. (2006). MMPI-2 at the crossroads: Aging technology or radical retrofitting? Journal of Personality Assessment, 87, 175–178. Rogers, R., Sewell, K. W., & Gillard, N. D. (2010). Structured Interview of Reported Symptoms–2 (SIRS-2) and professional manual. Lutz, FL: Psychological Assessment Resources. Rogers, R., Sewell, K. W., Martin, M. A., & Vitacco, M. J. (2003). Detection of feigned mental disorders: A meta-analysis of the MMPI-2 and malingering. Assessment, 10, 160–177.

Rogers, R., Vitacco, M. J., Cruise, K. R., Sewell, K. W., & Neumann, C. S. (2002). Screening for adolescent psychopathy among at-risk youth: Initial validation of the Survey of Attitudes and Life Experiences (SALE). Assessment, 9, 343–350. Sellbom, M., Wygant, D., & Bagby, M. (2012). Utility of the MMPI-2-RF in detecting non-credible somatic complaints. Psychiatry Research, 197(3), 295–301. Sokol, M. C., McGuigan, K. A., Verbrugge, R. R., & Epstein, R. S. (2005). Impact of medication adherence on hospitalization risk and healthcare cost. Medical Care, 43, 521–530. Strauss, G. P., Morra, L. F., Sullivan, S. K., & Gold, J. M. (2015). The role of low cognitive effort and negative symptoms in neuropsychological impairment in schizophrenia. Neuropsychology, 29(2), 282–291. Streiner, D. L. (2003). Diagnosing tests: Using and misusing diagnostic and screening tests. Journal of Personality Assessment, 81, 209–219. Sullivan, K., Deffenti, C., & Keane, B. (2002). Malingering on the RAVLT: Part II. Detection strategies. Archives of Clinical Neuropsychology, 17(3), 223–233. Tearnan, B. H., & Lewandowski, M. J. (1997). The Life Assessment Questionnaire (LAQ). Unpublished test, Pendrake, Inc., Reno, NV. Tearnan, B. H., & Ross, S. A. (2012). The development and classification accuracy of the Life Assessment Questionnaire in the detection of pain-related malingering. Behavioral Sciences and the Law, 30(4), 516–536. Tombaugh, T. N. (1996). Test of Memory Malingering (TOMM). Toronto: Multi-Health Systems, Inc. Waddell, G., McCulloch, J. A., Kummel, E., & Venner, R. M. (1980). Nonorganic physical signs in low-back pain. Spine, 5, 117–125. Wechsler, D. (2008). Wechsler Adult Intelligence Scale (4th ed.). San Antonio, TX: Pearson Assessment. Widows, M. R., & Smith, G. P. (2005). SIMS: Structured Inventory of Malingered Symptomatology: Professional manual. Lutz, FL: Psychological Assessment Resources. Wiggins, J. S. (1959). Interrelationships among MMPI measures of dissimulation under standard and social desirability instructions. Journal of Consulting Psychology, 23, 419–427. Wilkinson, L., & Task Force on Statistical Inference. (1999). Statistical methods in psychology journals: Guidelines and explanations. American Psychologist, 54, 594–604. Williams, M. A., Putzke, J. D., LaMarche, J. A., Bourge, R. C., Kirklin, J. K., McGiffin, D. C., et al. (2000). Psychological defensiveness among heart transplant candidates. Journal of Clinical Psychology in Medical Settings, 7(3), 167–174. Wygant, D. B., Ben-Porath, Y. S., Berry, D. T. R., & Arbisi, P. A. (2006, March). An MMPI-2 valid­ ity scale designed to detect somatic over-reporting in civil forensic settings. Paper presented at the American Psy­chology–Law Society conference, St. Pete’s Beach, FL.

CHAPTER 3

Neuropsychological Models ofFeigned Cognitive Deficits ScottD.Bender,PhD RichardFrederick,PhD

Valid neuropsychological conclusions require that the examinee’s symptoms and test performance are themselves valid. Because malingering must be ruled out as a possible explanation for an examinee’s presentation, many tests and methods have been developed to assess performance and symptom validity, under the basic assumption that such measures can detect malingering. Our primary aim in this chapter is to review and critique the current state of neuropsychological malingering detection, including recent advances in probabilistic modeling. Though the majority of the chapter focuses on indicators of feigned impairment, we review salient markers of genuine symptomatology as well. A definition specific to malingered neurocognitive dysfunction (MND; Slick, Sherman, & Iverson, 1999) and the development of detection models involving performance validity have not only advanced our understanding of the construct of malingering but have also served as reminders of how much we still do not know. For instance, which detection strategies work best and for which types of symptoms? Does the simple presence of a substantial incentive approximate closely enough the individual’s true motivations to warrant its use as a proxy for motivation? While possible answers to the first question have emerged with research, the answer to the second has eluded researchers, at least partially, because it has not been subjected to 42

sufficient critical analysis. This chapter critically reviews recent developments within neuropsychological malingering detection, focusing on Slick and Sherman’s (2013) revised diagnostic criteria for MND, symptom validity testing (SVT), and performance validity testing (PVT). In addition, the utility of using multiple tests (i.e., chaining likelihoods) to identify malingered performance is considered. In 1978, Heaton, Smith, Lehman, and Vogt published what is now a long-famous article suggesting that neuropsychologists cannot reliably detect malingered cognitive test performance. The diagnostic accuracy of the study’s 10 neuropsychologists, who reviewed test data blindly, ranged from chance level to roughly 20% above chance level. Though the article was faulted for not including malingering as one of the diagnostic possibilities, it was the first to raise the issue; indeed, no references to previous published work on malingered cognitive deficits appear in the Heaton et al. (1978) study. Since then, however, thousands of studies of malingered neuropsychological deficits have been published (Martin, Shroeder, & Odland, 2015). Unfortunately, research investigating conceptual assumptions regarding feigned impairment and its detection has not kept pace with that of test developers. Before examining current neuropsychological models of feigned cognitive defi-

3. Neuropsychological Models of Feigned Cognitive Deficits 43

cits, we begin with two definitions of malingering, one more general and well-known, and the other specific to cognitive impairment.

DEFINITIONS Malingering As has been the case since DSM-III (American Psychiatric Association, 1980), DSM-5 (American Psychiatric Association, 2013) defines malingering as the intentional fabrication or gross exaggeration of symptoms. The concept of motivation was not included in the early definition; instead, the individual’s “circ*mstances” were emphasized. The second half of the definition states that the exaggeration or fabrication must be motivated by external incentives. In theory, the terms intentional and external in the current definition are key to differentiating malingering from other conditions—for instance, intentionality is present in malingering but is absent in somatic symptom disorders. And, any gain that is present must be characterized as external in malingering but internal in factitious disorder. In short, motivation is at the heart of the definition. Unfortunately, the definition of malingering seems to be stronger in theory than in practice, because such clean demarcations rarely occur in real life. For instance, arguably, the most critical aspect of the definition of malingering (i.e., whether or not the fabrication of gross exaggeration was truly motivated by external incentive) cannot be determined with absolute confidence. The existence of such unknowns is not unique to malingering detection, but the consequence of being wrong with regard to malingering may be devastatingly harmful (Drob, Meehan, & Waxman, 2009; Rogers & Bender, 2013). Misdiagnosing social anxiety as generalized anxiety disorder before sending the patient for treatment, while not ideal, would very likely not put the patient’s wellbeing at risk. The patient would still qualify for treatment, and treatment would be similar to that received otherwise. Being falsely labeled a malingerer, in sharp contrast, could result in profound and irreversible losses. DSM-5 (American Psychiatric Association, 2013) conceptualization of malingering states that avoiding work or military duty, obtaining financial compensation, evading criminal prosecution, or obtaining drugs are examples of external incentives that may lead to malingering. However, it also acknowledges that malingering can be adaptive, and even principled, in some situations (e.g.,

feigning illness while held captive in a time of war or faking symptoms to divert unwanted attention away from another person). DSM-5 cautions practitioners to suspect malingering for two or more of the following: (1) a forensic context, (2) major differences between subjective reports and objective data, (3) the examinee’s (or client’s) lack of cooperation, and (4) an antisocial personality disorder (ASPD) diagnosis. While intuitively appealing, combining any two of the four criteria leads to an unacceptably high false-positive rate (Rogers, 1990). Also, at least some of the criteria, while certainly present in some cases, do not actually appear to be closely associated with malingering. ASPD, for example, is not necessarily associated with an increased incidence of malingering (Niesten, Nentjes, Merckelbach, & Bernstein, 2015; Sumanti, Boone, Savodnik, & Gorsuch, 2006). This serves as a reminder that a potentially common characteristic is not necessarily a distinguishing characteristic. Simple reliance on common characteristics (see Rogers, Chapter 2, this volume) will inevitably result in unnecessary false-positive errors.

Malingered NeurocognitiveDysfunction Greiffenstein, Baker, and Gola (1994) appear to be the first to have proposed malingering diagnostic criteria for use in neuropsychological settings, specifically regarding memory dysfunction. But their article lacked a clear definition of malingering, certain terms were vague, and the criteria referred to memory loss only. In 1999, Slick et al. proposed a more comprehensive set of criteria specific to malingered cognitive dysfunction. The criteria became widely accepted (Heilbronner et al., 2009), subsequently became known as “the Slick criteria,” and fueled an explosion of malingering test development. Slick et al. (1999) defined malingered cognitive dysfunction as the “volitional exaggeration or fabrication of cognitive dysfunction for the purpose of obtaining substantial material gain, or avoiding or escaping formal duty or responsibility” (p.552). The authors acknowledge the difficulty inherent in determining intention and motivation, and note that knowing these internal states in someone else may be impossible. In light of this problem and consistent with Rogers’s (1990) recommendation to use gradations of certainty when referring to malingering, the Slick criteria (1999) included possible, probable, and definite malingering as indications of professionals’ confidence in the diagnosis.

44

I.  Concep t ua l Fr a me work

The Slick criteria (1999) articulated a formal and comprehensive approach to the classification of MND. They were intended to (1) provide an explicit definition of malingering, (2) specify rule-out conditions, (3) include behavioral observations, (4) fully specify criteria, and (5) provide guidelines for the evaluation of neurocognitive domains other than memory. The Slick criteria have been endorsed by the American Academy of Clinical Neuropsychology as “a reliable means of operationalizing diagnostic decisions related to the determination of malingering” (Heilbronner, et al., 2009, p.1098). Nevertheless, the Slick criteria have only been proposed as a framework for identifying malingering. In the original Slick criteria, definite malingering is identified by (1) incentive to malinger plus (2) “below-chance” performance on a forced-choice procedure. Below-chance performance is seen when test-takers score significantly lower than 50% on tasks in which choices determined by a coin flip alone would yield a score near 50% (Frederick & Speed, 2007). The classification of below-chance performance is typically reserved for those instances in which only 5% or fewer of nonmalingerers would produce the below-chance score at or below the designated level. Probable malingering is identified by “two or more types of evidence from neuropsychological testing, excluding definite negative response bias” (Slick et al., 1999, p.552). Types of evidence include positive test scores (consistent with malingering) on “wellvalidated” psychometric tests and “discrepancies” between test data or self-report and known patterns of brain functioning, observed behavior, collateral reports, or documented background history. Two positive test scores would minimally meet the criteria for probable malingering, absent any other evidence. Positive scores on validity scales of personality tests such as the Minnesota Multiphasic Personality Inventory–2 (MMPI-2) or MMPI-2 Restructured Form (MMPI-2-RF) are also included in the analysis. Possible malingering is identified by substantial external incentive and just one positive element of the probable malingering category elements. Consequently, when using the Slick et al. (1999) criteria, a single positive test score of feigning is often sufficient to generate classifications of probable or possible MND. Limitations totheSlickCriteria

The Slick criteria have been central to an operationalized understanding of neurocognitive malingering, and their influence on the current state of

malingering detection cannot be overstated. Yet, as would be expected in any set of preliminary criteria regarding such a complex construct as malingering, several problems with the criteria emerged. For instance, Larrabee, Greiffenstein, Greve, and Bianchini (2007), while generally affirming the importance of the MND model, recommended improvements, especially with regard to determinations of probable MND (e.g., they recommended the use of an aggregate of malingering indicators). Boone (2007), in contrast, was more critical of the MND model, most notably regarding the nonspecificity of the exclusion criteria. Rogers, Bender, and Johnson (2011) were similarly critical, and noted that the criteria were potentially biased toward findings of malingering. Delis and Wetter (2007) underscored the difficulty inherent in determining intentionality as required by criterion A of the MND model. They observed that “intentionality is likely multifactorial in nature” (p.592) and “practitioners may not have access to sufficient background information about a person’s life to assess if external incentives are operative in the case” (p.594). In short, their analysis highlighted the criterion’s problematic tendency to equate possible incentive with true motivation. Whether psychometric tests are “well-validated” was not defined by Slick et al. (1999), and there is no consensus document from any organization that lists which tests are designated “well validated,” which are “adequately validated,” or which are “experimental.” Also, none of the Slick criteria have been empirically evaluated for their accuracy in classifying malingering. In recognition of these issues, Slick and Sherman (2013) revised the MND criteria. Though the definition of neurocognitive malingering was not changed, Slick and Sherman did propose modifications to important constructs, which we review here briefly. For example, below-chance scores are no longer necessary to diagnose definite malingering. As an important change, rule-out criteria do not need to be “fully” accounted for by psychiatric, neurological, or developmental factors; the new criteria use the term substantially instead of fully. The first of these two changes would appear to increase sensitivity, while the latter change seems to improve specificity. The category of possible malingering has been dropped altogether, apparently in recognition of its lack of diagnostic utility (see Rogers, Bender, et al., 2011). The revised Slick and Sherman (2013) criteria acknowledge the possibility that malingering may occur due to neurocognitive compromise (second-

3. Neuropsychological Models of Feigned Cognitive Deficits 45

ary malingering), and add both “compelling inconsistencies” (see Bianchini, Greve, & Glynn, 2005; Larrabee et al., 2007) and “posterior probabilities” (Larrabee, 2008) to the list of methods that can be used to detect malingered performance. In addition, it concedes that there are fundamental limitations to the concept of secondary gain. These additions to the criteria are reviewed briefly below. The use of compelling inconsistencies as potential indications of feigning was espoused by Bianchini et al. (2005) to facilitate identification of malingered pain. Such inconsistencies can be qualitative—an examinee’s complaint of severe and incapacitating memory loss as compared to intact memory observed during the interview, for example. Performance evaluations, collateral reports, and surveillance video that contradict the claimant’s report fall under this category as well. Several examples of what appear to be more quantitative compelling inconsistencies are included in the Slick and Sherman (2013) criteria; they use the term “marked and implausible discrepancies” to refer to compelling inconsistencies that are central to six specific criteria (criteria 3.2–3.7, p.63). Slick and Sherman (2013) also recommend establishing posterior probabilities, that is, using multiple indicators of poor effort or symptom exaggeration in aggregate to aid detection. A growing body of research suggests that this approach can increase diagnostic certainty above that associated with single indicators. However, the particular mechanics and utility of this procedure are still being debated. Finally, the term secondary gain has typically been used to distinguish incentives coming from internal mechanisms (primary gain) from those involving external or material motivations. But as noted by Rogers and Reinhardt (1998), the term is often confusing due to its multiple context-dependent meanings. Slick and Sherman (2013) seem to acknowledge this in their proposal of the new diagnostic category of adjustment problem/disorder with specious symptoms, for use “in cases in which a person exaggerates or fabricates symptoms to obtain psychosocial secondary gains, rather than material–legal secondary gains” (p.68). This is a potentially important distinction, but its clinical utility has yet to be tested. While many of the modifications to the original Slick criteria likely represent significant advances, they require validation as a unitary set of criteria for malingering. A distillation of the substantive changes to the 1999 criteria may be found in Table 3.1.

SYMPTOM VALIDITYTESTING The term symptom validity testing (SVT) originally referred to the forced-choice paradigm to which the binomial theorem could be applied to identify significantly below-chance performance (Pankratz, 1979; Bender & Rogers, 2004). Over time, SVT became synonymous with effort testing and validity scales in general, with no assumptions about significantly below-chance performance. Under this rubric, many tests of cognitive effort and validity scales were developed. However, more recently, SVT has been used to refer to tests that detect feigned psychiatric symptoms only, in an effort to distinguish them from tests developed to detect neurocognitive dysfunction. This partitioning is a clinically important one given that symptom exaggeration and poor cognitive effort are not always correlated, and that tests have been shown not to be equally effective across these domains (Rogers, Gillard, Berry, & Granacher, 2011). For a review of common SVTs used in forensic evaluations, see Rogers and Bender (2013), Young (2014), and Bender (Chapter 7, this volume).

PERFORMANCE VALIDITYTESTING Larrabee (2012) is credited with suggesting that neuropsychologists distinguish between tests that identify cognitive feigning and those used to detect feigned psychopathology, giving rise to the term performance validity testing (PVT). So-called “stand-alone” PVTs were designed expressly to detect poor effort during neurocognitive assessment. For reasons primarily involving ease (of both test development and clinical implementation), most of these tests have employed the floor effect strategy. A major strength of this strategy is that it is conducive to forced-choice formats, which allow for statistical comparisons to chance levels of performance, thereby potentially providing highly specific findings. The floor effect strategy capitalizes on the use of tests that appear to be more difficult than they actually are. Poor performance on such a test, a test on which even cognitively impaired individuals have been shown to perform quite well, suggests poor effort and possibly feigning (Schutte, Axelrod, & Montoya, 2015). While norms for those with severe traumatic brain injury (TBI) and a handful of other clinical populations have been established for comparison purposes, studies establishing norms for multiple comorbid conditions have been widely lacking for these

46

I.  Concep t ua l Fr a me work

TABLE 3.1.  Summary Table of Major Changes to Slick Criteria for MND

Slick, Sherman, and Iverson (1999)

Slick and Sherman (2013)

Specific language of 1999 criteria

Criterion number

How issue was addressed in2013 criteria

Minimal level of certainty

Reliance on information that simply “suggests” exaggeration or fabrication

B4, B5, and C1

References to “suggest” have been removed

Minimal criteria for probable MND

≥ 2 criteria (≥ 20% of 10 criteria)

B2–B6 and C1–C5

Criteria now require three or more indicators of exaggeration/ fabrication (criterion 2)

Use possible incentive as sufficient for MND

External incentive is equated with motivation

A

Not addressed (criterion 1)

Denial of past history “proves” MND

Denial of psychiatric history or premorbid dysfunction

C1 and C4

No substantial change (criterion 3.5)

Feigning in a noncognitive domain still “proves” MND

“Exaggerated or fabricated psychological dysfunction”

C5

No longer a specific reference to psychological symptoms

Maximal level of certainty

“Fully accounted for” (100%)

D

“Fully accounted for” is now “substantially accounted for” (“definite” criterion 3)

High criteria for ruling-out motivation

“Significantly diminished capacity to appreciate laws and mores” or “inability to conform behavior”

D

Problematic language removed from 2012 criteria

Poor insight and impairment are not considered

Self-reported symptoms are discrepant

C2 and C3

“Secondary MND” added to criteria

Problems with 1999 criteria

Note. In a critical analysis of the Slick et al. (1999) criteria for MND, Rogers, Bender, and Johnson (2011) pointed out several problems with the criteria, including the specific language in question, and made recommendations to improve the criteria. This table reflects some of those issues and how they were addressed in Slick and Sherman (2013).

tests. The virtual neglect of comorbidity poses a formidable problem, in light of the fact that few examinees present with a single set of symptoms or a single diagnosis. Embedded PVTs are tests that, although not developed to detect poor effort, were later utilized to this end. The strength of these tests rests in their time-effectiveness and dual purpose. However, because they tend to measure actual ability, at least to some degree, they are often less specific to the effects of poor effort alone. For example, using traditional cutoff scores for the Reliable Digit Span (RDS) from the Wechsler Adult Intelligence Scale–IV (WAIS-IV) incorrectly classifies genuine patients with dementia as feigning (Zenisek, Millis, Banks, & Miller, 2016). While not a conceptually based strategy per se, research has demonstrated the potential of aggre-

gating PVTs to improve classificatory accuracy. For instance, in a known-groups design (26 litigating/ compensation-seeking participants who scored below chance levels vs. 31 patients with moderate to severe TBI), Larrabee (2008) found that posttest probabilities of malingering increased markedly when three PVTs were failed as compared to one PVT failure. The pretest odds of each test correctly classifying a performance as feigned were set at .667 (which are the odds associated with a base rate of 40%). When the score from each effort test was combined with the others, posttest probabilities increased from 60–70 to 98% (at base rates of 30% or above). Whiteside et al. (2015) used a similar aggregation strategy via logistic regression to establish a “cross-domain, logistically-derived embedded PVT (CLEP)” (p.795). The criterion group consisted of

3. Neuropsychological Models of Feigned Cognitive Deficits 47

67 patients with mild TBI who failed two or more PVTs, while the clinical group comprised 66 patients with severe TBI. From a battery of tests, three scores entered the regression and yielded superior classification to that of each test alone, according to AUC (area under the curve) analyses. While promising as a detection strategy, it remains unknown whether other combinations of tests from the battery hold utility, and CLEP would not be helpful in cases in which a malingerer fakes a deficit in only one domain. Multiple assumptions must be made when using these complex statistical models, and the devil may be in the details, as we discuss later in this chapter.

EffortTesting For many years, malingering was thought to be determinable by experienced neuropsychologists, with no need of effort test data. In contrast, probabilistic methods involving effort tests are now considered a critical part of neuropsychological testing, particularly in forensic settings. The major governing bodies within neuropsychology have issued statements about the importance of effort testing when assessing malingering, and have indicated that the absence of effort testing must be justified (Bush et al., 2005; Heilbronner et al., 2009). The move from subjective judgment to actuarial decision making represents an important improvement. Yet despite an abundance of effort tests, regrettably few employ conceptually based detection strategies. The original rationale for formal effort testing likely began with a recognition that without sufficient examinee effort, neuropsychological test scores are invalid and uninterpretable. Over time, and despite cautions against drawing equivalences, effort tests have become more and more synonymous with malingering detection. Current tests of effort, as used in neuropsychological assessment, do not measure the full spectrum of effort. Rather, they identify poor effort only, without any regard to the cause or multiple causes of poor effort. A “normal” score does not necessarily indicate good effort, and a low score does not imply a particular cause, such as feigning. The importance of this point is highlighted in Box 3.1. As Millis (2008) stated, there is a “pressing need to determine how effort tests relate to each other and to standard cognitive measures” (p.897). In the years since his call for a better understanding of intertest relationships, the question remains largely unanswered (see Larrabee, 2014, for an example of progress in this area).

BOX 3.1.  Twin Fallacies about Effort Testing andMalingering 1. Rule-out fallacy: “Normal” scores do not preclude poor effort or feigning. 2. Rule-in fallacy: Poor effort cannot be equated with feigning.

Base Rates ofMalingering A base rate (or prevalence) can be defined simply as the percentage of a group that possesses a given characteristic (Elwood, 1993). Large epidemiological studies are needed for precise estimates. Base rates are instrumental to establishing the utility of a test; the test must prove to be superior to chance, and the odds of a correct determination must incorporate the probability of making that determination by base rate alone (i.e., with no test data). Despite what might appear to be well-established base rates for malingered mild TBI (mTBI), precise base rate estimates are probably impossible to obtain, and current estimates may be too high. Mittenberg, Patton, Canyock, and Condit (2002) reported a base rate of 38.5%, and Larrabee (2003) reported a rate of 40%. It bears noting, however, that the Mittenberg et al. (2002) data are (1) from a survey and (2) asked American Board of Clinical Neurophysiology (ABCN) diplomates to estimate the percentage of their annual cases that involved “probable symptom exaggeration or malingering” (p.1101, emphasis added), which incorrectly suggests that the two terms are interchangeable, thereby inflating the estimate of actual malingering. Larrabee’s (2003) 40% base rate estimate is the mean of 11 past studies with base rate estimates ranging from 15 to 64%. Although Larrabee’s estimate has an element of empiricism that surveys do not, it included studies utilizing the highly problematic differential prevalence design (i.e., simple presence of litigation; see Rogers, Chapter 1, this volume) to designate “motivated performance deficit suggestive of malingering” (p.411). Other surveys are markedly discrepant with the previously mentioned base rates. For example, Sharland and Gfeller’s (2007) survey of 188 members of the National Academy of Neuropsychology (NAN) yielded highly variable rates for probable malingering and definite malingering that depended on the wording of the question. In litigants, the median estimate of deliberate exaggeration of “cognitive impairment” was 20%, but

48

I.  Concep t ua l Fr a me work

the median for “definite malingering” was only 1% (Sharland & Gfeller, 2007; see Table 3, p.216). In addition, Slick, Tan, Strauss, and Hultsch (2004) found that only 12.5% of their group of experts believed that definite malingering occurs in over 30% of cases. In stark contrast to Mittenberg et al. (2002), roughly two-thirds reported that “definite malingering” occurs in 20% or less of cases. Base rate estimates of neurocognitive malingering are based on (1) an improbably low test score suggestive of poor effort, and (2) a presumption not only that the poor effort was deliberate but also that the effort test failure reflects the examinee’s intentional and external motivation for the poor effort (i.e., to obtain an external incentive). In other words, whether or not the examinee is actually incentivized by the incentive is not considered. Whereas the first criterion seems reasonable, the latter may require an impermissibly large leap of faith based entirely on the mere presence of incentive. Deliberate failure on an effort test does not indicate or even imply the reason(s) for failure. As a striking example, even scoring significantly below chance levels on the Test of Memory Malingering (TOMM; Tombaugh, 1997) does not equate to failing expressly to obtain some material gain, even when the case involves financial compensation. This problem of determining motivation is not unique to SVT and PVT prevalence estimation; rather, it is inherent to all research on malingering. While the current definition requires that malingering be motivated by external incentives, it does not provide any advice on how to determine the causal role of motivation. Instead, it provides only token guidance regarding context and risk factors. In light of these factors, most base rate estimates of malingering are very likely to be substantial overestimates. Similarly, the fact that findings of probable malingering are often lumped in with those of definite malingering likely has inflated base rate estimates. A recent study by Ruff, Klopfer, and Blank (2016) suggests that “lumping” has significant implications with regard to classification rates. The authors disentangled levels of classificatory certainty in a group of consecutively referred personal injury cases. They examined rates of “definite” malingering independently from “probable” and “possible,” as defined by Slick et al. (1999), and applied hom*ogenous testing methods across all participants. When analyzed by this approach, the base rates of definite and probable malingering plummeted: 4.7% for probable malingering and 2.0% for definite malingering. The percentages stand in stark contrast to previous studies that

(1) grouped definite with probable malingering, and (2) used more heterogeneous tests and methods. Interestingly, most of the claimants in their study presented with TBI, with severity ranging from mild to severe. When the group of patients with mTBI (n = 89) was evaluated independently, the rates increased only slightly (6.7% for probable and 3.4% for definite malingering). Given that prior studies have reported much higher base rates of malingering in mTBI, these data warrant replication. Note that the base rate of “possible” malingering reported in the Ruff et al. (2016) study was 21.3%; when this number is added to the other gradations of malingering from that study, the overall base rate is more in line with previous estimates (28% according to Ruff et al. vs. 32–38% according to prior estimates). In short, it appears that previous studies have relied on overinclusive criteria; using more specific criteria reduces the base rate. The Ruff et al. data are fairly well-aligned with the most comprehensive review of base rates to date by Young (2015), who reported that the base rate of malingering in forensic contexts was likely to be 10–20%.

DETECTION STRATEGIES FORNEUROCOGNITIVEMALINGERING Numerous psychometrically derived detection strategies can be systematically applied to neurocognitive feigning detection. For example, the strategy of “severity indexing” uses information about dose–response relationships between injury severity and test performances to identify unlikely or amplified presentations. In short, performances that are indicative of a severe injury (according to normative standards), while simultaneously coupled with a mild injury, raise questions about the validity of those performances primarily because the scores appear to be marked amplifications of the actual degree of dysfunction. Amplified presentations appear to be especially well-suited for use in forced-choice testing (FCT) formats. FCT performance can be compared to expected performance levels based on clinical groups, and to below-chance levels of performance. Both can be compelling indicators of feigning but clearly have different strengths and weaknesses. For example, significantly below-chance performance is virtually synonymous with deliberate failure, but it lacks sensitivity, which means that few feigners will be detected this way. In contrast, FCT comparisons to clinically derived cutoff

3. Neuropsychological Models of Feigned Cognitive Deficits 49

scores tend to be more sensitive but lack specificity, which means that some examinees may be incorrectly categorized as feigning. Both freestanding and embedded FCTs employing the amplified presentation strategy have been incorporated into the Advanced Clinical Solutions (ACS) scoring package for the WAIS-IV and Wechsler Memory Scale–IV (WMS-IV) (Holdnack, Millis, Larrabee, & Iverson, 2013). The ACS includes five effort tests (one freestanding and four embedded), each of which relies primarily on the floor effect detection strategy, thereby falling under the umbrella of the amplified presentations strategy. The likelihood of failing one measure can be added to the likelihood of also failing others, thereby increasing the probability of a correct classification. A strength of the ACS program is that it presents its classificatory estimates in the context of a range of base rates. In other words, it shows which effort tests were failed, how often such a performance occurred in various clinical samples, and the probability of that particular number of tests being failed by chance. The importance of comparing performance to clinical groups before making a determination of feigning is evident in Table 3.2, which is taken from the ACS Clinical and Interpretive Manual. This methodology is not new to statistics, but it is new to neuropsychological assessment and represents a significant advance in psychometric detection of feigned cognitive impairment. However, users of the ACS should be aware of potential methodological issues involving the potential for inflated estimates associated with making multiple comparisons and with the degree to which the measures may be correlated. As we discuss later, each of these issues warrants further investigation. The Response Bias Scale (RBS) is an example of an amplified presentation strategy applied to the MMPI-2-RF. The RBS was developed by Gervais, Ben-Porath, Wygant, and Green (2007) as a measure of exaggerated cognitive complaints based on items that correlated with failure on the Word Memory Test (WMT; Green, 2003). The RBS is empirically derived and appears to be effective because its items capture overreporting of cognitive symptoms without being correlated to actual memory ability or memory complaints (Gervais, Ben-Porath, Wygant, & Green, 2008; Gervais, Ben-Porath, Wygant, & Sellbom, 2010). In a criterion-group design investigating intentional versus unintentional feigning, Peck et al. (2013) found that the RBS could differentiate individuals with nonepileptic epilepsy from genuine

and nongenuine (≥ 2 PVT failures) patients with TBI, especially when used in conjunction with the Symptom Validity Scale (SVS; formerly called the Fake Bad Scale [FBS]). False-positive rates ranged from good to excellent, but the sample size was small. Also, they assumed a base rate of 40% for malingering (which may be double the true base rate), thereby artificially increasing classification rates. Nonetheless, the study suggests that using the two scales together has potential, and that the RBS in particular represents a relatively rare opportunity to identify feigned symptoms of TBI via a psychiatric symptom inventory. With regard to the unlikely presentation strategy, Rogers, Robinson, and Gillard (2014) created a Symptom Combinations (SC) scale for the Structured Interview of Malingered Symptomatology (SIMS; Smith & Burger, 1997) based on item pairs that were uncorrelated or negatively correlated in patients but were frequently endorsed by feigners. The strategy resulted in an effect size of 2.01, and while it still requires further validation, it appears to be a promising means of expanding the utility of the SIMS. New technologies are also being considered. For example, Bigler (2015), while acknowledging the general utility of current effort testing methods, has proposed that neuroimaging be incorporated into the neuropsychological assessment of effort and test validity. Part of his contention stems from mounting evidence, including the Institute of Medicine’s (IOM) position paper, suggesting that data obtained from PVTs and SVTs are insufficient when determining the motivations for effort test failure (Freedman & Manly, 2015). Bigler (2015) advances a framework to test hypotheses about the neural substrates of effort and motivation. In short, he argues that because all PVTs and SVTs involve at least some degree of task engagement, and task engagement can be impaired by neurocognitive disorders, neuroimaging should be part of the formulation when determining where cutoff scores should be on PVTs and SVTs. He also notes that nonconscious factors likely play a role in motivation, and imaging could uncover these processes. Bigler’s arguments appear to be logical and to have intuitive appeal, but establishing empirical bases to his points would appear to be a huge undertaking involving hundreds of patients with TBI and other clinical groups with various neurocognitive and mental disorders. Until more research supports forensic applications, imaging should not be used as an indicator of malingering (see McBride, Crighton, Wygant, & Granacher, 2013).

50

I.  Concep t ua l Fr a me work TABLE 3.2.  Percentage of Cases in Groups of Interest Having Various Numbers of Scores belowClinical 5% Cutoff

Number of scores at 5% cutoff 1

2

3

4

5 —

Groups of interest No stimulus group

100

100

88

    .8

Simulators

 40

 24

14

10

2

Overall clinical sample

  9

  2

Traumatic brain injury

  9

  3

Temporal lobectomy

  8

Schizophrenia

 15

  2

Major depressive disorder

  4

0 —

Anxiety disorder

Intellectual disability—mild

 38

 13

 3

 3

Autistic disorder

 23

  5

Asperger’s disorder

  3

Reading disorder

Mathematics disorder

 14

ADHD

  6

  3

Nonclinical sample

  2

Education level ≤8 years

 15

  2

9–11 years

  9

  1

12 years

  2

13–15 years

  1

≥16 years

  1

  1

Race/ethnicity White African American

  5

  1

Hispanic

  6

  1

Asian

General Ability Index 69 or less

 19

  6

70–84

  7

  1

85–99

  2

100–114

  1

115 or higher

Note. This table pertains to a 5% cutoff score only. The ACS also provides percentage estimates based on 2, 10, 15, and 25% cutoff scores, yielding a range of confidence levels.

Advanced Clinical Solutions for WAIS-IV and WMS-IV. Copyright © 2009 NCS Pearson, Inc. Reproduced with permission. All rights reserved. “WAIS” and “WMS” are trademarks, in the United States and/or other countries, of Pearson Education, Inc. or its affiliates(s).

3. Neuropsychological Models of Feigned Cognitive Deficits 51

THE CURRENT STATE OFVALIDITY TESTING INNEUROPSYCHOLOGICALASSESSMENT As mentioned, a number of embedded and freestanding tests are sold by major publishers that, by design, evaluate feigned cognitive impairment. These cognitive feigning tests have been described in the literature in various compendia (e.g., Lezak, Howieson, Bigler, & Tranel, 2012). Additionally, Schutte and Axelrod (2013) reported on “empirically-derived embedded measures that are typically derived from commonly administered neuropsychological tests” (p.160). According to Schutte and Axelrod, advantages of “embedded measures” include efficiency (reducing time constraints by administering stand-alone malingering tests), reducing ease of coaching, assessment in multiple domains of ability, multiple opportunities for assessing malingering, and assessment at multiple time points. Schutte and Axelrod reported on no fewer than 50 unique embedded measures for identifying feigned cognitive impairment. Likewise, Martin, Schroeder, and Odland (2015; see their Tables 21 and 22) reported a wide range of standalone and embedded tests currently employed in neuropsychological assessments. Martin et al. (2015) surveyed 316 neuropsychologists regarding their practices for using stand-alone and embedded measures of validity test measures. The mean number of validity tests administered per assessment was about six, with up to 16 stand-alone and embedded measures included in any one forensic assessment. On a smaller scale, Schroeder, Martin, and Odland (2016) surveyed 24 neuropsychologists whom they identified as experts in the assessment of feigned cognitive impairment. These experts administered an average of eight stand-alone and embedded measures, using up to 12 such measures in forensic evaluations. The authors provided respondents with a number of possible explanations for positive scores (i.e., scores suggestive of feigning) to see how the alternatives were ranked in forensic and nonforensic evaluations. None of the alternative explanations included “test unreliability.” In other words, false positives were assumed to always be related to genuine cognitive dysfunction. In standard criterion groups analysis, however, false-positive classification errors refer to a failure of the test to classify correctly on the construct used to divide the criterion groups; that is, effort or feigning—not cognitive dysfunction. A false-positive classification in this context refers to the inherent unreliability of the classification method, and that should be the

first consideration of reasons for assumptions of “false positives” in the clinic. It is correct to say: “Individuals with bona fide pathology nevertheless inaccurately generate positive scores, not because they have genuine pathology, but because the test is inherently unreliable to some degree, which is reflected in the false-positive rate.”

Classification Errors Produced byTest DecisionRules For any useful classification test, there are characteristics of the test that result in differential rates of positive scores for people with a condition (for our purposes, faking symptoms) and for people without the condition (not faking symptoms). Because some tests have more than one potential score/basis for classification, we prefer to use the term trials rather than tests. Any decision about feigning based on an objective test score will be referred to as a trial. Not all feigning receives a positive score on a feigning test trial, but the rate at which a single trial generates positive scores for feigners is the true positive rate (TPR, 0 ≤ TPR ≤ 1). Some persons who are not faking (compliant responders) nevertheless generate positive tests scores on malingering test trials—these are falsepositive classification errors. The reasons for these errors of classification include the inherent unreliability of the classification process and errors in measurement. The rate at which the trial generates incorrect positive scores is called the “falsepositive rate” (FPR, 0 ≤ FPR ≤ 1). The inherent unreliability of the classification process can be explained by the fact that compliant responders often have attributes in common with feigners on many neuropsychological assessment procedures (for a list of alternative clinical explanations for misclassification, see Martin et al., 2015; Schroeder et al., 2016). It is important to recognize that TPR and FPR are not the test information that is essential for individuals who administer tests; clinicians must consider positive predictive power (PPP; the probability that a positive test score resulted from feigning). As noted by Frederick and Bowden (2009a), values of PPP are not always obvious, as PPP depends on the probability that someone is feigning before testing even begins (i.e., the local base rate of feigning). As an example, consider the Failure to Maintain Set (FMS) score as applied by Larrabee (2008), who assumed that the base rate (BR) of feigning is 40% in compensation-seeking persons alleging TBI. Larrabee reported that FMS

52

I.  Concep t ua l Fr a me work

> 1 has a TPR = .48 and FPR = .129. FPR = .129 might suggest to many that there exists only a 12.9% probability that a person alleging TBI and seeking compensation would be classified as feigning despite not actually feigning impairment. The risk is actually much higher as the BR of feigning falls below 50%, as more and more compliant individuals are at risk for a false-positive classification error. For example, consider Larabee’s BR estimate of .40 for 100 examinees—40 feigners. True positive classification of feigning is found by 40 * .48, which is to say 19 true feigners are classified by FMS > 1. False-positive classifications of feigning for the 60 individuals who are compliant is 60 * .129, which means that eight individuals will be falsely classified as feigning; 27 individuals are classified as feigning—the rate of false classifications for the 100 individuals is 8/27 or 30%, which is far in excess of the estimated FPR. If the BR is actually much lower, say 15%, then the rate of false classifications is (85 * .129) / ([85 * .129] + [15 * .48]) = 11/18 = 61%. Decisions based on classification scores inherently have error. Even if TPR and FPR are known precisely (which is never the case), decisions are essentially exercises in choosing which form of error one wishes to risk. If one decides that a score is consistent with feigning, then one chooses to risk some form of false-positive classification error: “I am willing to characterize this person as feigning at the risk of incorrectly mislabeling someone who is not faking as a faker.” If one decides that a test score is not consistent with feigning, then one chooses to risk some form of false-negative error: “I am willing to label this person as compliant at the risk of not identifying a faker as a faker.” Forensic practitioners should keep in mind that error-free decisions are not possible, and it is often impossible to recognize when one is making an error.

Use ofMultiple ValidityTests Common sense would argue that two positive scores on feigning tests predict feigning better than one positive score only. Given that there is usually more than one trial related to a decision in current examinations, and sometimes there are many trials, the overall rate of making a falsepositive error must be considered. Given the large numbers of tests and embedded measures that exist and apparently are administered in forensic assessment of cognitive abilities (Martin et al., 2015), it seems important to have a strategy for evaluating multiple positive test scores. Authors (Boone

& Lu, 2003; Larrabee, 2008; Ross, Putnam, Millis, Adams, & Krukowski, 2006) have discussed methods of combining test results to better evaluate the probability of malingering. MultipleComparisons

When multiple tests are evaluated in hypothesis testing, the associated error rate is referred to as the familywise error rate (FWER; e.g., see Keselman, Miller, & Holland, 2011). Keselman et al. noted that the FWER is computed as 1 – (1 – alpha)m, where “m” is the number of hypothesis tests and alpha is the predetermined acceptable error rate (p.421). The FWER inflates rapidly when alpha = .05 and when there are many hypothesis tests. For instance, for five hypothesis tests at alpha = .05, the overall false-positive error rate, FWER, is 1 – (1 – .05)5 = .23. It is common practice in experimentation to adjust the individual hypothesis error rates (i.e., to reduce alpha) in a series of decisions (e.g., post hoc hypothesis tests) to keep the overall error rate at or below 5%. Unlike alpha error rates chosen a priori in hypothesis testing, classification test FPRs often vary considerably. For a series of five classifications tests, FWER can be computed as FWER = 1 – [(1 – FPR1)*(1 – FPR2) *(1 – FPR3)*(1 – FPR4)*(1 – FPR5)] Larrabee (2008) was interested in demonstrating how information about TPR and FPR could be combined for multiple trials to improve the certainty about classification of malingering. He reported on four embedded measures of malingering from the Visual Form Discrimination Test (VFD; TPR = .48, FPR = .069), the finger-tapping test (FTT; TPR = .40, FPR = .065), the RDS (TPR = .50, FPR = .065), and FMS on the Wisconsin Card Sorting Test (FMS; TPR = .48, FPR = .129), plus the MMPI-2 FBS (TPR = .808, FPR = .138). Based on these FPR estimates, we can compute the FWER for combining these five classification trials: FWER = 1 – [(1 – .069) * (1 – .065) * (1 – .065) * (1 – .129) * (1 – .138)] = .389 At least 39% of non-fakers will get at least one positive score when completing this sequence of tests. However, when considering the meaning of just one positive score for these trials, Larrabee es-

3. Neuropsychological Models of Feigned Cognitive Deficits 53

timated (at BR = .40) that one positive score represented an 80–84% likelihood of faking (see Larrabee, 2008, Table 2). If 39% of non-fakers get at least one positive score for these five tests, it cannot be true that getting one score positive in five tries represents an 80–84% likelihood of faking. The error lies in the way the likelihood ratios were chained.

Chained LikelihoodRatios Calculating the probability of feigning only in terms of the number of positive classifications does not take into account the meaning of negative classifications. The probability of negative classifications must be considered in the context of the complete evaluation—if five feigning tests were used, the outcomes of all five tests must be evaluated. For example, if five feigning tests are administered and two are failed, it is easy to understand that this means something much different than if 1,000 tests are administered and two are failed. Larrabee (2008) proposed the use of likelihood ratios (LRs) to understand the information present in multiple validity tests. LRs are ratios that assess the potential utility of a classification test. When a test score is positive, LR+ = TPR/FPR. When a test score is negative, LR– = (1 – TPR) ÷ (1 – FPR). For validity tests, LR+ essentially means “increased likelihood of feigning” and is calculated only when a test score is positive. LR– essentially means “decreased likelihood of feigning,” and is calculated only when a test score is negative. “Chained likelihood ratios” refers to the process of considering how test information changes the initial likelihood of feigning represented by BR. BR is the pretest likelihood of feigning, and chaining LRs results in a posttest likelihood of feigning. The mathematics of chaining LRs essentially derives a probability value for a process, in which X tests are passed and Y are failed: 1. Start: Pretest probability of feigning (i.e., BR, converted to pretest odds of feigning) a. First test passed; pretest odds decrease; therefore there is a new value for odds of feigning b. Second test failed; odds of feigning from (a) therefore increase; new value for odds of feigning c, d, e, etc. Iterative changes to odds of feigning based on passing or failing remaining tests, until 2. End: Posttest odds of feigning (converted to post-test likelihood of feigning).

It is important to note that a likelihood expressed as a “probability of an event” is not the same as the “odds of an event,” even though the values are easily transformed from one to the other. For example, if the pretest likelihood of feigning (BR) = .4, then we compute the pretest odds of feigning, which is BR ÷ (1 – BR), or .4/.6 = .67. To “chain” likelihoods, we must convert BR to an odds value (pretest odds), then multiply the pretest odds by each test’s LR, depending on whether it was a pass or a fail. In the case in which two of five tests were failed, Pretest odds * LR1– * LR2– * LR3– * LR4+ * LR5+ =posttest odds → posttest probability of feigning The method of chaining likelihoods has existed for decades and is commonly cited in medical decision-making texts such as that of Straus, Richardson, Glasziou, and Haynes (2011). When Larrabee (2008) reported a range of probabilities for one positive score, he calculated the values this way, using the information for only the one failed test, Pretest odds * LRn+ = posttest odds → posttest probability of feigning For a more balanced estimate, the information from all the tests administered should be used (Straus et al., 2011). In Larabee’s example, test 1 is failed, but tests 2–5 are passed: Pretest odds * LR1+ * LR2– * LR3– * LR4– * LR5– = posttest odds → posttest probability of feigning Because it appeared to us that all five tests were used for all participants in Larrabee’s study, we applied his method of chaining LRs in order to recalculate Tables 1, 2, and 3 in Larrabee (2008, pp.671–672) using both positive and negative LRs; that is, like Larrabee, we first calculated the probability of a single positive trial, but we then recalculated that probability in light of having four negative trials follow the positive trial; we chained both positive and negative LRs. For Larrabee’s Table 2, we calculated the probability of two trials being positive, but then modified the probability by the necessary implication that three trials were negative. For his Table 3, we calculated the probability of three trials positive and two trials negative. In our own Table 3.3, we report the calculated values based on Larrabee’s assertions, then we provide estimates of the likelihood of faking

54

I.  Concep t ua l Fr a me work

TABLE 3.3.  Probability of MND Given Number of Positive Trials out of Five Trials

Larrabee (2008)

Recalculation

One positive trial

.796 to .837

.096 to .308

Two positive trials

.936 to .973

.569 to .865

Three positive trials

.989 to .995

.927 to .988

Note. The recalculated estimates include data from both negative and positive trials.

given n positive trials by use of simulation study. It is important to note that Table 3.3 applies only to a situation in which five trials are planned and presented. The recalculated probability values for three positive trials would be much lower if there were 10 or 20 total trials, for example. Ethically, clinicians are obligated to report the outcomes for all trials that they consider appropriate indicators of validity; that is, if they use RDS < 7 on one case, they must use it consistently on all cases when validity assessment is invoked. In other words, they cannot vary the cutoff score and sometimes use RDS < 8. It is certainly inappropriate for clinicians to choose validity measures after they have seen the outcomes. Berthelson, Mulchan, Odland, Miller, and Mittenberg (2013) investigated the implications of using multiple tests to evaluate feigning. They obtained data from 22 independent studies of nonmalingering individuals without pending compensation issues or litigation, who had completed a variety of validity measures. They found substantial intercorrelation of test scores within most of the samples, with a mean correlation of .31 across all studies. Based on these findings, they provided tables (based on Monte Carlo simulations) that estimate the overall false-positive rates when using as many as 20 validity tests in an assessment. For example, with an assumption that every validity test has an FPR = .10, they show that the actual FPR for 20 tests > .10, until at least 6 tests are failed. If the FRP = .15 for each test, they show the actual FPR for 20 tests > .10, until at least 8 tests are failed. A more detailed Monte Carlo analysis can be found in Odland, Lammy, Martin, Grote, and Mittenberg (2015). Their results refuted the findings of Larrabee (2008), who claimed that even one failure is associated with an 80% chance of faking, and their results are inconsistent with the Slick and Sherman (2013) criteria that indicate feigned cognitive impairment is “probable” when two or more tests are failed.

Larrabee (2014) and Davis and Millis (2014) contended that validity tests are not nearly as correlated as the rates reported by Berthelson et al. (2013). Davis and Millis (2014) used up to nine validity tests for 87 outpatient physiatry referrals, which constituted a neurological no-incentive group, and found a nonsignificant correlation of .13 among the tests. Six of the tests had FPRs much lower than .10, two had FPRs = .10, and one test had an FPR = .15. Among the participants, 18 failed at least one test (20.7%), seven failed at least two tests (8.0%), and four failed three or more tests (4.6%). Davis and Millis reported that using two or more positive scores as a basis for a classification of MND results in a 12.6% false classification rate. Berthelson et al. (2013, Table IV) had suggested that when nine tests with FPRs = .10 were used, 22.8% would fail two or more tests. The difference between 22.8 and 12.6% might be related to the more conservative test scores used by Davis and Millis (2014), but Davis and Millis argued that Bertheleson et al. (2013) were simply wrong to assume that correlations among tests would make failure rates more likely. Larrabee (2014) also argued that the tables generated by Berthelson et al. (2013) produced inflated FPRs for multiple tests because of undue consideration of correlation among validity measures. However, as can be seen in several tables generated by Larrabee (2014), the differences in data from Larrabee and estimates from Berthelson et al. (2013) were not significantly different, which Larrabee attributed to a problem in statistical power. Pella, Hill, Shelton, Elliott, and Gouvier (2012) also showed quite high rates of false positives when using multiple validity indicators. Within a group of 478 students who had no external incentive to malinger cognitive impairment, they found that when administered eight validity indicators, 21.1% failed two or more, 4.6% failed three or more, 1.9% failed four or more, and 0.4% failed five indicators. These findings comport well with the values reported by Berthelson et al. (2013), and suggest that if the Berthleson et al. figures were wrong, they were not wrong by much. Regardless of whether the tests are highly correlated or not, Pella et al. (2012) replicated Berthelson et al.’s (2013) finding that the use of multiple validity indicators inherently leads to high rates of false-positive classification errors. Berthelson et al. (2013) reported a mean correlation coefficient of .31 to characterize the relationships of validity indicators to each other for the studies they sampled, and Odland et al. (2015) used the value of .31 to compute false-positive

3. Neuropsychological Models of Feigned Cognitive Deficits 55

rates for using multiple validity tests. But the Berthleson et al. (2013) meta-analysis included studies whose tests were not independent. For example, they included both the Rey 15-Item Test and the Rey 15-Item Recognition Test—the score for the latter test includes the score for the former test. The reported correlation for this study was .63. Additionally, Berthleson et al. included many studies that included both Digit Span scaled score and RDS. Both measures are from the same test. These studies included correlations of .83 and .92. In contrast, Davis and Millis (2014) reported r = .13 among the validity tests administered in their neurological no-incentive group, an apparently pure group of nonfeigners. When validity tests are independent in content, there is no reason to expect that trials will be significantly correlated within pure groups of nonfeigners or pure groups of feigners. Validity test scores that are obviously dependent on other measures (as in the earlier examples) should not be used to make classifications of validity. It seems reasonable to conclude that the use of multiple feigning tests sharply increases the possibility of false-positive classification of individuals who are compliant—not just when the tests are correlated—but because the false-positive error rate accumulates with the administration of each additional validity measure.

What Do These TestsMeasure? Neuropsychologists and others who evaluate feigned cognitive impairment need to consider what the tests measure. Bigler (2011, 2014) has provided thoughtful analyses of what effort tests actually measure, questioning the indiscriminate use of the term effort. Frederick and Bowden (2009b) considered a number of ways to evaluate the constructs measured by the Validity Indicator Profile (VIP), finding support for more narrowly honed “effort” constructs, “intention to respond correctly” and “magnitude of effort.” Careful construction of criterion groups is not commonly accomplished. Both Frederick (2000) and Frederick and Bowden (2009a) explored problems in validity test research relating especially to problems with criterion group contamination (not knowing which members of criterion groups are actually compliant or faking). Within the neuropsychology literature, Greve and Bianchini (2004) have emphasized the importance of establishing specificity for validity indicators by carefully identifying groups of individuals who have no identifiable incentives to perform poorly but who other-

wise match potentially malingering samples with regard to the types of injuries and demographic characteristics. Some researchers have promoted methods that do not even depend on knowing a priori which participants belong to which groups. Meehl (1995) proposed numerous mathematical strategies (taxometrics) that do not depend on knowing which individual members of samples are actually feigning or not feigning. For example, Strong, Glassmire, Frederick, and Greene (2006) investigated the characteristics of the MMPI-2 F(p) validity indicator using Meehl’s taxometric mean above minus below a cutoff (MAMBAC) and maximum eigenvalue (MAXEIG) procedures. Walters et al. (2008) also evaluated the Structured Interview of Reported Symptoms (SIRS) using Meehl’s maximum covariance (MAXCOV) and MAMBAC methods. Similarly, Mossman, Wygant, and Gervais (2012) investigated the utility of latent class modeling (LCM) to identify test diagnostic statistics. They applied each of two LCM approaches. In the first, the agnostic approach, the only basis for analysis is the test scores—no knowledge of participant status is required. In the second, the partial truth approach, confident classifications of a subset of participants (in their study, about 10% of extreme cases) promoted the modeling of test behavior.

Common versusDistinguishing Characteristics andMarkers ofGenuineResponding Characteristics common to cognitive feigning include poor effort, exaggeration, litigation or presence of potential compensation, criminal prosecution, and adversarial proceedings. Such characteristics can be helpful in the early stages of determinations of feigning. However, they are merely correlates and do not have any clinical value as distinguishing characteristics (see Rogers, Chapter 2, this volume). This problem was illustrated earlier in the example involving antisocial personality and malingering: Antisocial personality may be fairly common among malingerers, but it is not a distinguishing characteristic of malingering. Establishing just what constitutes a distinguishing characteristic can be a challenge. For example, postconcussion syndrome (PCS) is reliably characterized by several cognitive, somatic, psychiatric symptoms—headache, irritability, sleep disturbance, attention problems, and other symptoms are very common in PCS. Research has shown,

56

I.  Concep t ua l Fr a me work

however, that these symptoms are also common in other conditions, and in healthy individuals as well (Maruta et al., 2016). Before a distinguishing characteristic can have any meaningful degree of utility, it must be shown to differentiate groups and individuals through studies employing appropriate criterion-group research designs. The probability of correctly recognizing genuine (and feigned) responses may be increased through the use of indeterminate groups in psychometric research. When relying on cutoff scores to differentiate individuals, occasional incorrect classifications are inevitable, especially those at or very near the cutoff score. Instead of forcing tooclose-to-call scores into a classification group, they are removed and are not included for calculation of utility estimates. While this procedure does mean that all participants cannot be classified, it also means that far fewer will be misclassified. There is certainly room for discussion about this research methodology with strengths and weakness to be debated. Rogers and Bender (2013) reviewed the implications of using this methodology for clinical decision making, and Rogers, Gillard, Wooley, and Ross (2012) applied it to the Personality Assessment Inventory (PAI) with regard to feigned posttraumatic stress disorder (PTSD). The overall classification rates improved only modestly when removing too-close-to-call cases. Nonetheless, it avoided impermissibly high error rates within the indeterminate group. For example, removing the range of Negative Impression Management (NIM) scores that were ±5T from the cutoff score (70T) removed 38.5% of the false-positive errors. There are other implications as well, but with regard to markers of genuine responding, the procedure resulted in reduced error. Practitioners should be knowledgeable about markers of feigned impairment, of which there are many. But indications that an examinee has responded honestly should be considered carefully as well. Rarely is an individual’s presentation accurately categorized as entirely invalid (or entirely valid), and the absence of invalid scores does not mean that the examinee’s presentation is therefore valid. Both effort and malingering are likely dimensional rather than dichotomous (Bigler, 2014; Walters et al., 2008). Forcing dimensional/ continuous variables into a dichotomy forces out important information. Yet this is what is typically demanded in forensic settings. Similarly, indications of invalid performance or exaggerated symptomatology are not synonymous with feign-

ing or malingering. This important distinction is even more important in forensic contexts in which examinees may have little reason to be forthcoming or honest (Rogers & Bender, 2013) but may not feel compelled to malinger. Genuine neuropsychological presentations are characterized by a general consistency both between the syndrome in question and the symptoms presented, and among test performances. But these variables depend on the phase/acuity of the disorder. A mild TBI, for example, produces generally mild symptoms. However, this may not be true in the acute phase of recovery, when symptoms may be more pronounced. Scores within the same domains that hang together are more likely to be valid reflections of performance than those that do not. This is not to say that an odd departure in a score means the performance is invalid. Rather, it simply may require follow-up. While consistent performance can be a sign of validity, inconsistent performance is not necessarily a sign of invalidity. Some degree of inconsistency in test performance is expected in healthy adults (Binder, Iverson, & Brooks, 2009). Thus, the normal vagaries of consistency make it difficult to establish in individual cases. It is incumbent on the forensic practitioner to know the general number of low scores expected given the examinee’s age and IQ that might otherwise be misattributed to cognitive dysfunction. A troubling situation arises when what was genuine effort devolves into a more cynical presentation involving effort test failure and exaggeration simply due to the pressures of protracted litigation (Bender & Matusewicz, 2013; Hall & Hall, 2012; Silver, 2015). Many questions emerge in such cases: What responsibility for malingering (if any) lies with the process of litigation and serial neuropsychological assessments, during which the examinee’s complaints are repeatedly contested? And would this suggest that the examinee is less accountable? While by no means exculpatory, Rogers’s (1990) adaptive explanatory model for malingering seems particularly germane here. Finally, if repeated neuropsychological assessments represent a risk factor for malingering (Vanderploeg, Belanger, & Kaufmann, 2014), does this imply a professional obligation to reduce the risk or otherwise prevent malingering from happening in the first place? The interested reader should see Horner, Turner, VanKirk, and Denning (2017) for an interesting, albeit preliminary, study of this issue.

3.  Neuropsychological Models of Feigned Cognitive Deficits 57

SUMMARY

TABLE 3.4.  Neuropsychological Decision Model for Feigned Cognitive Deficits

Malingering detection has been a focus of research for decades. Test development has burgeoned, and with it, our ability to detect malingering has improved. However, the growth in our understanding of malingering as a construct has been far less satisfying, and most malingering research has not pursued strategically based approaches. Nonetheless, neuropsychological models have recently emerged, and criteria specific to MND have been proposed. On the one hand, both the recent emphasis on BRs and the publication of the revised Slick criteria represent progress; on the other hand, it is humbling to see how much had been overlooked (or simply unknown) in the past. If we are to continue what has been substantial progress, existing diagnostic criteria for malingering must be validated, research must include criterion-group designs and multiple comorbidities, and both diagnostic probabilities and BRs must be accurately calculated. Moreover, with no means to measure motivation directly, the DSM (American Psychiatric Association, 2013) and Slick (Slick & Sherman, 2013) criteria are left conflating incentive with motivation, and practitioners are forced to rely on indirect indications of motivation, such as incentives. Important questions about incentives remain: Is the incentive a source of motivation for the examinee in the same way the examiner assumes it to be (Rogers & Bender, 2013)? Should the examiner attempt to establish that the incentive is indeed the source of motivation to malinger? What explains the (more common than not) absence of malingering in the presence of external incentives to malinger? In the meantime, based on the best available evidence, we recommend the following steps for forensic practitioners using symptom and performance validity tests for determinations of malingering (see Table 3.4).

1.  Determine whether the reported symptoms and/or cognitive profile(s) make neurological sense a.  Determine whether test performance fits known cognitive profiles (e.g., better performance on free recall than on recognition is unusual in genuine amnesia) b.  Is the amount of test scatter within base rate expectations for age and IQ? See Binder, Iverson, and Brooks (2009), Iverson, Brooks, and Holdnack (2008), and Schretlen, Munro, Anthony, and Pearlson (2003) for guidelines.

REFERENCES American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disorders (3rd ed.). Washington, DC: Author. American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: Author. Bender, S. D., & Matusewicz, M. (2013). PCS, iatrogenic symptoms, and malingering following concussion. Psychological Injury and Law, 6, 113–121. Bender, S. D., & Rogers, R. (2004). Detection of neu-

2.  Determine whether the degree of cognitive impairment reported is beyond that expected given the severity of the injury (Amplified and Unlikely Presentations) a.  Look for compelling inconsistencies (both qualitative and quantitative). b.  Is the degree of reported impairment inconsistent with the degree of functional disability? c.  Employ strategic detection tests and methods (e.g., amplified and unlikely presentation strategies) for both freestanding and embedded measures. d.  Use of posterior probabilities incorporating multiple tests, clinical comparison groups, and estimates of base rates (e.g., see ACS tables for the WAIS-IV and WMS-IV). e.  Use of multiple detection tests in aggregate but must know the number of negative trials on PVTs as well as positive. f.  Acknowledge existence of “indeterminate groups” in research. 3.  Attempt to determine whether there is evidence of motivation to perform poorly, with intent to secure material gain. a.  Evaluate the examinee’s perceptions of the incentive and whether it is incentivizing to the examinee. 4.  Remain aware of construct drift (i.e., broadening the conceptualization of malingering to embrace any manifestation of inadequate motivation) that results in imprecise and likely misleading results.

rocognitive feigning: Development of a multi-strategy assessment. Archives of Clinical Neuropsychology, 19(1), 49–60. Berthelson, L., Mulchan, S. S., Odland, A. P., Miller, L. J., & Mittenberg, W. (2013). False positive diagnosis of malingering due to the use of multiple effort tests. Brain Injury, 27(7–8), 909–916. Bianchini, K. J., Greve, K. W., & Glynn, G. (2005). On the diagnosis of malingered pain-related disability:

58

I.  Concep t ua l Fr a me work

Lessons from cognitive malingering research. The Spine Journal, 5(4), 404–417. Bigler, E. D. (2011). Effort–what is it, how should it be measured? Journal of the International Neuropsychological Society, 17(4), 751–752. Bigler, E. D. (2014). Effort, symptom validity testing, performance validity testing and traumatic brain injury. Brain Injury, 28(13–14), 1623–1638. Bigler, E. D. (2015). Neuroimaging as a biomarker in symptom validity and performance validity testing. Brain Imaging and Behavior, 9(3), 421–444. Binder, L. M., Iverson, G. L., & Brooks, B. L. (2009). To err is human: “Abnormal” neuropsychological scores and variability are common in healthy adults. Archives of Clinical Neuropsychology, 24(1), 31–46. Boone, K. B. (2007). A reconsideration of the Slick et al. (1999) criteria for malingered neurocognitive dysfunction. In K. B. Boone’s (Ed.), Assessment of feigned cognitive impairment (pp.29–49). New York: Guilford Press. Boone, K. B., & Lu, P. (2003). Noncredible cognitive performance in the context of severe brain injury. Clinical Neuropsychologist, 17(2), 244–254. Bush, S. S., Ruff, R. M., Tröster, A. I., Barth, J. T., Koffler, S. P., Pliskin, N. H., et al. (2005). Symptom validity assessment: Practice issues and medical necessity NAN policy & planning committee. Archives of Clinical Neuropsychology, 20(4), 419–426. Davis, J. J., & Millis, S. R. (2014). Examination of performance validity test failure in relation to number of tests administered. Clinical Neuropsychologist, 28(2), 199–214. Delis, D. C., & Wetter, S. R. (2007). Cogniform disorder and cogniform condition: Proposed diagnoses for excessive cognitive symptoms. Archives of Clinical Neuropsychology, 22(5), 589–604. Drob, S. L., Meehan, K. B., & Waxman, S. E. (2009). Clinical and conceptual problems in the attribution of malingering in forensic evaluations. Journal of the American Academy of Psychiatry and the Law, 37(1), 98–106. Elwood, R. W. (1993). Psychological tests and clinical discriminations: Beginning to address the base rate problem. Clinical Psychology Review, 13(5), 409–419. Frederick, R. I. (2000). Mixed group validation: A method to address the limitations of criterion group validation in research on malingering detection. Behavioral Sciences and the Law, 18(6), 693–718. Frederick, R. I., & Bowden, S. C. (2009a). Evaluating constructs represented by symptom validity tests in forensic neuropsychological assessment of traumatic brain injury. Journal of Head Trauma Rehabilitation, 24, 105–122. Frederick, R. I., & Bowden, S. C. (2009b). The test validation summary. Assessment, 16(3), 215–236. Frederick, R. I., & Speed, F. M. (2007). On the interpretation of below-chance responding in forced-choice tests. Assessment, 14(1), 3–11. Freedman, D., & Manly, J. (2015). Use of normative data

and measures of performance validity and symptom validity in assessment of cognitive function. Available from www.iom.edu/~/media/files/perspectivesfiles/2015/freedmanmanlycommissioned%20paper.pdf. Gervais, R. O., Ben-Porath, Y. S., Wygant, D. B., & Green, P. (2007). Development and validation of a Response Bias Scale (RBS) for the MMPI-2. Assessment, 14(2), 196–208. Gervais, R. O., Ben-Porath, Y. S., Wygant, D. B., & Green, P. (2008). Differential sensitivity of the response bias scale (RBS) and MMPI-2 validity scales to memory complaints. Clinical Neuropsychologist, 22(6), 1061–1079. Gervais, R. O., Ben-Porath, Y. S., Wygant, D. B., & Sellbom, M. (2010). Incremental validity of the MMPI2-RF over-reporting scales and RBS in assessing the veracity of memory complaints. Archives of Clinical Neuropsychology, 25, 274–284. Green, P. (2003). Green’s Word Memory Test for Windows: User’s manual. Edmonton, Alberta, Canada: Green’s. Greiffenstein, M. F., Baker, W. J., & Gola, T. (1994). Validation of malingered amnesia measures with a large clinical sample. Psychological Assessment, 6(3), 218–224. Greve, K. W., & Bianchini, K. J. (2004). Setting empirical cut-offs on psychometric indicators of negative response bias: A methodological commentary with recommendations. Archives of Clinical Neuropsychology, 19(4), 533–541. Hall, R. C., & Hall, R. C. (2012). Compensation neurosis: A too quickly forgotten concept? Journal of the American Academy of Psychiatry and the Law, 40(3), 390–398. Heaton, R. K., Smith, H. H., Lehman, R. A., & Vogt, A. T. (1978). Prospects for faking believable deficits on neuropsychological testing. Journal of Consulting and Clinical Psychology, 46(5), 892–900. Heilbronner, R. L., Sweet, J. J., Morgan, J. E., Larrabee, G. J., Millis, S. R., & Conference Participants. (2009). American Academy of Clinical Neuropsychology consensus conference statement on the neuropsychological assessment of effort, response bias, and malingering. Clinical Neuropsychologist, 23(7), 1093–1129. Holdnack, J. A., Millis, S. R., Larrabee, G., & Iverson, G. L. (2013). Assessing performance validity with the ACS. In J, Holdnack, L. Drozdick, L. Weiss, & G. Iverson (Eds.), WAIS-IV, WMS-IV, and ACS (pp.331–365). Bloomington, MN: Pearson. Horner, M. D., Turner, T. H., VanKirk, K. K., & Denning, J. H. (2017). An intervention to decrease the occurrence of invalid data on neuropsychological evaluation. Archives of Clinical Neuropsychology, 32(2), 228–237. Iverson, G. L., Brooks, B. L., & Holdnack, J. A. (2008). Misdiagnosis of cognitive impairment in forensic neuropsychology. In R. L. Heilbronner (Ed.), Neuropsychology in the courtroom: Expert analysis of reports

3.  Neuropsychological Models of Feigned Cognitive Deficits 59

and testimony (pp. 243–266). New York: Guilford Press. Keselman, H. J., Miller, C. W., & Holland, B. (2011). Many tests of significance: New methods for controlling type I errors. Psychological Methods, 16, 420–431. Larrabee, G. J. (2003). Detection of malingering using atypical performance patterns on standard neuropsychological tests. Clinical Neuropsychologist, 17(3), 410–425. Larrabee, G. J. (2008). Aggregation across multiple indicators improves the detection of malingering: Relationship to likelihood ratios. Clinical Neuropsychologist, 22(4), 666–679. Larrabee, G. J. (2012). Performance validity and symptom validity in neuropsychological assessment. Journal of the International Neuropsychological Society, 18(4), 625–630. Larrabee, G. J. (2014). Test validity and performance validity: Considerations in providing a framework for development of an ability-focused neuropsychological test battery. Archives of Clinical Neuropsychology, 29(7), 695–714. Larrabee, G. J., Greiffenstein, M. F., Greve, K. W., & Bianchini, K. J. (2007). Refining diagnostic criteria for malingering. In G. J. Larrabee (Ed.), Assessment of malingered neuropsychological deficits (pp.334–371). New York: Oxford University Press. Lezak, M. D., Howieson, D. B., Bigler, E. D., & Tranel, D. (2012). Neuropsychological assessment (5th ed.). New York: Oxford University Press. Martin, P. K., Schroeder, R. W., & Odland, A. P. (2015). Neuropsychologists’ validity testing beliefs and practices: A survey of North American professionals. Clinical Neuropsychologist, 29(6), 741–776. Maruta, J., Spielman, L. A., Yarusi, B. B., Wang, Y., Silver, J. M., & Ghajar, J. (2016). Chronic post-concussion neurocognitive deficits: II. Relationship with persistent symptoms. Frontiers in Human Neuroscience, 10, 45. McBride, W. F., Crighton, A. H., Wygant, D. B., & Granacher, R. P. (2013). It’s not all in your head (or at least your brain): Association of traumatic brain lesion presence and location with performance on measures of response bias in forensic evaluation. Behavioral Sciences and the Law, 31(6), 779–788. Meehl, P. E. (1995). Bootstraps taxometrics: Solving the classification problem in psychopathology. American Psychologist, 50(4), 266–275. Millis, S. R. (2008). Assessment of incomplete effort and malingering in the neuropsychological examination. In J. E. Morgan & J. H. Ricker (Eds.), Textbook of clinical neuropsychology (pp.891–904). New York: Taylor & Francis. Mittenberg, W., Patton, C., Canyock, E. M., & Condit, D. C. (2002). Base rates of malingering and symptom exaggeration. Journal of Clinical and Experimental Neuropsychology, 24(8), 1094–1102. Mossman, D., Wygant, D. B., & Gervais, R. O. (2012). Estimating the accuracy of neurocognitive effort

measures in the absence of a “gold standard.” Psychological Assessment, 24(4), 815–822. Niesten, I. J. M., Nentjes, L., Merckelbach, H., & Bern­ stein, D. P. (2015). Antisocial features and “faking bad”: A critical note. International Journal of Law and Psychiatry, 41, 34–42. Odland, A. P., Lammy, A. B., Martin, P. K., Grote, C. L., & Mittenberg, W. (2015). Advanced administration and interpretation of multiple validity tests. Psychological Injury and the Law, 8(1), 46–63. Pankratz, L. (1979). Symptom validity testing and symptom retraining: Procedures for the assessment and treatment of functional sensory deficits. JournalofConsulting and Clinical Psychology, 47(2), 409–410. Peck, C. P., Schroeder, R. W., Heinrichs, R. J., VonDran, E. J., Brockman, C. J., Webster, B. K., et al. (2013). Differences in MMPI-2 FBS and RBS scores in brain injury, probable malingering, and conversion disorder groups: A preliminary study. Clinical Neuropsychologist, 27(4), 693–707. Pella, R. D., Hill, B. D., Shelton, J. T., Elliott, E., & Gouvier, W. D. (2012). Evaluation of embedded malingering indices in a non-litigating clinical sample using control, clinical, and derived groups. Archives of Clinical Neuropsychology, 27(1), 45–57. Rogers, R. (1990). Development of a new classificatory model of malingering. Journal of the American Academy of Psychiatry and the Law, 18(3), 323–333. Rogers, R., & Bender, S. D. (2013). Evaluation of malingering and deception. In A. M. Goldstein (Ed.), Handbook of psychology: Vol. 11. Forensic psychology (2nd ed.). Hoboken, NJ: Wiley. Rogers, R., Bender, S. D., & Johnson, S. F. (2011). A critical analysis of the MND criteria for feigned cognitive impairment: Implications for forensic practice and research. Psychological Injury and Law, 4(2), 147–156. Rogers, R., Gillard, N. D., Berry, D. T. R., & Granacher, R. P., Jr. (2011). Effectiveness of the MMPI-2-RF validity scales for feigned mental disorders and cognitive impairment: A known-groups study. Journal of Psychopathology and Behavioral Assessment, 33(3), 355–367. Rogers, R., Gillard, N. D., Wooley, C. N., & Ross, C. A. (2012). The detection of feigned disabilities: The effectiveness of the Personality Assessment Inventory in a traumatized inpatient sample. Assessment, 19(1), 77–88. Rogers, R., & Reinhardt, V. (1998). Conceptualization and assessment of secondary gain. In G. P. Koocher, J. C. Norcross, & S. S. Hill (Eds.), Psychologist’s desk reference (pp.57–62). New York: Oxford University Press. Rogers, R., Robinson, E. V., & Gillard, N. D. (2014). The SIMS screen for feigned mental disorders: The development of detection-based scales. Behavioral Sciences and the Law, 32(4), 455–466. Ross, S. R., Putnam, S. H., Millis, S. R., Adams, K. M., & Krukowski, R. A. (2006). Detecting insufficient

60

I.  Concep t ua l Fr a me work

effort using the seashore rhythm and speech-sounds perception tests in head injury. Clinical Neuropsychologist, 20, 798–815. Ruff, R. M., Klopfer, J., & Blank, J. (2016). hom*ogenous base rates for malingering in neuropsychological examination of litigants. Psychological Injury and Law, 9(3), 143–153. Schretlen, D. J., Munro, C. A., Anthony, J. C., & Pearlson, G. D. (2003). Examining the range of normal intraindividual variability in neuropsychological test performance. Journal of the International Neuropsychological Society, 9(6), 864–870. Schroeder, R. W., Martin, P. K., & Odland, A. P. (2016). Expert beliefs and practices regarding neuropsychological validity testing. Clinical Neuropsychologist, 30(4), 515–535. Schutte, C., & Axelrod, B. N. (2013). Use of embedded cognitive symptom validity measures in mild traumatic brain injury cases. In D. A. Carone & S. S. Bush (Eds.), Mild traumatic brain injury: Symptom validity assessment and malingering (pp.151–181). New York: Springer. Schutte, C., Axelrod, B. N., & Montoya, E. (2015). Making sure neuropsychological data are meaningful: Use of performance validity testing in medicolegal and clinical contexts. Psychological Injury and Law, 8(2), 100–105. Sharland, M. J., & Gfeller, J. D. (2007). A survey of neuropsychologists’ beliefs and practices with respect to the assessment of effort. Archives of Clinical Neuropsychology, 22(2), 213–223. Silver, J. M. (2015). Invalid symptom reporting and performance: What are we missing? NeuroRehabilitation, 36(4), 463–469. Slick, D. J., & Sherman, E. M. S. (2013). Differential diagnosis of malingering. In D. A. Carone & S. S. Bush (Eds.), Mild traumatic brain injury: System validity assessment and malingering (pp.57–72). New York: Springer. Slick, D. J., Sherman, E. M. S., & Iverson, G. L. (1999). Diagnostic criteria for malingered neurocognitive dysfunction: Proposed standards for clinical practice and research. Clinical Neuropsychologist, 13(4), 545–561. Slick, D. J., Tan, J. E., Strauss, E. H., & Hultsch, D. F. (2004). Detecting malingering: A survey of experts’ practices. Archives of Clinical Neuropsychology, 19(4), 465–473.

Smith, G. P., & Burger, G. K. (1997). Detection of malingering: Validation of the Structured Inventory of Malingered Symptomatology (SIMS). Journal of the American Academy of Psychiatry and the Law, 25(2), 183–189. Straus, S. E., Richardson, W., Glasziou, P., & Haynes, R. B. (2011). Evidence-based medicine: How to practice and teach it (4th ed.). New York: Elsevier. Strong, D. R., Glassmire, D. M., Frederick, R. I., & Greene, R. L. (2006). Evaluating the latent structure of the MMPI-2 F(p) scale in a forensic sample: A taxometric analysis. Psychological Assessment, 18(3), 250–261. Sumanti, M., Boone, K. B., Savodnik, I., & Gorsuch, R. (2006). Noncredible psychiatric and cognitive symptoms in a workers’ compensation “stress” claim sample. Clinical Neuropsychologist, 20(4), 754–765. Tombaugh, T. N. (1997). The Test of Memory Malingering (TOMM): Normative data from cognitively intact and cognitively impaired individuals. Psychological Assessment, 9(3), 260–268. Vanderploeg, R. D., Belanger, H. G., & Kaufmann, P. M. (2014). Nocebo effects and mild traumatic brain injury: Legal implications. Psychological Injury and Law, 7(3), 245–254. Walters, G. D., Rogers, R., Berry, D. T., Miller, H. A., Duncan, S. A., McCusker, P. J., et al. (2008). Malingering as a categorical or dimensional construct: The latent structure of feigned psychopathology as measured by the SIRS and MMPI-2. Psychological Assessment, 20(3), 238–247. Whiteside, D. M., Gaasedelen, O. J., Hahn-Ketter, A. E., Luu, H., Miller, M. L., Persinger, V., et al. (2015). Derivation of a cross-domain embedded performance validity measure in traumatic brain injury. Clinical Neuropsychologist, 29(6), 788–803. Young, G. (2014). Resource material for ethical psychological assessment of symptom and performance validity, including malingering. Psychological Injury and Law, 7(3), 206–235. Young, G. (2015). Malingering in forensic disability-related assessments: Prevalence 15 ± 15%. Psychological Injury and Law, 8(3), 188–199. Zenisek, R., Millis, S. R., Banks, S. J., & Miller, J. B. (2016). Prevalence of below-criterion Reliable Digit Span scores in a clinical sample of older adults. Archives of Clinical Neuropsychology, 31(5), 426–433.

CHAPTER 4

Beyond Borders Cultural andTransnational Perspectives ofFeigning andOther ResponseStyles AmorA.Correa,PhD

Psychological assessments are highly dependent on the honesty and openness of examinees. Historically, mental health practitioners developed professional relationships in therapy that built on trust and confidentiality, enabling many clients to openly express their symptoms and concerns. In the United States, the traditional professional–client relationship is sometimes overshadowed by a complex mental health system and opportunities for secondary gain for the patient (Rogers, 2008). Possible motives for falsifying symptoms include financial compensation, exemption from duty, or leniency from the criminal justice system; therefore, malingering is encountered in a variety of clinical and forensic situations (Reid, 2000). The increased complexity of mental health service delivery in other countries such as Spain may exert a similar effect on the demand for adequate feigning measures (Salvador-Carulla, Garrido, McDaid & Haro, 2006). Additionally, a “disability paradox” has been noted in the United Kingdom, where the number of disability claims in the country has risen despite evidence of an overall improvement in citizens’ health (Merten et al., 2013). Psychologists are now being asked to make important determinations of feigning and other response styles in an increasingly international context. The growing international spotlight on feigning presents an important quandary for practicing clinicians. In the realm of response styles, valida 61

tion research confirming the adequacy of current assessment measures is still catching up to the need for their use with increasingly culturally diverse clientele. As Merten et al. (2013) highlight, research on malingering was largely taboo in Germany and the United Kingdom until the early portion of the 21st century. Similarly, Spanish authors published their first book on the detection of feigning in 2012, and the majority of European Spanish feigning tests have only been published and made available over the course of the past 5 years (González Ordi, Santamaria Fernández, & Capilla Ramírez, 2012; Merten et al., 2013). This chapter will serve as a guide for practitioners conducting evaluations with culturally and linguistically diverse clients. The first section discusses major factors to consider in the field of multicultural assessment and test translation. The next section is an informative overview of currently available feigning measures. Finally, the chapter concludes with guidance on making culturally competent clinical interpretations of assessment data.

AN OVERVIEW ONPSYCHOLOGICAL ASSESSMENT ANDTESTTRANSLATIONS In the United States, ethical guidelines from the American Psychological Association (2002) require that psychologists working with ethnically,

62

I.  Concep t ua l Fr a me work

linguistically, and culturally diverse populations recognize these characteristics as important factors affecting a person’s experiences, attitudes, and psychological presentation (Bersoff, 2004). Mental health professionals are generally aware that standardized assessment measures in the United States were developed with normative groups that comprised individuals proficient in English and sharing in mainstream American culture. Psychological assessment with culturally diverse and international populations must be conducted with the knowledge that most standard assessment instruments were not constructed or validated with these populations in mind. Additional considerations are needed to ensure test validity when a psychological measure is translated into another language. The Test Translation and Adaptation Guidelines developed by the International Test Commission (ITC) in 1992, and updated in 2005, called for test developers and publishers to apply appropriate research methods and statistical techniques to establish the validity of a test for each population the adapted version is intended to assess. First, empirical research must be conducted. The research results should be used to improve the accuracy of the translation/ adaptation process and to identify problems in the adaptation that may render the new version inadequate for use with the intended populations. Additionally, test developers should strive to establish the equivalence of the different language versions of the test, to make them as parallel to the original as possible. Last, the validity of the translated version must be determined separately from that of the original measure. It should not be assumed that a translated version has acceptable validity simply because that of the original English language version is adequate (Allalouf, 2003; Anastasi, 1988). The ITC standards are consistent with the Standards for Educational and Psychological Testing, coauthored by the American Psychological Association (American Educational Research Association, American Psychological Association, & National Council on Measurement in Education, 1999, 2014), and with adaptation guidelines set forth by the World Health Organization (www. who.int/substance_abuse/research_tools/translation/ en). These standards require that psychologists and other professionals refrain from using a translated version until the reliability and validity of that new measure has been firmly established. The danger in administering tests that have not been validated is that clinicians interpret the results based on an assumption that the test continues to

function in the intended manner (Fantoni-Salvador & Rogers, 1997). Until the reliability and validity of the translated assessment measures have been determined, mental health professionals should refrain from using them, just as they would refrain from administering any other unvalidated measure (Allalouf, 2003; Hambleton, 2001). For mental health practitioners seeking to conduct a culturally competent assessment battery, it is imperative to begin by choosing the most appropriate tests for each particular client. This decision involves choosing tests with appropriate translations, interpretive norms, and validation studies. The following sections highlight important considerations for clinicians when selecting which tests to administer.

TRANSLATIONMETHODS Many measures of feigning were originally created in the English language (Correa & Rogers, 2010a). Therefore, practitioners often find themselves using translated versions of widely researched measures. When clinicians are knowledgeable regarding the test translation process, they are better able to choose psychological measures with high linguistic validity, which is an important component of construct validity, because it denotes that the different versions of a test are conceptually equivalent in each of the target languages. This section aims to describe methods most commonly used in translating psychological assessment measures. The test translation process has been equated to construction of a new test, requiring evidence for construct validity, statistical support, and assessment of bias at the item level (Jeanrie & Bertrand, 1999). Test developers must be prepared to provide each of these requisite components for a valid measure. Three basic approaches are generally used in translating written documents from one language to another: one-way translation, translation by committee, and back-translation (Marín & Marín, 1991). Each technique varies in complexity and has its own set of strengths and limitations.

One-WayTranslations One-way translations employ the simplest of translation techniques. Here, one bilingual individual uses dictionaries, reference materials, and his or her knowledge of both languages to create the

4.  Cultural and Transnational Perspectives 63

translated product (Marín & Marín, 1991). This approach is appealing because it is time-efficient and cost-effective, and uses the resources of only one person to achieve a translation. However, its simplicity also serves as the basis of most criticisms. Relying on a single person to translate the material leaves the translation vulnerable to error, because the translator’s work is left unchecked. When misinterpretations make their way into the final product, quality of the translated measure is adversely affected. For instance, Berkanovic (1980) demonstrated that a health survey with one-way translation into Spanish had different psychometric properties and lower reliability than its original English-language counterpart. One sound recommendation is to focus on the quality of the translator for improving the quality of a one-way translation. On this point, Hambleton and Kanjee (1995) stress that translators should be (1) highly proficient in both languages involved and (2) familiar with cultural groups associated with both languages. The latter recommendation helps in constructing translated items that flow well in the new language, retain the intended meanings, and are readily understood by the target population. If translators also have an understanding of the construct being measured, they are better able to preserve the intended meaning of test items. One-way translations may also be made more thorough via the use of judges to evaluate the final product (Jeanrie & Bertrand, 1999). Judges may evaluate the following three areas: 1. Content equivalence: relevance to that cultural group 2. Conceptual equivalence: maintaining construct validity 3. Linguistic equivalence: maintaining as direct a translation as possible, without jeopardizing content and conceptual equivalence. Test items attaining the highest scores on the relevant constructs can be compiled and edited in the final step. This framework attempts to remedy some of the major criticisms of one-way translation. However, no published data are found to indicate whether this leads to a better oneway translation or simply takes up resources that could best be put to use in implementing a more well-researched translation model. Despite these suggestions, researchers (Brislin, 1970; Marín & Marín, 1991) tend to agree that one-way translations should not be used. Instead, they conclude

that more translators should be involved, and that back-translation should be used for quality control.

Translation byCommittee A second translation technique involves a translation by committee (Marín & Marín, 1991). This approach utilizes two or more individuals who are highly conversant in both languages. Professionals independently produce their own translation without consulting the other translators. After the initial translations are complete, the coordinator can ask all translators to meet, compare their individual versions, and discuss and resolve the differences. In this manner, they create a final version incorporating the changes they have discussed. The goal of this process is to prevent problems, such as misinterpretation and awkward wording, that arise from relying too heavily on a single translator. Alternatively, the coordinator can ask one more persons (not involved in the original translations) to review each translator’s work and choose the best version (Marín & Marín, 1991). This option still falls under the rubric of “translation by committee,” because multiple translators are involved in the process. Translation by committee can be more accurate than one-way translation. Marín and Marín (1991) are quick to point out, however, that traits shared by the translators, such as cultural background, education level, and social class, might lead them to make similar errors in their independent translations. Ensuring that the committee consists of individuals with diverse cultural backgrounds within the culture of the target language reduces the risk of error caused by uniform interpretations (Martinez, Marín, & Schoua-Glusberg, 2006). However, committee discussion can never ensure that all possible mistakes are pointed out, because committee members may not feel comfortable criticizing each other’s translation (Marín & Marín, 1991).

Back-Translation A final translation procedure, commonly known as “back-translation,” is the most recommended by researchers (Brislin, 1986; Moreland, 1996), yet it remains the least used translation technique (Jeanrie & Bertrand, 1999). Its lack of use may be due to its time-consuming nature. Back-translation makes use of at least two bilingual individuals. As in one-way translation, one independently translates the original language (e.g., English) into the

64

I.  Concep t ua l Fr a me work

new language. At this point, a second translator takes the newly translated version and translates it back into the original language. It is critical that the translators work independently throughout this process and not be permitted to consult with one another. Two English-language versions of the measure now exist: the original version and the back-translated version. The two English versions are compared, and inconsistencies are identified. When differences are found, it is imperative to approach both translators, determine why the difference exists, and reach an agreement about the best option (Marín & Marín, 1991). A third party not involved in the original translation or backtranslation may also be commissioned to evaluate the two English versions (Brislin, 1970). Back-translation can be improved if the process is conducted more than once and with multiple translators. These iterations make the procedure more time-consuming and complex. However, the measure is reviewed by more bilingual professionals, which produces a better version of the instrument in the end (Marín & Marín, 1991). Backtranslation has been used extensively in creating Spanish-language versions of assessment tools as diverse as general health questionnaires (Marín, Perez-Stable, & Marín, 1989), structured interviews for the diagnosis of mental health problems (Burnam, Karno, Hough, Escobar, & Forsyth, 1983), and structured interviews for the assessment of feigned psychopathology (Correa & Rogers, 2010b; Rogers, Sewell, & Gillard, 2010). Marín et al. (1989) advocated the back-translation process finding the Spanish version of their survey was, indeed, equivalent to the English version after administering both versions to bilingual speakers. Likewise, Sireci, Yang, Harter, and Ehrlich (2006) conducted a study designed to determine how a more rigorous translation procedure (back-translation) compared to a simple translation procedure (one-way translation). They found that for many of the test items, back-translation produced results that were more comparable to the original English measure. Using their design, the Spanish Diagnostic Interview Schedule (DIS) was also found to be acceptably equivalent to the English DIS for bilingual participants (Burnam et al., 1983). An inherent limitation in the process of backtranslation is that it still relies on the translator’s interpretation of item meaning (Marín & Marín, 1991). For this reason, it is important to employ the same precautions that should be used for

“translation by committee.” Recruiting translators from varied educational, cultural, and social backgrounds minimizes errors caused by uniform interpretations (Martinez et al., 2006). Another criticism of back-translations involves the absence of guidelines as to how many independent translators are sufficient for a good translation (Cha, Kim, & Erlen, 2007). Some experts instead advocate using a combined technique (Jones, Lee, Phillips, Zhang, & Jaceldo, 2001) that employs back-translation and administers both versions of the test to bilingual participants in order to identify discrepancies before creating the final version. This method appears to incorporate equivalence testing (a recommended step for final validation) into the translation procedure (Hambleton, 2001).

CULTURALLY SPECIFIC RESPONSE STYLES THATAFFECT VALIDITY SCALESCORES Culturally specific response patterns emerge in several diagnostic measures for psychopathology. Consistent patterns of score elevations may be culturally normative for some test-takers and should not be construed as symptom exaggeration. Conversely, score elevations are often less evident for other groups due to culturally normative defensiveness, leaving existing assessment measures less sensitive in the detection of feigning for groups with the tendency to underreport. Distinct patterns are often apparent on clinical and validity scales of multiscale inventories used to detect minimization or exaggeration of symptoms (Correa & Rogers, 2010a; Molina & Franco, 1986). This section explores cultural characteristics that may affect the validity profile of an examinee’s responses on standardized assessment measures. The construct of machismo is among the response patterns that can significantly impact a patient’s self-report measures. Machismo is a gender schema consisting of behaviors, attitudes, and beliefs often espoused by Latin American men (Casas, Wagenheim, Banchero, & Mendoza-Romero, 1995). Factors of machismo contain positive aspects related to chivalry and negative aspects related to chauvinism. Despite little research in this area, available studies examining machismo, gender roles, and mental health have found that higher levels of machismo and restrictive emotionality may be associated with higher levels of depression and stress among Hispanic American men (Fragoso & Kashubeck, 2000).

4.  Cultural and Transnational Perspectives 65

Therefore, machismo bolsters the theory that low symptom endorsem*nt does not necessarily indicate subjective well-being among Hispanic Americans. Rather than indicate an absence of symptoms, underreporting on assessment measures may be more reflective of a culture-based hesitation in this clinical population to disclose symptoms of psychological distress (Correa & Rogers, 2010b). Cultural patterns for Hispanic males are particularly apparent on validity scales related to minimization of symptoms (Molina & Franco, 1986). Besides machismo, the conceptualization of extreme response style suggests that individuals of Hispanic and Mediterranean cultures have a tendency to respond either very low or very high when given choices on Likert-type scales, commonly used in the United States (Hui & Triandis, 1989). It is believed that these individuals consider extreme responses to be more sincere than “less decisive” responses located in the middle of a Likert-type scale. This distinction is most evident for individuals from Hispanic and Mediterranean cultures when contrasted with those from Asian cultures, who tend to respond in the middle of the scale (Zax & Takahashi, 1967). Notably, the language of a test can further magnify this cultural response style. In a study that administered the same items in two different languages to bilingual individuals, Gibbons, Zellner, and Rudek (1999) found that participants used more extreme ratings when responding in Spanish than in English. The theory of extreme response style suggests the possibility that Hispanic Americans may be just as likely to overreport as to underreport symptoms on a measure. More research is needed in this area to adequately understand the disparity in research findings to date. It is widely acknowledged that African Americans tend to score higher than European Americans on measures of social aggression and psychotic symptoms, even when studies control for age, gender, and level of education (Cuellar & Paniagua, 2000; Dana, 1993, 2000). Researchers believe that these results are closely tied to cultural identity and beliefs; as such, they are not necessarily indicative of psychopathology. Culture-related stressors, such as racism and adjustment to mainstream American culture, play an important role in the types of distress African Americans tend to report. Guarding against prejudice and residual anger toward racial tensions can manifest itself as aggression, hypervigilance, or paranoia on multiscale inventories (Marcella & Yamada, 2000; Ewart & Suchday, 2002).

LANGUAGE The effects of language are vitally important to consider when accounting for the accuracy of the assessment process. The psychometric properties of standardized assessment measures are likely to change when administered to individuals who differ culturally from the normative sample (Marín & Marín, 1991). Furthermore, multilingual individuals who are not tested in their preferred language may suffer a detachment effect (Bamford, 1991); they fail to adequately connect with the assessment questions and are unable to fully express their emotional and psychological issues. The detachment effect can result in poor communication about symptoms and less self-disclosure (Dana, 1995); however, it is often remedied when individuals are tested in their preferred language. For example, Guttfreund (1990) shows that bilingual Hispanic American patients who prefer to speak Spanish are more able to effectively express their emotions when tested in their preferred language rather than English. For multilingual individuals, clinicians must take into account a client’s language preference prior to beginning the assessment process. When a client is conversant in English and a second language but expresses only a minor preference for the second language, practitioners might choose the English version— given that extensive validation studies are more likely to be available. When a strong preference is expressed for another language and/or Englishlanguage abilities are limited, the translated version of the test is most appropriate.

ACCULTURATION In addition to the different response patterns among distinct cultures, level of acculturation should be assessed for each examinee taking a test normed on a cultural group different from their own. Acculturation is defined as the changes that occur in an individual’s beliefs and behaviors, as a result of interaction with his or her own ethnic group and another cultural group. Assessing acculturation allows clinicians to determine the applicability of interpretive norms and to decide whether culture-specific cutoff scores should be employed if they are available and well validated. Individuals with higher levels of acculturation may possess a greater understanding of their new culture and begin to accept and incorporate aspects

66

I.  Concep t ua l Fr a me work

of it into their daily lives. Individuals with low levels of acculturation continue chiefly to identify with the values of their ethnic group despite interaction with members of a different culture (Wagner & Gartner, 1997). In 1989, Berry, Kin, Power, Young, and Bujaki proposed a widely accepted, two-dimensional model of acculturation in which individuals feel a need to identify with both their own culture and a new culture. The individual can maintain one of four possible relationships with majority and minority cultures: • Assimilation: sole identification with the majority culture • Integration: identification with both cultures • Separation: sole identification with the minority culture • Marginality: no identification with either culture Berry et al.’s (1989) bidimensional model of acculturation takes into account an individual’s varying degrees of affiliation with minority and mainstream cultures. In contrast, unidimensional models of acculturation (e.g., Gordon, 1964) contend that one relationship must always be stronger than the other. In unidimensional models, individuals relinquish their ethnic culture, as they become more assimilated to the mainstream culture. In both models, distinct levels of acculturation increase the variety of possible response patterns, because salient differences also exist within cultures, not just between them, depending on how much an individual identifies with each of the cultures in question. However, unidimensional models might obscure the complexity of individual acculturation, by failing to recognize bicultural individuals, who identify strongly with both cultures (Ryder, Alden, & Paulhus, 2000). However, both models emphasize the notion that it is erroneous to assume that all members of the same ethnicity will respond similarly when interpreting assessment results. How acculturation affects responses to test items should also be established when characterizing new normative samples and cut scores.

APPROPRIATENESS OFAVAILABLE INTERPRETIVENORMS An important issue in the validity of any assessment measure used with ethnic/minority or nonEnglish-speaking populations is consideration of

the etic and emic qualities of the test (Dana, 1993, 2005). Etic measures are those with “universal” applications, whose constructs are equally applicable to individuals of all different groups. It is expected that an individual’s assessment results on an etic measure can be interpreted based on the same set of norms, regardless of the individual’s membership in any particular cultural group. Emic measures, on the other hand, are culture-specific measures; their clinical applications can be specific to populations based on age, gender, ethnicity, or any other grouping classification. Emic measures are only appropriate for use with the groups for whom they were designed. Researchers (Berry 1969, 1988; Dana 2005) have observed for some time that most standardized assessment measures were created and normed on samples that comprised mainly European Americans in the United States. Based on current clinical practices, tests are administered, scored, and interpreted uniformly according to guidelines established in the manual. Since the majority of interpretive norms were developed primarily on individuals of European American heritage, these assessment measures fall into the category of imposed etic tests. Without further testing on other cultures, the tests remain empirically valid for only the European American culture. The practice of adding proportionate but comparatively small samples of minority populations to test norms improves their representativeness. However, it is not a substitute for minority group validation studies to establish culturally relevant cutoff scores and interpretation guidelines. Without such work, test developers inadvertently imply that European American–based cutoff scores are universally valid and generalize to all cultures. Indeed, Henrich, Heine, and Norenzayan (2010) have noted that researchers routinely make claims based on samples of American college students and other Western, Educated, Industrialized, Rich, and Democratic (WEIRD) societies. The omission of ethnic minorities in test development effectively forces minority individuals into the same interpretative categories as European Americans, thereby creating a substantial possibility for assessment bias, misdiagnosis, and misinterpretation of test results for individuals from different nationalities or cultural groups (Dana, 1993; Todd, 2004). Assessment bias may be minimized when clinicians are well informed about the populations they are testing, recognize limitations of their measures, and use additional cultural measures to aid in their interpretation of assessment results (Dana,

4.  Cultural and Transnational Perspectives 67

2005). Adequate knowledge of concepts such as acculturation makes clinicians better equipped to choose culturally appropriate tests, as well as understand how best to apply test norms and interpretive guidelines. Differences between the test scores of individuals from other cultural groups and those from the dominant culture become problematic when interpretive norms lead to inaccurate predictions or diagnoses for minority individuals (Graham, 1990). Thus, mental health professionals should consider the validity of assessment data derived from standardized measures. Clinicians must always weigh the cultural characteristics of the person being assessed and how such attributes likely affect a person’s responses. In this manner, professionals working in the mental health field can refine their clinical interpretations of standardized measures, in order to make more accurate assessments of individuals from diverse cultural backgrounds and create treatment recommendations that are best suited for such patients (Dana, 2005). Clinical interpretations of test data should be made with adequate awareness of the limited generalizability inherent in many currently available assessment tools. In the last two decades, normative studies have increasingly included census-matched standardization samples with representation from minority cultural groups in the United States (Correa & Rogers, 2010a). European countries have also begun to develop their own norms for use with specific international populations (Giger & Merten, 2013; Merten et al., 2013). Such developments are invaluable for making culturally sensitive interpretations of clinical data. They represent an important step in fully validating psychological measures for three related but distinct constructs: culture, ethnicity, and diverse linguistic backgrounds. To ensure culturally competent test interpretation, clinicians must consciously search for alternative normative datasets. Often, culturally specific norms are not available in the published test manuals. Mental health practitioners are tasked with searching the available literature for published validation studies, then determining which of the researched normative groups best fits the demographic characteristics and acculturation of each individual examinee (Committee on Psychological Testing, Including Validity Testing, for Social Security Administration Disability Determinations; Board on the Health of Select Populations; & Institute of Medicine, 2015). For example,

the demographically adjusted norms developed by Heaton, Miller, Taylor, and Grant (2004) provide a useful resource. Colloquially referred to as the “Heaton Norms,” the book compiled by Heaton et al. is used for the culturally sensitive interpretation of over 50 cognitive and neuropsychological tests. The norms provide alternative cutoff scores for examinees based on differences in demographic variables, including age, gender, level of education, and race/ethnicity. The goal of such an expansive collection of demographically adjusted norms is to improve diagnostic accuracy by controlling for variables that have continually proved to significantly affect test performance. A vigorous debate continues in the literature regarding the use of race-specific norms (Manly & Echemendia, 2007). Many researchers note that specialized norms improve the sensitivity, specificity, and diagnostic accuracy of psychological assessment measures (Ardila, 1995; Heaton et al., 2004; Manly & Echemendia, 2007). Others contend that race-specific norms are often lower and more lenient, which can obfuscate deficits and potentially disqualify individuals from beneficial services (Manly, 2005). When it comes to feigning, being suspected of intentionally reporting false symptoms can also have significant consequences for an examinee. In clinical settings, this classification can preclude clients from receiving mental health interventions (Rogers, 1997, 2008), because in settings where resources are scarce, many mental health professionals believe it is their responsibility to ensure that only the truly sick receive access mental health treatment (Resnick, 1984). When encountered in a forensic setting, the ramifications can be even more serious. Not only might individuals be denied mental health care but the classification of malingering could also be used to discredit them throughout the trial process (Rogers & Shuman, 2005). Once individuals have been classified as malingerers, often it is difficult for them to prove the genuineness of their disorders in future situations (Berry, Baer, Rinaldo, & Wetter, 2002). As previously discussed, several culturally specific response styles may affect a test-taker’s presentation on validity scales (Casas, et al., 1995; Correa & Rogers, 2010a; Dana 1993, 2000; Hui & Triandis, 1989). Another point of contention when using racespecific norms centers on their theoretical legitimacy. Some researchers contend that using race as the defining construct for the creation of normative groups is misleading, because it disregards cultural or educational variables that may actually

68

I.  Concep t ua l Fr a me work

account for the differences in test performance (Manly & Echemendia, 2007). For example, Lynn (2002) found a correlation between skin tone and scores on a vocabulary test, attributing the higher scores to an increased proportion of European American genes in those participants. Lynn’s data were reanalyzed by Hill (2002), who controlled for childhood environmental factors such as educational achievement, parental education, and geography. Skin tone and “race” no longer correlated after controlling for these other variables, indicating that the differences in scores were unrelated to participants’ race. For practicing clinicians, these findings highlight the complexities involved in deciding which interpretive norms to use for a given examinee. It is generally accepted that clinical determinations for non-English-speaking individuals will be more accurate if their performance is compared to demographically and linguistically similar individuals (Heaton et al., 2004; Manly & Echemendia, 2007). Mindt, Byrd, Saez, and Manly (2010) recommended making decisions on a case-by-case basis, as well as using the best assessment measures and norms possible. It is important to note that demographically appropriate norms are not always available, so professionals may have to choose the closest approximation and note the limitations in the reports. Intelligence testing provides an excellent example of how test norms and translation procedures may affect clinical interpretations; researchers have long since pointed out that demographic variables such as age, gender, and culture affect an individual’s performance on cognitive tests (Kaufman & Lichtenberger, 2006). Using data from the English-language Wechsler Adult Intelligent Scale–Wechsler Memory Scale (WAISWMS) co-norming project, Heaton, Taylor, and Manly (2001) found that Hispanic American individuals generally achieved lower scores than their European American counterparts when both groups were tested in English. Using standard norms, between 15 and 25% of Hispanic individuals were misclassified as “impaired” on the WAIS and WMS, even when corrections were made for other factors, such as age, gender, and level of education. In order to reduce an apparent bias in the interpretation of the measure, normative adjustments were suggested by Heaton et al. (2001). Predictably, when using the corrected cutoff scores, Hispanic American individuals have nearly the same likelihood of being misclassified as their European American counterparts.

Kaufman and Lichtenberger (2006) hypothesized that lower scores for Hispanic individuals on verbal measures reflect (1) unfair language demands placed on individuals for whom English is a second language, and (2) the cultural content of some verbal test items. Similarly, the Standards for Educational and Psychological Testing from the American Educational Research Association, American Psychological Association, and National Council on Measurement in Education (1999, 2014) specify that any oral or written test is also inherently a measure of an examinee’s verbal skills, whether it aims to measure this construct or not. Therefore, seemingly discrepant performance on translated measures may be a product of the translation process, not necessarily an examinee’s response style. At a minimum, clinicians must provide caveats while interpreting assessment data and tailor treatment recommendations to different groups of ethnically diverse clients (Correa & Rogers, 2010a). Clinicians must weigh the pros and cons of each measure in choosing the most appropriate test for their clients. Manly and Echemendia (2007) advocate using a “cost–benefit matrix” to evaluate the potential outcomes of each decision, then choosing the interpretive norms with the least cost to the assessment client. High test specificity is important in classifications of malingering, because the consequences of a false-positive result can be very damaging to the client (Rogers & Shuman, 2005). For other clinical decisions, higher test sensitivity may be desirable.

USING ACCULTURATION MEASURES TOCHOOSE TESTS ANDINTERPRETIVENORMS When norms do not fully match the characteristics of a particular client, approximations will need to be used by the clinician. Assessing acculturation as part of the evaluation aids in these approximations, as well as case conceptualization (Mindt et al., 2010). With low levels of acculturation, clients continue to identify with the values of their ethnic group despite frequent interactions with the mainstream culture/normative group (Wagner & Gartner, 1997). This process may account for within-culture differences in response patterns, personality characteristics, and psychopathology (Okazaki & Sue, 1995). Some clients have particular difficulty with the acculturation process and experience “marginality” with distress and alienation, because they do not identify with either culture (Berry et al., 1989).

4.  Cultural and Transnational Perspectives 69

Therefore, comprehensive evaluations often benefit from the inclusion of an acculturation measure as part of the assessment process (Dana, 1993; Wagner & Gartner, 1997). The use of an acculturation measure often clarifies the level of confidence of the interpretations and highlights important caveats to include in the final report. Specifically, acculturation measures provide tangible data about the limitations of interpretive test norms when the client’s current acculturation level deviates from that of the normative group. Many acculturation measures are available. Psychologists may wish to consider the Scale of Ethnic Experience (SEE; Malcarne, Chavira, Fernandez, & Liu, 2006), which has been widely adapted for diverse cultures, and the Acculturation Rating Scale for Mexican Americans–II (ARSMA-II; Cuellar, Arnold, & Maldonado, 1995), which has a widely researched and validated Spanish Language version.

MEASURES OFFEIGNEDPSYCHOPATHOLOGY This section investigates properties and clinical utility of the most widely researched measures of feigned psychopathology currently available for international populations. For many of these assessment measures, there remains a dearth of available research validating clinical use with populations different than the normative groups published in test manuals. Consequently, it is difficult to estimate the extent to which individuals are affected by “imposed etic” interpretative practices. This chapter addresses only those tests with published feigning research involving international populations. An in-depth examination of the each test’s construction and validation research is included. The goal is to provide clinicians with important insights when testing clients of diverse ethnic backgrounds, as well as to discuss important caveats.

The Structured Interview ofReported Symptoms—2ndEdition The original version of the Structured Interview of Reported Symptoms (SIRS; Rogers, Bagby, & Dickens, 1992) was a comprehensive measure designed to evaluate feigned mental disorders. Items were generated for eight primary scales, with each scale devoted to a single detection strategy: Rare Symptoms, Symptom Combinations, Improbable and Absurd Symptoms, Blatant Symptoms, Subtle Symp-

toms, Selectivity of Symptoms, Severity of Symptoms, and Reported versus Observed Symptoms. More recently, the SIRS–2nd Edition (SIRS-2; Rogers, Sewell, & Gillard, 2010) has demonstrated excellent reliability, validity, and classification accuracy (sensitivity = .80; specificity = .975; positive predictive power = .91; negative predictive power = .91) that results in an excellent overall classification of .91 and a very small false-positive rate of 2.5%. The SIRS-2 retained the original primary scales and sought to reduce false-positives rates through additional scales and a new decision model. These additional scales include the Rare Symptom Total (RS-Total) Index and the Modified Total (MT) Index. Using the MT Index, the SIRS2 uses two unclassified groups: indeterminate–evaluate and indeterminate–general. Individuals in the indeterminate–evaluate group should be further screened for feigning. The indeterminate–general category is composed of individuals with a relatively low likelihood of feigning. The final index is the Supplementary Scale (SS) Index, involves the sum of four supplementary scales, one of which is a new scale on the SIRS-2. This new addition, the Improbable Failure (IF) scale is composed of simple cognitive tasks. It is important to note that whereas there are no published data to date, clinical evidence indicates that the IF scale may not be appropriate for clients with limited proficiency in English or those educated outside the United States. This important caveat is explored further in the next section. The SpanishSIRS-2

The Spanish SIRS-2 was translated from the original English-language version utilizing the backtranslation method with multiple translators. In the first phase, three bilingual psychologists independently translated the English-language SIRS to Spanish. These three translators then met and reviewed any discrepancies in language or intended meaning. They developed a consensus on the best Spanish translation. For the second phase, a fourth psychologist used that Spanish version for the backtranslation into English. To avoid bias in the translation process, this psychologist had no knowledge of the original English-language SIRS. A fifth bilingual psychologist then independently compared the original and back-translated English versions, resolving a small number of discrepancies. Before its publication, two additional bilingual psychologists reviewed the Spanish SIRS-2 for any grammatical errors (Correa & Rogers, 2010b, 2013).

70

I.  Concep t ua l Fr a me work

RELIABILITY, VALIDITY, ANDCLINICALUTILITY

The reliabilities of the Spanish and English SIRS2 are compared for internal consistencies (alphas) and interrater reliabilities (Table 4.1). Alpha coefficients for the Spanish SIRS-2 equal and sometimes slightly exceed the English version. A similar pattern is also observed for interrater reliabilities. On average, interrater reliability for the Spanish SIRS-2 indicates nearly perfect agreement between evaluators (mean r = .995). Spanish SIRS-2 validation has focused primarily on discriminant validity, or its ability to accurately distinguish feigners from genuine responders with true mental disorders (Correa & Rogers, 2010b; Rogers et al., 2010). The primary scales of the Spanish SIRS-2 evidence very large effect sizes, with Cohen’s ds from 1.38 to 2.47. Utility estimates for the decision model produce very similar results to the English SIRS-2 (e.g., sensitivity = .88; specificity = .92). Its false-positive rate is 8.0%. This finding underscores the importance of using more than one feigning measure in determinations of feigned mental disorders. Rogers et al. (2010) also conducted a small, ethnically diverse study of linguistic equivalence with bilingual outpatients in Miami. Each patient was administered both the Spanish and English versions of the SIRS-2. As reported in the SIRS2 manual, the differences between the English and Spanish administrations were very small for TABLE 4.1.  Reliability of the Spanish SIRS-2 in Comparison to the English SIRS-2

Alpha Scale

Spanish

SEM

English

Spanish

English

RS

.81

.81

.00

0.44

SC

.89

.80

.00

0.46

IA

.84

.83

.00

0.38

BL

.96

.91

.00

0.74

SU

.95

.91

.90

0.68

SEL

NA

NA

.00

0.66

SEV

NA

NA

.67

0.00

RO

.76

.81

.17

0.59

Average

.89

.85

.22

.49

RS-Total

.82

.86

.00

.71

Note. Derived from Rogers et al. (2010) and Correa and Rogers (2010b). English alphas do not include partial SIRS administrations.

the primary scales, averaging less than half of one point. Overall categorization of feigning versus genuine disorders remained identical across both language versions. With respect to cultural equivalence, one important caveat involves the Improbable Failure (IF) scale, which is a supplementary scale used to screen for feigned cognitive impairment. The IF scale requires examinees to quickly complete simple verbal tasks (rhyming and opposites) commonly practiced in U.S. elementary schools. These tasks are likely to be unfamiliar for native Spanish speakers with low levels of acculturation, low educational attainment, or those educated outside of the United States. The effectiveness of the Spanish IF scale remains to be investigated. Commonsensically, the Spanish IF scale should not be utilized with any examinees unfamiliar with rhyming and opposites. Fortunately, these skills can easily be tested prior to the SIRS-2 administration. The Spanish IF scale cannot be administered to participants who cannot comprehend or apply the measure’s instructions. As a broader principle, simple cognitive tasks commonly used in mental status exams (or on the IF scale) may be answered incorrectly by individuals who lack familiarity with the tasks. This point applies to those putting forth a genuine effort, even highly educated, unimpaired individuals from foreign countries (Ostrosky-Solís et al., 2007). Therefore, the IF cognitive tasks may be culturally inappropriate for some examinees, and low scores do not necessarily screen for cognitive feigning. By pretesting the required cognitive skills, clinicians may prudently omit the IF scale from their SIRS-2 administrations (Richard Rogers, personal communication, June 7, 2016). The two Spanish SIRS-2 validation studies included outpatients with varied cultural backgrounds. The Miami study (Rogers et al., 2010) predominantly included clients of Puerto Rican, Central American, and South American cultural heritage. In contrast, the monolingual (Spanish only) study included outpatients, mostly of Mexican descent, with low levels of acculturation to mainstream American culture. Looking to the future, TEA Ediciones, a leading test publisher in Spain, is developing a European Spanish version of the SIRS-2 and plans to conduct further studies in Spain and other Spanish-speaking countries. The ChineseSIRS-2

Currently, one published article (Liu et al., 2013) described two validity studies conducted on a

4.  Cultural and Transnational Perspectives 71

Chinese-language version of the SIRS-2. The researchers refer to their translation process as rigorous. In a personal communication (Richard Rogers on June 2, 2016), the process was briefly described: The English to Mandarin translation was completed by a research assistant who had no knowledge of the SIRS-2; this translation was reviewed and corrected by a clinical psychologist with knowledge of the SIRS-2. The back-translation was independently carried out by a graduate student in clinical psychology. The first study by Liu et al. (2013) was a simulation design, in which Chinese undergraduate students were asked to (1) respond honestly or (2) feign symptoms of mental illness. These two groups were then compared to a clinical comparison sample (i.e., psychiatric inpatients responding honestly). The second study, a known-groups design, compared psychiatric outpatients to a group of suspected malingerers from forensic settings in China. RELIABILITY, VALIDITY, ANDCLINICALUTILITY

For reliability, internal consistences were calculated for each of the Chinese SIRS-2 primary and supplementary scales. Cronbach’s alpha coefficients were mostly high for the primary scales (M = .83) but only moderate for Reported versus Observed (RO; alpha = .74). Supplementary scales ranged from .79 to .84. For the simulation study, convergent validity of the Chinese SIRS-2 was examined in relationship to the Chinese Minnesota Multiphasic Personality Inventory–2 (MMPI-2) F scale, producing low to moderate correlations (.22 to .48) with the SIRS-2 primary scales. Of greater significance, discriminant validity remained impressively high for Chinese SIRS-2 primary scales in both the simulation (mean d = 1.79) and known-groups (mean d = 1.79). These effect sizes strongly support the Chinese SIRS-2 ability to discriminate feigners from forensic inpatients with genuine disorders. Finally, Liu et al. (2013) tested the effectiveness of the Chinese SIRS-2 decision model. Focusing in the known-groups design, the utility estimates produced outstanding results, with a sensitivity of .85 and specificity of 1.00. Qualified psychologists and psychiatrists, fluent in Mandarin, may wish to include the Chinese SIRS-2 in their clinical assessments when feigned mental disorders are suspected. As always, assessments of malingering should include a multifaceted evaluation, with multiple measures and sources of data. As high-

lighted by Liu et al.’s research, some individual scales may vary from those reported in the SIRS-2 manual. Therefore, conclusions should be drawn from the SIRS-2 decision model rather than individual scales. As a caution, additional research is clearly needed for cultural minorities in China and Taiwan, where Mandarin is predominantly the spoken language. For Cantonese (e.g., Hong Kong) and other Chinese variants, different translations of the Chinese SIRS-2 are likely warranted.

The Personality AssessmentInventory The Personality Assessment Inventory (PAI; Morey, 1991) is a 344-item multiscale inventory designed to assess personality traits and patterns of psychopathology. The measure contains 11 clinical scales, five treatment scales, and two interpersonal scales. In addition, the PAI contains four standard validity scales for measuring response styles and profile validity (Morey, 1991, 2007). The normative samples included in the PAI manual create some limitations in interpreting results for members of different cultural groups. Ethnic differences are examined for the census-matched standardized sample but were not considered for the representative clinical sample. Clinical standardization samples, described in the more recent version of the PAI manual (Morey, 2007) are composed of 78.8% European Americans, 12.6% African Americans, and 8.6% “other” minority groups. Normative data for all ethnic/minority groups except African Americans is collapsed into a single “other” group. This grouping propagates the erroneous assumption that all other ethnic minorities are alike, except for African Americans, and potentially masks important between-culture differences in response style (see Romain, 2000; Todd, 2004). For instance, high scores for one cultural group on a particular scale might be balanced out by low scores from another group, masking disparities in scale elevations. The SpanishPAI

To date, the PAI has been published in English and Spanish. The PAI test manuals (Morey, 1991, 2007) do not describe the translation process for the Spanish version. However, its publisher, Psychological Assessment Resources, has standardized its translation process to include an independent back-translation and review/approval by the test’s author (see www4.parinc.com/prores/permis-

72

I.  Concep t ua l Fr a me work

sions.aspx). When a bilingual client expresses only a minor language preference, practitioners might choose the English version due to its extensive validation. When a strong preference is expressed for Spanish, or English language abilities are limited, the Spanish PAI is the most appropriate. Clinicians should note that the Spanish PAI is currently interpreted using the same norms as the English version. CULTURALLY SPECIFIC RESPONSESTYLES

Using the PAI, Hopwood, Flato, Ambwani, Garland, and Morey (2009) found evidence of socially desirable response styles in Hispanic Americans. Hispanic American undergraduates attained higher scores than mainstream culture Americans on the three PAI socially desirable response indicators: PAI Defensiveness Index (DEF), PAI Cashel Discriminant Function (CDF), and the Positive Impression Management (PIM) scale. These differences produced only modest effect sizes (ds = 0.28, 0.37, and 0.38, respectively), so it is unclear whether they indicate a cultural response style that potentially affects test validity. Romain’s (2000) dissertation indicates that more than 40% of the PAI protocols from Hispanic Americans were considered “invalid” based on the standard validity scale cutoff scores outlined in the PAI manual (Morey, 1991), as compared to 20% of the European American profiles. In her sample, Hispanic Americans had higher PIM scores in comparison to mainstream culture Americans (Cohen’s d = 0.60). Romain’s data suggest the presence of a culturally specific response style not previously discussed in the PAI literature. Hispanic Americans are scoring in a non-normative or atypical manner on items unrelated to psychopathology (i.e., Infrequency [INF] scale). Because INF elevations may reflect carelessness, confusion, or reading difficulties, psychologists may wish to consider issues of reading comprehension, knowledge of mainstream culture, and acculturation for the Spanish PAI. Given its large effect size in Romain’s sample (d = 1.00), INF scale may indicate a culturally specific response pattern beyond differences in reading abilities. Correa (2013) also found that approximately one-third of profiles were considered invalid for Spanish-speaking outpatients using the cutoff scores recommended in the PAI manual for the Inconsistency (ICN) scale and the INF scale. Significant differences in INF scores among honest, malingering, and defensive participants in

a simulation study suggest the possibility of idiosyncratic responding among Hispanic American patients in both underreporting and overreporting symptoms on the Spanish PAI. There is no clear pattern in INF elevations based on experimental condition. Each condition demonstrated endorsem*nt of different INF items. To date, specific properties of the INF scale on the Spanish PAI and the possibility of a culturally specific response style have not been explicitly researched. Elevated ICN scores are generally evidence of inconsistent responding. However, the ICN scale was significantly elevated in a clinical sample, regardless of experimental condition. Using Morey’s (2007) general guideline, only participants with ICN scores lying 2 standard deviations above the sample mean in Correa’s (2013) study were considered significantly elevated and excluded from analysis. This practice yielded far fewer invalid protocols and may indicate a need for revised cutoff scores on the Spanish version of the PAI. RELIABILITY, VALIDITY, ANDCLINICALUTILITY

Rogers and his colleagues (Fantoni-Salvador & Rogers, 1997; Rogers, Flores, Ustad, & Sewell, 1995) conducted several studies on the clinical usefulness of the Spanish PAI. Rogers et al. (1995) conducted the first validation study on mostly first- and second-generation Mexican Americans involved in mental health services. The internal consistencies for clinical and treatment scales were generally lower than those found in the representative clinical sample and other non-Hispanic studies (see Morey, 2007). Measures of internal reliability were adequate for the clinical and treatment scales. Research has also found good test–retest reliability for the Spanish PAI. The equivalency between the English and Spanish versions of the PAI was examined for a small bilingual sample and produced moderately high correlations (Fernandez et al., 2008; Rogers et al., 1995). Only two investigations have addressed Spanish PAI validity indicators, one published study that specifically investigates feigning and defensiveness (Fernandez, Boccaccini, & Noland, 2008) and one unpublished dissertation (Correa, 2013). According to Correa, Spanish PAI validity indicators generally produced moderate to large effect sizes in a sample of Spanish-speaking patients with “traditional” levels of acculturation (mean d = 1.08; range from 0.72 to 1.35). Specifically, PAI scales utilizing rare-symptoms strategies (Negative Impression Management [NIM] and Negative Dis-

4.  Cultural and Transnational Perspectives 73

tortion Scale [NDS]) demonstrated moderate to large effect sizes. In contrast, the spurious patterns strategies (Malingering Index [MAL] and Rogers Discriminant Function [RDF]) that focus on patterns of response that are characteristic of malingering but very uncommon in clinical populations (MAL and RDF) appeared to be generally less effective, with ds < 1.00 (Correa, 2013). The discriminability of validity scales was also explored for PAI measures of defensiveness and socially desirable responding, specifically, the PIM, DEF, and CDF. Spanish PAI validity indicators demonstrated moderate to very large effect sizes (mean d = 1.27; range from 0.94 to 1.68). Notably, CDF produced the smallest effect size (d = 0.94) of all Spanish PAI validity indicators (Correa, 2013). This finding is unexpected, because it has been found to be more accurate in detecting defensiveness in the English version of the PAI than either the PIM or DEF scores alone (Cashel, Rogers, Sewell, & Martin-Cannici, 1995; Morey, 2007). Clinical utility of the Spanish PAI increases as different cutoff scores are employed. Correa (2013) suggests various optimized cut off scores for the PAI validity index for use with mostly monolingual Spanish-speaking outpatients with “traditional” levels of acculturation. Fernandez et al. (2008) suggested cutoff scores to maximize the overall classification rate in a sample of bilingual university students. Clinicians may wish to implement the cutoff scores established with the population that best matches the demographic characteristics of each particular client they assess. It is important to note that a European Spanish version of the PAI is now available with European norms (Ortiz-Tallo, Santamaria, Cardenal, & Sanchez, 2011). LINGUISTICEQUIVALENCE

In a study using bilingual university students and community members, Fernandez et al. (2008) noted that validity scales on the English and Spanish PAI showed relatively equivalent levels of performance when differentiating honest responders and individuals asked to feign or respond defensively. The Spanish PAI clinical scales showed a moderate to good correspondence between Spanish and English versions (mean r = .72) and good test–retest reliability between Spanish language administrations (mean r = .79). Rogers et al. (1995) also found that the Spanish PAI demonstrated moderate correspondence between both language versions (mean r = .68). Additionally Rogers et al. (1995) demonstrated generally adequate alpha co-

efficients for Spanish PAI clinical scales (M = .68; range from .40 to .82) and treatment and interpersonal scales (M = .62; range from .40 to .82). For the PIM scale, Fernandez et al. (2008) found moderately high English-to-Spanish correlations for honest responders and the underreporting condition. These correlations are in stark contrast to the PIM correlation found by Rogers et al. (1995) in a population of Hispanic American patients. In addressing the disparity between these two studies, Fernandez et al. (2008) noted that marked differences in linguistic equivalence may contribute to differences in the samples of the two Spanish PAI studies. Specifically, they utilized a nonclinical, better educated sample than the Rogers et al. (1995) clinical outpatient sample. Neither study examined level of acculturation, so it is not possible to determine whether that also played a role in the disparity between the two studies. Another contributing factor may be that qualities specific to the PIM scale limit its effectiveness and stability among certain samples of Hispanic American individuals. Specifically, Rogers et al. found a modest correlation of .21 for the PIM scale, but much higher correlations for the remaining validity scales (i.e., INC, INF, and NIM), which ranged from .58 to .83. PIM was also identified as having the smallest effect size (d = 0.13) when differentiating between Hispanic American and European American students (Hopwood et al., 2009). Clinicians may wish to weigh the results of Rogers et al. (1995) more heavily, since they correspond better to the demographics of typical clinical referrals.

The Structured Inventory ofMalingeredSymptomatology The Structured Inventory of Malingered Symptomatology (SIMS; Smith, 1997; Smith & Burger, 1997) is a 75-item, self-report screening measure for feigned psychopathology and cognitive impairment. The DutchSIMS

The Dutch SIMS was created in a multistep process. English SIMS items were translated to Dutch, then back-translated. Also, culturally specific references, such as American dollars, were replaced by Dutch equivalents. The interpretive cutoff score of 16, recommended by Rogers, Hinds, and Sewell (1996) was used in a validation study by Merckelbach and Smith (2003).

74

I .  C o n c e p t u a l F r a m e w o r k

RELIABILITY, VALIDITY, ANDCLINICALUTILITY

Merckelbach and Smith (2003) conducted three studies. They administered the Dutch SIMS to undergraduate students and a small sample of 10 psychiatric inpatients. In the first study, undergraduate students completed the Dutch SIMS twice, with instructions to respond honestly. Three weeks elapsed between testing sessions, so that test–retest reliability could be evaluated. In the second study, undergraduates completed the Dutch SIMS and other self-report measures of anxiety and depression under honest conditions. The third study consisted of undergraduates responding honestly, inpatients responding honestly, and undergraduate students simulating amnesia, schizophrenia, or neurological problems. Findings demonstrate that the Dutch translation of the SIMS has good test–retest reliability and internal consistency. Findings of the simulation study show that undergraduate students obtain higher SIMS scores than both the undergraduate control group and psychiatric inpatients. Sensitivity, specificity, and positive predictive power rates were all high (≥ 0.90). Finally, Merckelbach and Smith (2003) noted a 16% false-positive rate for the SIMS misclassifying undergraduates who scored high on a depression questionnaire. They noted that the SIMS is a screening measure, and a classification of malingering warrants further investigation and additional psychological testing.

The Miller Forensic Assessment ofSymptomsTest The Miller Forensic Assessment of Symptoms Test (M-FAST) is a 25-item screening measure used to detect feigned psychopathology by implementing multiple detection strategies (Miller, 2001). The M-FAST has been translated into Korean and Spanish. However, most of the existing research on the M-FAST has been conducted on European Americans (Montes & Guyton, 2014). The SpanishM-FAST

The Spanish version of the M-FAST parallels the English version. The adaptation process used to create the Spanish M-FAST rigorously followed the adaptation guidelines set forth by the ITC (Hambleton, 2001). Developers used a multistep back-translation process with multiple bilingual and bicultural psychologists of diverse Latin American backgrounds. All translators worked

independently and produced a final version of the M-FAST that was highly comparable to the English M-FAST. The developers of the Spanish M-FAST have conducted the only published validation study (Montes & Guyton, 2014). Results of the study are discussed below. RELIABILITY, VALIDITY, ANDCLINICALUTILITY

Internal consistency for the Spanish M-FAST total score was very high (alpha = .97), with average interitem correlations in the optimal range (M = .42). All scales also demonstrated good internal consistency, with very high interitem correlations. In a simulation study using a sample of bilingual Hispanic American inmates, participants in both coached and uncoached malingering conditions demonstrated significantly higher total scores on the Spanish M-FAST than participants in the honest condition. Utility estimates were used to evaluate the effectiveness of both the English and Spanish M-FAST to correctly identify feigning in bilingual Hispanic incarcerated males. Notably, Montes and Guyton (2014) employed a total cutoff score of 5 instead of the cutoff score of 6 recommended by Miller (2001) in the M-FAST manual. Miller’s cutoff score produced lower sensitivity and negative predictive power estimates (.91 and .85, respectively) in the English M-FAST for this sample. However, the absence of a clinical comparison sample limits the generalizability of these results. LINGUISTICEQUIVALENCE

For participants in the honest condition, the English and Spanish versions of the M-FAST correlated highly (r = .96, p = .01). Significant positive correlations were also found between the total scores of the English- and Spanish-language versions of the M-FAST for the uncoached and coached malingering conditions. The English M-FAST (M = 0.82, SD = 1.24) were very similar to the Spanish M-FAST (M = 0.82, SD = 1.14). Results provide evidence for good language equivalence between the English and Spanish versions of the M-FAST.

MEASURES OFFEIGNED COGNITIVEIMPAIRMENT Assessment of feigned cognitive impairment is a rapidly growing field in Europe. To date, significantly more research has been conducted in North America, as feigning studies have long been con-

4.  Cultural and Transnational Perspectives 75

sidered taboo in international markets (Merten et al., 2013). Ruiz Sanchez de Leon and Gonzalez Marquez (2012) identify a number of neuropsychological tests that assess cognitive feigning in European Spanish speakers. A smaller amount of research has been conducted in Germany, Great Britain, The Netherlands, and Spain, and there are a handful of published studies from Austria, Portugal, and Switzerland. There is virtually no published research from other large countries, such as Italy and France. This disparity in distribution of research could be a result of the aforementioned acceptance or reluctance in some countries regarding the exploration of malingering (Merten et al., 2013). Currently, the most widely used measures of cognitive feigning fall into two camps: (1) forcedchoice/below chance performance and (2) tests using the “floor effect.” Many of the available measures have been widely researched and validated in North America, using normative groups composed of Americans, with small representations of ethnic/minority groups. North American English norms are the only interpretive guidelines currently available for several of these measures, so clinicians must remain aware of limited generalizability for international or non-English-speaking clients. The Montreal Cognitive Assessment (MoCA; Nasreddine et al., 2005) provides a good example of how test translations are sometimes widely available yet validity studies may continue to lag behind. Developed in Montreal, Canada, the MoCA is a brief neurological screen for mild cognitive impairment. It has been translated into approximately 60 different languages and dialects, but validation studies have only been conducted on 29 of the translations and specific normative data are available for five versions. The following sections focus on measures of cognitive feigning with published validity data and specialized norms for international populations.

Forced-Choice/Below-ChancePerformance Research indicates that symptom validity tests appear to retain construct validity and remain effective tools when adapted into other languages and other Western cultures. The primary reason for this may be that these measures have such a low test ceiling that they are successfully performed even by patients with low IQ or significant neurological damage (Giger & Merten, 2013). Clinicians must guard against blindly generalizing this guideline, however. Research also indicates that efficacy of an adapted test is highly dependent on the quality

of the translation/adaptation. Cultural differences may still lead to misdiagnosis even when test adaptations are of high quality, because certain concepts are lost in translation. For example, concepts such as seasonal changes, school-based tasks, and visual stimuli (e.g., drawings of certain foods) may be so culturally irrelevant that these test items are missed by highly educated, neurologically healthy individuals. Often, available interpretive norms do not account for these differences in international populations, so clinicians must always interpret results with caution (Ostrosky-Solis et al., 2007). The following tests employ forced choices to assess symptom validity and have published research to bolster their suitability for use with some foreign populations. The Test ofMemoryMalingering

For a detailed description of the Test of Memory Malingering (TOMM; Tombaugh, 1996), please see Frederick (Chapter 17, this volume). A European Spanish version of the TOMM was published in 2011 with European Spanish norms (Vilar López, Pérez García, & Puente, 2011). The original English version of the TOMM is widely used in Great Britain, but no British validation studies have been published. There is an indication that the TOMM may also be frequently used in other European countries, but there are no validation studies on its use with international populations (Merten et al., 2013). The Word MemoryTest

The Word Memory Test (WMT; Green, 2003) is available in 10 languages. Please refer to Frederick (Chapter 17, this volume) for a detailed description of the WMT. Validation research has been conducted on some of the adapted versions, but most versions rely on interpretive guidelines established for the original English-language version. THE DUTCHWMT

Rienstra, Spaan, and Schmand (2009) published reference data for the memory subtests of the WMT. They established linguistic equivalence between the English and Dutch versions of the WMT, with comparable mean subtest scores for the Canadian and Dutch samples. Unfortunately, an unknown number of participants were removed from the study because their scores demonstrated poor effort, so the performance of the Dutch

76

I.  Concep t ua l Fr a me work

WMT cannot be investigated as a test for feigning. Additionally, the absence of a clinical comparison sample limits its generalizability. THE BRITISHWMT

The WMT is one of the most commonly used symptom validity tests in the United Kingdom (McCarter, Walton, Brooks, & Powell, 2009). Hall, Worthington, and Venables (2014) conducted the only published U.K. study with nonlitigant, nonmalingering patients classified as having mild head injury. Patients were established as nonmalingerers through their scores on a battery of six different feigning tests. WMT performance data were analyzed only for the patients who demonstrated adequate effort on the other tests. WMT analysis demonstrated that Immediate Recognition (IR) and Delayed Recognition (DR) scales had acceptable false-positive rates of less than 10%. The Consistency Index (CI), however, yielded an unacceptably high false-positive rate of 18%. Hall et al. specified that participants who “failed” the CI managed to pass the other feigning tests in the administered battery, which implies that the WMT may misclassify genuine patients. The researchers recommended an investigation of alternative cutoff scores for populations with acute mild head injury. A cutoff score of 75% on the WMT provided more acceptable levels of specificity for their sample. In the context of this chapter, it is important to note that cultural differences between U.K. examinees and the original normative sample on the WMT may also affect the generalizability of recommended cutoff scores.

SUMMARY ANDRECOMMENDATIONS FORCLINICALPRACTICE Validity research on the use of standardized tests with culturally and linguistically diverse clients is still in its infancy on an international level. To date, validation studies are particularly scarce for the assessment of feigning (Merten et al., 2013). As a result, test adaptations are often made available for clinical use without published research to demonstrate its cultural or linguistic appropriateness (Fernandez, Boccaccini, & Noland, 2007; Montes & Guyton, 2014). In order to responsibly choose appropriate psychometric measures, clinicians must be knowledgeable in test construction, test translation, current validation studies, and

cultural factors that potentially influence test validity. The following guidelines are meant to provide useful recommendations for mental health practitioners assessing feigned symptomatology in individuals who are culturally different than those included in the normative sample: 1.  If questioning the appropriateness of a testing measure for a particular client, administration of an acculturation measure may be helpful. Acculturation data could preclude the use of certain tests, illuminate interpretive caveats that could be included in the report, and provide guidance as to which interpretive norms may be most appropriate. 2. Cautionary statements should be included for all interpretations involving clients with low levels of acculturation. Many tests normed in the United States have not yet been extensively researched for clients identifying closely with a minority culture. 3. Many tests used to detect feigned psychopathology in the United States do not have sufficient research to indicate how the validity scales perform with international populations. In choosing which tests to administer, it is best to select a measure with published validity data to avoid imposed etics in test interpretation. If sufficiently established, acceptable validity data may include alternative cutoff scores for certain populations. 4.  When alternative cutoff scores or interpretive norms are not available for a particular population, clinicians may opt to cautiously interpret the test, using what they know about test performance for that particular cultural group. Cautionary/descriptive statements are immensely useful in clarifying any clinical interpretations that diverge from published test norms. Examples of suitable statements include a. “Research indicates elevation on this particular scale could be due to cultural differences.” b. “On this test, English norms were used for scoring, though research indicates the Spanish version performs similarly to the original English version.” c. “Demographically adjusted normative data were utilized when available, using minority norms to most accurately reflect this patient’s environmental, cultural, and educational background. Otherwise, specific test publisher norms were used.”

4.  Cultural and Transnational Perspectives 77

5.  If a client’s background clearly does not fit into one set of available interpretive norms, it is acceptable to score the tests using more than one set of norms. A comparison and discussion of the examinee in reference to each normative group may be helpful in contextualizing how the examinee’s culture may or may not affect interpretation of results. 6.  In communicating test results to either patients or other mental health professionals, it is important to clearly convey the limitations of each test with members of international populations when the test has been designed for persons with mainstream American values.

REFERENCES Allalouf, A. (2003). Revising translated differential item functioning items as a tool for improving crosslingual assessment. Applied Measurement in Education, 16(1), 55–73. American Educational Research Association, American Psychological Association, & National Council on Measurement in Education. (1999). Standards for educational and psychological testing. Washington, DC: American Educational Research Association. American Educational Research Association, American Psychological Association, & National Council on Measurement in Education. (2014). Standards for educational and psychological testing. Washington, DC: American Educational Research Association. American Psychological Association. (2002). Ethical principles of psychologists and code of conduct. American Psychologist, 57, 1060–1073. Anastasi, A. (1988). Psychological testing (6th ed.). New York: Macmillan. Ardila, A. (1995). Directions of research in cross-cultural neuropsychology. Journal of Clinical and Experimental Neuropsychology, 17(1), 143–150. Bamford, K. W. (1991). Bilingual issues in mental health assessment and treatment. Hispanic Journal of Behavioral Sciences, 13(4), 377–390. Berkanovic, E. (1980). The effect of inadequate language translation on Hispanics’ responses to health surveys. American Journal of Public Health, 70(12), 1273–1276. Berry, D. T. R., Baer, R. A., Rinaldo, J. C., & Wetter, M. W. (2002). Assessment of malingering. In J. N. Butcher (Ed.), Clinical personality assessment: Practical approaches (2nd ed., pp.269–302). New York: Oxford University Press. Berry, J. W. (1969). On cross-cultural comparability. International Journal of Psychology, 4(2), 119–128. Berry, J. W. (1988). Imposed etics–emics-derived etics: The operationalization of a compelling idea. International Journal of Psychology, 24(6), 721–735.

Berry, J. W., Kin, U., Power, S., Young, M., & Bujaki, M. (1989). Acculturation attitudes in plural societies. Applied Psychology: An International Review, 38(2), 185–206. Bersoff, D. N. (Ed.). (2004). Ethical conflicts in psychology. Washington, DC: American Psychological Association. Brislin, R. W. (1970). Back-translation for cross-cultural research. Journal of Cross-Cultural Psychology, 1(3), 185–216. Brislin, R. W. (1986). The wording and translation of research instruments. In W. J. Lonner & J. W. Berry (Eds.), Field methods in cross-cultural research (pp.137–164). Newbury Park, CA: SAGE. Burnam, M. A., Karno, M., Hough, R. L., Escobar, J. I., & Forsyth, A. B. (1983). The Spanish Diagnostic Interview Schedule: Reliability and comparison with clinical diagnoses. Archives of General Psychiatry, 40(11), 1189–1196. Casas, J. M., Wagenheim, B. R., Banchero, R., & Mendoza-Romero, J. (1995). Hispanic masculinity: Myth or psychological schema meriting clinical consideration. In A. Padilla (Ed.), Hispanic psychology (pp.231–244). Newbury Park, CA: SAGE. Cashel, M. L., Rogers, R., Sewell, K., & Martin-Cannici, C. (1995). The Personality Assessment Inventory (PAI) and the detection of defensiveness. Assessment, 2(4), 333–342. Cha, E., Kim, K., & Erlen, J. (2007). Translation of scales in cross-cultural research: Issues and techniques. Journal of Advanced Nursing, 58(4), 386–395. Committee on Psychological Testing, Including Validity Testing, for Social Security Administration Disability Determinations; Board on the Health of Select Populations; and Institute of Medicine. (2015). Cognitive tests and performance validity tests. Washington, DC: National Academies Press. Available from www. ncbi.nlm.nih.gov/books/nbk305230. Correa, A. A. (2013). An investigation of malingering and defensiveness using the Spanish PAI among Spanish-speaking Hispanic American outpatients. Dissertation Abstracts International, 76,(1-B)(E). Correa, A. A., & Rogers, R. (2012). Entrevistas de simulación [Structured interviews for malingering]. In H. González Ordi, P. Santamaria Fernández, & P. Capilla Ramírez (Eds.), Estrategias de detección de la simulación: Un manual clínico multidisciplinar [Detection strategies for malingering: A multidisciplinary clinical manual] (pp.153–183). Madrid: TEA Ediciones. Correa, A. A., & Rogers, R. (2010a). Cross-cultural applications of the PAI. In M. Blais, M. Baity, & C. Hopwood (Eds.), Clinical applications of the Personality Assessment Inventory (pp.135–148). New York: Routledge/Taylor & Francis. Correa, A. A., & Rogers, R. (2010b). Validation of the Spanish SIRS with monolingual Hispanic outpatients. Journal of Personality Assessment, 92(5), 458–464.

78

I.  Concep t ua l Fr a me work

Cuellar, I., Arnold, B., & Maldonado, R. (1995). Acculturation Rating Scale for Mexican Americans–II: A revision of the original ARSMA Scale. Hispanic Journal of Behavioral Sciences, 17(3), 275–304. Cuellar, I., & Paniagua, F. A. (Eds.). (2000). Handbook of multicultural mental health: Assessment and treatment of diverse populations. San Diego, CA: Academic Press. Dana, R. H. (1993). Multicultural assessment perspectives for professional psychology. Boston: Allyn & Bacon. Dana, R. H. (1995). Culturally competent MMPI assessment of Hispanic populations. Hispanic Journal of Behavioral Sciences, 17(3), 305–319. Dana, R. H. (2000). Handbook of cross-cultural and multicultural personality assessment. Mahwah, NJ: Erlbaum. Dana, R. H. (2005). Multicultural assessment principles, applications, and examples. Mahwah, NJ: Erlbaum. Ewart, C. K., & Suchday, S. (2002). Discovering how urban poverty and violence affect health development and validation of a Neighborhood Stress Index. Health Psychology, 21(3), 254–262. Fantoni-Salvador, P., & Rogers, R. (1997). Spanish versions of the MMPI-2 and PAI: An investigation of the concurrent validity with Hispanic patients. Assessment, 4(1), 29–39. Fernandez, K., Boccaccini, M. T., & Noland, R. M. (2007). Professionally responsible test selection for Spanish-speaking clients: A four-step approach for identifying and selecting translated tests. Professional Psychology: Research and Practice, 38(4), 363–374. Fernandez, K., Boccaccini, M. T., & Noland, R. M. (2008). Detecting over- and underreporting of psychopathology with the Spanish-language Personality Assessment Inventory: Findings from a simulation study with bilingual speakers. Psychological Assessment, 20(2), 189–194. Fragoso, J. M., & Kashubeck, S. (2000). Machismo, gender role conflict, and mental health in Mexican American men. Psychology of Men and Masculinity, 1(2), 87–97. Gibbons, J. L., Zellner, J. A., & Rudek, D. J., (1999). Effects of language and meaningfulness on the use of extreme response style by Spanish-English bilinguals. Cross-Cultural Research, 33(4), 369–381. Giger, P., & Merten, T. (2013). Swiss population-based reference data for six symptom validity tests. Clínica y Salud, 24(3), 153–159. González Ordi, H., Santamaria Fernández, P., & Capilla Ramírez, P. (Eds.). (2012). Estrategias de detección de la simulación: Un manual clínico multidisciplinar. [Detection strategies for malingering: A multidisciplinary clinical manual.] Madrid: TEA Ediciones. Gordon, M. M. (1964). Assimilation in American life: The role of race, religion, and national origins. New York: Oxford University Press. Graham, J. R. (1990). MMPI-2: Assessing personality and psychopathology (2nd ed.). New York: Oxford University Press.

Green, P. (2003). Green’s Word Memory Test for Microsoft Windows: User’s manual. Edmonton, Alberta, Canada: Green’s. Guttfreund, D. G. (1990). Effects of language usage on the emotional experience of Spanish–English and English–Spanish bilinguals. Journal of Consulting and Clinical Psychology, 58(5), 604–607. Hall, V. L., Worthington, A., & Venables, K. (2014). A UK pilot study: The specificity of the Word Memory Test effort sub-tests in acute minimal to mild head injury. Journal of Neuropsychology, 8(2), 216–230. Hambleton, R. K. (2001). The next generation of the ITC test translation and adaptation guidelines. European Journal of Psychological Assessment, 17(3), 164–172. Hambleton, R. K., & Kanjee, A. (1995). Increasing the validity of cross-cultural assessments: Use of improved methods for test adaptations. European Journal of Psychological Assessment, 11(3), 147–157. Heaton, R. K., Miller, S. W., Taylor, M. J., & Grant, I. (2004). Revised comprehensive norms for an Expanded Halstead–Reitan Battery: Demographically adjusted neuropsychological norms for African Americans and Caucasian adults. Lutz, FL: Psychological Assessment Resources. Heaton, R. K., Taylor, M., & Manly, J. (2001). Demographic effects and demographically corrected norms with the WAIS-III and WMS-III. In D. Tulsky, R. K. Heaton, G. J. Chelune, I. Ivnik, R. A. Bornstein, A. Prifitera, & M. Ledbetter (Eds.), Clinical interpretations of the WAIS-II and WMS-III (pp.181–210). San Diego, CA: Academic Press. Henrich, J., Heine, S. J., & Norenzayan, A. (2010). The weirdest people in the world? Behavioral and Brain Sciences, 33(2–3), 61–83. Hill, M. E. (2002). Skin color and intelligence in African Americans: A reanalysis of Lynn’s data. Population and Environment, 24(2), 209–214. Hopwood, C. J., Flato, C. G., Ambwani, S., Garland, B. H., & Morey, L. C. (2009). A comparison of Latino and Anglo socially desirable responding. Journal of Clinical Psychology, 65(7), 769–780. Hui, C., & Triandis, H. (1989). Effects of culture and response format on extreme response style. Journal of Cross-Cultural Psychology, 20(3), 296–309. Jeanrie, C., & Bertrand, R. (1999). Translating tests with the International Test Commission’s guidelines: Keeping validity in mind. European Journal of Psychological Assessment, 15(3), 277–283. Jones, P. S., Lee, J. W., Pillips, L. R., Zhang, X. E., & Jaceldo, K. B. (2001). An adaptation of Brislin’s translation model for cross-cultural research. Nursing Research, 50(5), 300–304. Kaufman, A. S., & Lichtenberger, E. O. (2006). Assessing adolescent and adult intelligence (3rd ed.). Hoboken, NJ: Wiley. Liu, C., Liu, Z., Chiu, H. F. K., Carl, T. W., Zhang, H., Wang, P., et al. (2013). Detection of malingering: Psychometric evaluation of the Chinese version of

4.  Cultural and Transnational Perspectives 79

the Structured Interview of Reported Symptoms-2. BMC Psychiatry, 13(1), 254–254. Lynn, R. (2002). Skin color and intelligence in African Americans. Population and Environment, 23(4), 365–375. Malcarne, V., Chavira, D., Fernandez, S., & Liu, P. (2006). The scale of ethnic experience: Development and psychometric properties. Journal of Personality Assessment, 86(2), 150–161. Manly, J. J. (2005). Advantages and disadvantages of separate norms for African Americans. Clinical Neuropsychologist, 19, 270–275. Manly, J. J., & Echemendia, R. J. (2007). Race-specific norms: Using the model of hypertension to understand issues of race, culture, and education in neuropsychology. Archives of Clinical Neuropsychology, 22(3), 319–325. Marcella, A. J., & Yamada, A. M. (2000). Culture and mental health: An introduction and overview of foundations, concepts, and issues. In I. Cuellar & F. A. Paniagua (Eds.), Handbook of mental health (pp.3–24). San Diego, CA: Academic Press. Marín, G., & Marín, B. V. (1991). Research with Hispanic populations. Newbury Park, CA: SAGE. Marín, G., Perez-Stable, E. J., & Marín, B. V. (1989). Cigarette smoking among San Francisco Hispanics: The role of acculturation and gender. American Journal of Public Health, 79, 196–198. Martinez, G., Marín, B. V., & Schoua-Glusberg, A. (2006). Translating from English to Spanish: The 2002 National Survey of Family Growth. Hispanic Journal of Behavioral Sciences, 28(4), 531–545. McCarter, R. J., Walton, N. H., Brooks, D. N., & Powell, G. E. (2009). Effort testing in contemporary UK neuropsychological practice. Clinical Neuropsychologist, 23(6), 1050–1067. Merckelbach, H., & Smith, G. P. (2003). Diagnostic accuracy of the Structured Inventory of Malingered Symptomatology (SIMS) in detecting instructed malingering. Archives of Clinical Neuropsychology, 18(2), 145–152. Merten, T., Dandachi-FitzGerald, B., Hall, V., Schmand, B. A., Santamaría, P., & González Ordi, H. (2013). Symptom validity assessment in European countries: Development and state of the art. Clínica y Salud, 24(3), 129–138. Miller, H. A. (2001). M-FAST: Miller Forensic Assessment of Symptoms Test professional manual. Odessa, FL: Psychological Assessment Resources. Mindt, M. R., Byrd, D., Saez, P., & Manly, J. (2010). Increasing culturally competent neuropsychological services for ethnic minority populations: A call to action. Clinical Neuropsychologist, 24(3), 429–453. Molina, R. A., & Franco, J. N. (1986). Effects of administrator and participant sex and ethnicity on selfdisclosure. Journal of Counseling and Development, 65(3), 160–162. Montes, O., & Guyton, M. R. (2014). Performance of Hispanic inmates on the Spanish Miller Forensic

Assessment of Symptoms Test (M-FAST). Law and Human Behavior, 38(5), 428–438. Moreland, K. L. (1996). Persistent issues in multicultural assessment of social and emotional functioning. In L. A. Suzuki, P. J. Mueller, & J. G. Ponterotto (Eds.), Handbook of multicultural assessment (pp.51–76). San Francisco: Jossey-Bass. Morey, L. C. (1991). The Personality Assessment Inventory. Odessa, FL: Psychological Assessment Resources. Morey, L. C. (2007). The Personality Assessment Inventory (2nd ed.). Lutz, FL: Psychological Assessment Resources. Nasreddine, Z. S., Phillips, N. A., Bédirian, V., Charbonneau, S., Whitehead, V., Collin, I., et al. (2005). The Montreal Cognitive Assessment, MoCA: A brief screening tool for mild cognitive impairment. Journal of the American Geriatrics Society, 53(4), 695–699. Okazaki, S., & Sue, S. (1995). Methodological issues in assessment research with ethnic minorities. Psychological Assessment, 7(3), 367–375. Ortiz-Tallo, M., Santamaria, P., Cardenal, V., & Sanchez, M. P. (2011). Inventario de evaluación de la personalidad PAI adaptación Española. [Personality Assessment Inventory PAI Spanish Adaptation]. Madrid: TEA Ediciones. Ostrosky-Solís, F., Gómez-Pérez, M. E., Matute, E., Rosselli, M., Ardila, A., & Pineda, D. (2007). Neuropsi attention and memory: A neuropsychological test battery in Spanish with norms by age and educational level. Applied Neuropsychology, 14(3), 156–170. Reid, W. H. (2000). Malingering. Journal of Psychiatric Practice, 6, 226–228. Resnick, P. J. (1984). The detection of malingered mental illness. Behavioral Sciences and the Law, 2(1), 21–38. Rienstra, A., Spaan, P. E. J., & Schmand, B. (2009). Reference data for the Word Memory Test. Archives of Clinical Neuropsychology, 24(3), 255–262. Rogers, R. (Ed.). (1997). Clinical assessment of malingering and deception (2nd ed.). New York: Guilford Press. Rogers, R. (2008). Clinical assessment of malingering and deception (3rd ed.). New York: Guilford Press. Rogers, R., Bagby, R. M., & Dickens, S. E. (1992). SIRS: Structured Interview of Reported Symptoms professional manual. Odessa, FL: Psychological Assessment Resources. Rogers, R., Flores, J., Ustad, K., & Sewell, K. W. (1995). Initial validation of the Personality Assessment Inventory–Spanish Version with clients from Mexican–American communities. Journal of Personality Assessment, 64(2), 340–348. Rogers, R., Hinds, J. D., & Sewell, K. W. (1996). Feigning psychopathology among adolescent offenders: Validation of the SIRS, MMPI-A, and SIMS. Journal of Personality Assessment, 67(2), 244–257. Rogers, R., Sewell, K. W., & Gillard, N. D. (2010). SIR2: Structured Interview of Reported Symptoms: Professional manual (2nd ed.). Lutz, FL: Psychological Assessment Resources. Rogers, R., & Shuman, D. W. (2005). Fundamentals of

80

I .  C o n c e p t u a l F r a m e w o r k

forensic practice: Mental health and criminal law. New York: Springer. Romain, P. M. (2000). Use of the Personality Assessment Inventory with an ethnically diverse sample of psychiatric outpatients. Dissertation Abstracts International, 61(11-B), 6147. Ruiz Sanchez de Leon, J. M., & Gonzalez Marquez, J. (2012). Evaluación Neuropsicológica [Neuropsychological evaluation]. In H. González Ordi, P. Santamaria Fernández, & P. Capilla Ramírez, P. (Eds.), Estrategias de detección de la simulación: Un manual clínico multidisciplinar [Detection strategies for malingering: A multidisciplinary clinical manual.] (pp.275–300). Madrid: TEA Ediciones. Ryder, A. G., Alden, L. E., & Paulhus, D. L. (2000). Is acculturation unidimensional or bidimensional?: A head-to-head comparison in the prediction of personality, self-identity, and adjustment. Journal of Personality and Social Psychology, 79(1), 49–65. Salvador-Carulla, L. L., Garrido, M. M., McDaid, D. D., & Haro, J. M. (2006). Financing mental health care in Spain: Context and critical issues. European Journal of Psychiatry, 20(1), 29–44. Sireci, S. G., Yang, Y., Harter, J., & Ehrlich, E. J. (2006). Evaluating guidelines for test adaptations: A methodological analysis of translation quality. Journal of Cross-Cultural Psychology, 37(5), 557–567.

Smith, G. P. (1997). Assessment of malingering with self-report instruments. In R. Rogers (Ed.), Clinical assessment of malingering and deception (pp.351–370). New York: Guilford Press. Smith, G. P., & Burger, G. K. (1997) Detection of malingering: Validation of the Structured Inventory of Malingered Symptomatology (SIMS). Journal of the Academy of Psychiatry and the Law, 25(2), 180–183. Todd, W. (2004). Race/ethnicity and the Personality Assessment Inventory (PAI): The impact of culture on diagnostic testing in a college counseling center. Dissertation Abstracts International, 65(10-B), 5425. Tombaugh, T. N. (1996). Test of Memory Malingering: TOMM. Toronto: Multi-Health Systems. Vilar López, R., Pérez García, M., & Puente, A. E. (2011). Adaptación española del TOMM, Test de simulación de problemas de memoria. [Spanish adaptation of the TOMM, Test of Memory Malingering]. Madrid: TEA Ediciones. Wagner, J., & Gartner, C. G. (1997). Highlights of the 1996 Institute on Psychiatric Services. Psychiatric Services, 48(1), 51–56. Zax, M., & Takahashi, S. (1967). Cultural influences on response style: Comparisons of Japanese and American college students. Journal of Social Psychology, 71(1), 3–10.

PA R T I I

DIAGNOSTIC ISSUES

CHAPTER 5

Syndromes Associated withDeception MichaelJ.Vitacco,PhD

Deception is a central component of malingering in the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5; American Psychiatric Association, 2013). Forensic evaluators must be aware of the possibility that examinees may engage in deception whenever they are undergoing court-ordered evaluations. Malingering is described by DSM-5 in terms of faked presentations and external motivations. Yet in traditional psychotherapeutic settings, honesty is often taken for granted given that most clients are self-referred for treatment and express a genuine desire and interest in their personal growth. As such, the base rate of deceptive behaviors in traditional treatment settings remains unknown given the dearth of programmatic research on their detection. In treatment settings, response styles are often considered inconsequential as the goal is designed to improve client functioning. However, in adversarial situations, such as court-ordered evaluations, examinees frequently engage in a variety of response styles. In these cases, deception should not be construed as inconsequential and should be evaluated as a standard part of the assessment process. Such a consideration is unsurprising, since base rates of symptom exaggeration in forensic evaluations range from 15 to 30% (Mittenberg, Patton, Canyock, & Condit, 2002; Vitacco, Rogers, Gabel, & Munizza, 2007). This chapter focuses on various diagnoses, syndromes, and situations in which deception is para 83

mount. As discussed by Vitacco and Rogers (2005), individuals have varyious motives for deception at different times. For instance, the same person may feign psychosis in a pretrial competency-to-proceed evaluation to delay going to trial, and after conviction feign psychosis in a correctional environment for a different reason (e.g., a preferential cell placement or unwarranted medications). Finally, the same individual may become defensive and deny mental health problems as a potential parole release date draws closer. This illustrative example exemplifies several features about deception that should become evident throughout this chapter. First, deception is a multidimensional construct that manifests differently across situations and settings. Second, relatedly, deception is not taxonic; rather, it should be viewed as a dimensional construct that can change in direction and intensity. Third, deception is frequently adaptive but not always. For instance, in the earlier example, trying to obtain medications through feigning may be both adaptive and criminogenic (see Rogers, Chapter 1, this volume). Finally, individuals who engage in deceptive behaviors are not always insightful and aware of their reasons for deceiving. For example, clients with anorexia nervosa may deceive others about the calories consumed or their robustness of their health, but these same clients may still lack insight into the pervasiveness of their illnesses and a strong desire to control their body images. In the context of these syndromes, this chapter examines

84

II .   D i a g n o s t ic I ss u e s

specific contexts within each syndrome in which deception is manifested. As the primary goal, this chapter addresses a variety of syndromes (and situations) associated with deceptions. The specific areas include DSM-5 disruptive and impulse control disorders, specifically, conduct and oppositional defiant disorders. Second, this chapter provides information on the relationship between deception and personality disorders, specifically, antisocial personality disorder (ASPD) and psychopathy. Third, it considers drug use and abuse, which have strong associations to deceptive behavior. Fourth, as mentioned earlier, individuals with eating disorders frequently use deception to avoid questions about their lack of caloric intake or to hide binging. Finally, individuals engage in a variety of deceptive practices related to sexual paraphilia. Again, the actual reasons for deception in sexual disorders are varied. Finally, I discuss other syndromes (e.g., false memory) and conditions (e.g., child custody) in relation to deception.

CHALLENGES ANDPITFALLS WHEN EVALUATING MALINGERING ANDDECEPTION A classification of malingering by itself is often the foundation for serious consequences. For example, the U.S. Code of Military Justice Article 115 provides criminal sanctions for a soldier who malingers for the purpose of avoiding work or duty. In the landmark case of the United States v. Greer (1998), the Fifth Circuit Court of Appeals allowed for enhanced punishment for malingering under the umbrella of obstruction of justice. Problematically, relying on DSM-5 screening indicators for malingering represents a surefire way to have an unacceptable number of false positives. As noted by Rogers and Vitacco (2002), the rate of false positives could be as high as 80% if relying solely on these indicators. These indicators include background (ASPD), setting (forensic), discrepancy (subjective reports unverified by objective data), and assessment attitude (uncooperative). In routinely evaluating pretrial defendants for court-ordered evaluations, most of the examinees I assess meet many, if not all, of the DSM-5 screening indicators for malingering. Yet the majority of these defendants do not appear to be engaging in intentional deception, such as malingering. Despite these enduring problems (e.g., see Rogers, 1990a, 1990b), DSM-5 continues to rely on screening indicators that are outdated and poorly validated.

In evaluating feigning and related response styles, several conceptual issues warrant consideration. First, evaluators must carefully consider examinees’ motivations. The definition of malingering requires a close review. It necessitates that feigning must be “intentional.” Factitious disorders, such as malingering, are typically viewed as occurring on a continuum (see Yates, Mulla, Hamilton, & Feldman, Chapter 11, this volume). Clinicians must consider motivations when making a diagnosis of malingering. As noted both in the Structured Interview of Reported Symptoms (SIRS; Rogers, Bagby, & Dickens, 1992) and the Structured Interview of Reported Symptoms–2 (SIRS-2; Rogers, Sewell, & Gillard, 2010)—the premiere instruments for the detection of feigned mental disorders—clinicians are asked to carefully evaluate the individual’s motivation for deceptive responding. As reported in the SIRS manuals, symptoms of malingering and factitious disorder are difficult to disentangle. A second point is that clinicians must not equate isolated test results with a classification of feigning or malingering. Psychological testing is useful, if not essential, to properly assessing malingering and deception. However, a single score on a psychological test must never be considered in a vacuum. False positives are an inherent part of any diagnostic test, be it psychological or medical, and psychological measures for evaluating response styles are not immune. Underscoring this idea, Berthelson, Mulchan, Odland, Miller, and Mittenberg (2013) reported a false-positive rate for malingering in neuropsychological evaluations to be as high as 38% (see also Larrabee, 2012). And, although superior to clinical judgment, objective measures, when used as the sole indices of malingering, continue to put clinicians at continued and serious risk for false conclusions. In general, clinicians are on empirically solid ground when using a multimodal approach to evaluating malingering (Denney, 2005). Such an approach integrates information from behavioral observations, mental health history (if any), criminal history (if any), and objective psychological testing. For instance, it would be highly unlikely for defendants with lengthy criminal histories and court proceedings to accurately claim virtually no knowledge about court or how the criminal justice system works. Historical information, when combined with careful objective testing of response styles, provides clinicians with well-rounded and comprehensive information on response styles. As a third conceptual issue, as noted earlier, the behaviors associated with malingering are not tax-

5.  Syndromes Associated with Deception 85

onic (Walters, Berry, Rogers, Payne, & Granacher, 2009; Walters et al., 2008). For instance, in a confirmatory factor analysis of the SIRS, Rogers, Jackson, Sewell, and Salekin (2005) found that two dimensional factors underpin the instrument: spurious presentation and plausible presentation. Using a similar analysis with the Miller Forensic Assessment of Symptoms Test (M-FAST; Miller, 2001), Vitacco et al. (2008) found a single factor that best represented the M-FAST structure. Beyond formal analyses, Resnick, West, and Payne (2008) subtyped malingering by the level of fabrication/exaggeration, from pure malingering (i.e., total fabrication) to false imputation (no exaggeration but misattributing symptoms to a compensable cause; see Resnick, West, & Wooley, Chapter 10, this volume). As a fourth conceptual issue, clinicians should consider explanatory models of malingering (see Rogers, Chapter 1, this volume). Table 5.1 offers a basic conceptualization of various clinical syndromes and situations in which deception is com-

mon. However, a close review of Table 5.1 demonstrates that explanatory models are not discrete categories; behaviors associated with deception can span two or even three of the explanatory models. It is clear that most types of deception are actually unrelated to malingering. What is clear is that clinicians must be aware of multiple determinants of motivations related of feigning and deceptive behaviors. For the fifth and final point, clinicians, inasmuch as possible, must be exacting in their use of language. For example, the term secondary gain was popular for many years and used by clinicians when conducting forensic evaluations. However, Rogers and Reinhardt (1998) warned against using the term secondary gain in forensic situations. By default, practically everyone undergoing a forensic evaluation has a stake in the outcome. If a clinician uses secondary gain as an indicator, it is akin to saying, “I know it is there, I just need to find it.” Based on such ambiguities and potential for misuse, Vitacco and Rogers (2005) recommended cli-

TABLE 5.1.  Explanatory Models of Motivations Related to Diagnoses and Dissimulation

Disorder

Explanatory model

Characteristics associated with deception

Conduct disorders

Criminogenic

Instrumental/poor impulse control

Reactive attachment disorder

Pathogenic

Secondary to extreme abuse and abandonment

Adaptational

Compensatory mechanisms in social situations

Criminogenic

Secondary to antisocial personality disorders

Adaptational

Financial motivations

Pathogenic

Rigidity

Criminogenic

Secondary to antisocial personality disorders

Adaptational

Escaping/avoiding responsibilities

Pathogenic

Self-medicating

Eating disorders

Pathogenic

Rigidity/distorted body image/maintaining control

Paraphilias

Criminogenic

Luring victims/maintaining offending

Pathogenic

Own abuse history leads to poor boundaries

Psychopathy

Criminogenic

Instrumental/game playing/poor impulse control

False-memory syndrome

Criminogenic

Secondary to antisocial personality disorder

Adaptational

Financial motivations (litigation)

Pathogenic

Regression/avoiding responsibilities

Criminogenic

Extortion/lying to turn child against parent

Adaptational

Strong desire to remain with child

Pathogenic

Rigidity/personality disorders

Factitious disorder

Substance abuse

Child custody

86

II .   D i a g n o s t ic I ss u e s

nicians abandon the term secondary gain in favor of more empirically validated terms. In summary, evaluating malingering and deception are complicated endeavors that require a sophisticated understanding of response styles, knowledge of strengths and limitations of objective measures, and the willingness to explore motivations as they relate to deception. Drob, Meehan, and Waxman (2009) indicated that there is pressure on clinicians to find malingering, and cultural factors complicate the detection of malingering. These challenges underscore the importance of developing a full appreciation of response styles prior to conducting forensic evaluations. In the next two sections I focus on areas in which deception is common. The initial section focuses on deception as it relates to formal DSM-5 diagnoses. The second addresses other clinical syndromes and conditions in which various aspects of deception are manifested.

DSM-5 DIAGNOSES ASSOCIATED WITHDECEPTION Clinicians in general, but especially forensic clinicians, must be aware of the multiple ways deception can be manifested as a part of psychopathology. Moreover, as will be demonstrated, clinicians must be cognizant that motivations for deceptions may vary. Clinicians may prescribe malevolent motives to deceptive behavior, when, in reality, the deception may be unconscious or simply adaptive. This section begins with disorders typically diagnosed in children and adolescents, then moves to disorders of adulthood in considering a full range of explanations for deceptive behaviors within the context of DSM diagnoses.

Deception andDisorders ofChildhood andAdolescence Oppositional Defiant andConductDisorders

Oppositional defiant disorder (ODD) and conduct disorder (CD) are now found in the section of DSM-5 labeled “Disruptive, Impulse-Control, and Conduct Disorders.” These disorders have been grouped together because they involve problems with self-control that may bring them into conflict with others. Sometimes these symptoms portray behavioral problems that continue to manifest into adulthood (Moffitt, 1993). Behaviors related to deceitfulness are a core component of both CD and ODD. For example, in CD, diagnostic criteria

explicitly place deceitfulness as part of the disorder with respect to falsehoods and conning. New DSM-5 criteria also allow for a specifier regarding whether behaviors include limited prosocial emotions, including lack of remorse and the presence of callousness. As such, for a CD diagnosis, it would be important to discern how adolescents think and feel about their deceptive behavior. In contrast to the explicit CD criteria related to deception, there are no explicit criteria linking ODD with deception. Indirectly, the deception in ODD is often manifested in poor attitudes displayed toward people in authority. Frick et al. (1991) evaluated the covariation of symptoms between ODD and CD in a sample of 177 clinicreferred boys. The authors found that lying loaded significantly on both ODD and CD across parent and teacher ratings. Although ODD and CD differ in their deceptive presentations, the deception associated with both diagnoses typically has a negative influence on the adolescent’s interpersonal relationships (Hughes & Hill, 2006). Relatedly, deception can also be part of behavior associated with psychopathic traits in adolescence. In studies evaluating traits of psychopathy in youth, deception was frequently part of the construct (Lynam et al., 2009; Lynam & Gudonis, 2005). In fact, two items on the Psychopathy Checklist: Youth Version (PCL:YV; Forth, Kosson, & Hare, 2003) are Pathological Lying and Conning. The mere act of engaging in deception portends significant problems in multiple domains of adolescents’ lives. Boys perceived as deceptive by their peers were reported to be more withdrawn, more aggressive, and less likeable (Gervais, Tremblay, & Héroux, 1998). A final consideration in this section involves how deceptions relate to sexual offending behavior with adolescents. Like adults, deception can take the form of both lying and minimization to avoid taking responsibility or to continue to allow access to victims (Barbaree, Marshall, & Hudson, 1993). Evidence supports the notion the large majority of adolescent sexual offenders do not persist into adulthood. Research (Caldwell, 2002; Caldwell & Dickinson, 2009) found sexually-based offending is not the strongest predictor of adult sex offenses. Nonetheless, deception and lying in adolescent sex offenders can be secondary to several factors including embarrassment, trying to get away with it, or minimizing their own histories of victimization. For instance, Baker, Tabacoff, Tornusciolo, and Eisenstadt (2001) compared sexually abused children with other children on welfare. Notably,

5.  Syndromes Associated with Deception 87

an identified factor labeled “family deception” was composed of family myths and active lying. Longitudinally, the presence of “family deception” was associated with subsequent sexual offending. In summary, this section highlights the relationship between externalizing disorders and deceptive behaviors. In ODD and CD, deception is likely voluntary and related to either general defiance or blatant antisocial behavior. Specifically, deception is embedded within the items on the Interpersonal facet of the PCL:YV. Clinicians working with children and adolescents with conduct problems are advised not only to assess for the presence of deception but also to consider deception as a viable treatment target. Reactive AttachmentDisorder

Similar to ODD and CD, children with reactive attachment disorder (RAD) are known to engage in deception. RAD is classified in DSM-5 under “Trauma and Stressor-Related Disorders”; children with this disorder often have experienced chronic neglect or abuse, leading to a disturbance in the ability to properly bond with individuals. The core feature for RAD involves emotional withdrawal and a lack of appropriate attachment. Research has linked RAD with lying and deception. Wilson (2001) characterized children with RAD as presenting with “sociopathic behavior which includes deception” (p.42). Multiple studies have identified associations between psychopathic traits and early neglect. For example, Schraft, Kosson, and McBride (2013) found a link between exposure to violence and the presence of psychopathic traits in a sample of adolescents. Schimmenti, Di Carlo, Passanisi, and Caretti (2015) found that emotional abuse and neglect were related to psychopathy scores in a sample of inmates with violent convictions. More broadly, Gostisha et al. (2014) reported a modest relationship between early life stress and psychopathic traits in a sample of incarcerated adolescents in a secure treatment facility (see also Shirtcliff et al., 2009). Although the deceptive behaviors of RAD may appear similar to those of ODD and CD, they are fundamentally different in terms of the underlying motivation. In RAD, the deception tends to be used as an adaptive mechanism to protect the individual from what he or she perceives as dangerous social relationships. Alternatively, the deception may be pathogenic, secondary to the extreme neglect by caregivers. Clinicians are advised to explore histories of abuse and/or neglect

when evaluating behavioral problems. It is possible that what appear to be conscious attempts at deception may instead be behaviors associated with RAD, which are secondary to abuse and neglect. As evidenced by the new DSM-5 criteria, the etiology and pathogenesis of RAD remain imprecise (Zilberstein, 2006); however, research points to an association between the symptoms of RAD and various forms of disruptive behaviors, including deception.

Deception andDisorders ofAdulthood FactitiousDisorders

Factitious disorders are frequently confused with malingering due to their similar clinical presentations. In DSM-5, factitious disorders are found in the category “Somatic and Related Disorders.” Like malingering, a key component of factitious disorder is deception. Notably, the individual often presents to others as highly impaired or seriously ill. As noted by both the SIRS and SIRS-2 (Rogers et al., 1992, 2010), factitious disorder must be ruled out before deciding whether an individual is malingering. The key differential with factitious disorders is the absence of clear external incentive for the deceptive behavior. Consistent with malingering, the feigning of symptoms is within the patient’s control. However, no explicit guidelines on psychological measures clearly differentiate between malingering and a factitious disorder. In these cases, clinicians must rely on their professional judgment and explore the need of the individual to continue to maintain his or her identity as a sick person. The lack of objective evaluation weakens both the validity of the factitious disorder and clinicians’ ability to render an accurate diagnosis. In considering factitious disorder, several factors of the patient’s presentation likely draw attention to him or her. DSM-IV-TR (American Psychiatric Association, 2000) and DSM-5 (American Psychiatric Association, 2013) list criteria required for diagnosing a factitious disorder. These include both behavioral/clinical (i.e., atypical or dramatic presentation inconsistent with a mental disorder or medical condition, pathological lying, covert use of substances, and extensive history of traveling) and treatment-related issues (i.e., symptoms that are present only when the individual is aware that he or she is being observed; disruptive behavior on an inpatient unit; arguing excessively with nurses and physicians; evidence of multiple treatment interventions; few, if any, hospital visi-

88

II .   D i a g n o s t ic I ss u e s

tors; and fluctuating hospital course). In a clinical case, I evaluated an offender for serious depression and suicidal ideations while he was serving a sentence for multiple charges of burglary. Objective psychological testing (e.g., Personality Assessment Inventory; Morey, 1991) and a specialized measure (SIRS) were consistent with feigned mental disorders. Despite feigning results, the multidisciplinary treatment team found no external incentives present. A review of the inmate’s history indicated several hospitalizations for unspecified physical and emotional disorders. Ultimately, the treatment team decided the most appropriate diagnosis was factitious disorder. This case illustrates that even in prison, where clinicians generally consider malingering first, it is prudent to consider the full range of potential motivations for deception. Second, this case is consistent with the notion that once it is determined that an individual is feigning, the next step needs to determine the motivation for the feigning (see Rogers et al., 1992). Another consideration is that factitious disorder includes various and diverse symptoms. Cunnien (1997) proposed the following as markers of factitious disorders: (1) strong masoch*stic needs; (2) sickness allowing regression and avoidance of adult responsibilities; (3) illness symbolic of anger or conflict with authority figures; (4) illness fulfilling dependency needs; and (5) illness symbolizing attempts at mastery of past trauma. In going beyond motivations, Cunnien provided a list of diagnoses often exhibited by individuals with factitious disorders. These include psychosis, bereavement, posttraumatic stress disorder, dissociative identity disorder, and even false claims of child abuse. Pope, Jonas, and Jones (1982) evaluated nine patients who seemingly had control over the presence of their psychotic symptoms and each was diagnosed with personality disorder(s). Such comorbidity is seemingly common with factitious disorders. The development and presentation of factitious disorders are not well understood. This is unfortunate given that factitious disorders are likely more prevalent than initially believed (Gregory & Jindal, 2006), with prevalence estimated at 1% in inpatient settings (American Psychiatric Association, 2013, p.326). Factitious Disorder Imposed byAnother (Munchausen byProxy)

In DSM-5, factitious disorder imposed by another (FDIA) is the new term for deceptions targeting another person either in symptom falsification or

surreptitious symptom induction, while typically pretending to be a concerned caregiver. This diagnosis was previously referred to as Munchausen by proxy. Often FDIA takes the form of a parent purposely making his or her child become sick or impaired. If discovered, it can lead to legal charges levied against the parent. If undiscovered, the child can be subjected to long-term disabilities or even death. The television news show 48 Hours presented a case in April 2016 of a young mother accused of injecting her son with salt in order to make him sick. Ultimately, the child died and the mother was convicted of second-degree murder. Unfortunately, FDIA is not rare, and it has been suggested that factitious disorder imposed by another constitutes 10% of factitious disorder cases (Reich & Gottfried, 1983). Diagnosing FDIA is extremely difficult due to the level of deception a parent or caretaker undertakes to convince others that someone is very sick. Waller (1983) discussed how significant deception and convincing authorities that a caretaker is making the child sick interact to form barriers to effective identification and treatment. Waller noted that most parents continue their strong denials even when confronted with overwhelming evidence implicating them as the cause of their child’s illness. I once was involved in a case in which a mother went so far as to shave her child’s head to convince others she was undergoing chemotherapy to counteract an aggressive leukemia. The mother was caught only when relatives came to the hospital after traveling a significant distance, due to their concerns about the child’s impending death. Libow and Schreier (1986) reported three subtypes of individuals with FDIA: (1) those calling for help, (2) active inducers, and, (3) doctor addicts. Active inducers garner the most attention, because the parent’s behavior in these cases is criminal and the children under their care are at risk of harm. For instance, in the case presented on 48 Hours, the mother was found guilty of causing the death of her son, which occurred while he was in the hospital, under close medical supervision. As noted by John Stirling, Jr., and the Committee on Child Abuse and Neglect (2007), FDIA is not just a mental health disorder, it also constitutes a serious form of child abuse (see Ayoub et al., 2002). The etiology of factitious disorder and FDIAs are not well understood (Mart, 2002). One potential way to evaluate such disorders is through the application of explanatory models of malingering proposed by Rogers (1990b). To that end,

5.  Syndromes Associated with Deception 89

underlying motivations for factitious disorders, according to Cunnien (1997), may be consistent with all three explanatory models of malingering (i.e., adaptational, pathogenic, and criminogenic). Adaptational explanations include financial motivations or an attempt to bring alienated family together to bond over a common cause (i.e., a sick child). Pathogenic explanations include rigid control of family or a dysfunctional and maladaptive attachment with the child (Rogers, 1990b). Finally, the criminogenic model might be used to explain how an individual with psychopathy or ASPD would attempt to generate an illness to manipulate or steal. Despite their potential usefulness, Rogers (2004) indicated that such explanatory models for factitious disorders are underdeveloped. As such, our understanding of these disorders and how to apply explanatory models also need to be fully addressed. EatingDisorders

Patients with diagnoses of anorexia nervosa and bulimia nervosa often engage in a variety of deceptive practices that enable them to continue their obsessive behavior focused on their weight and body image. This behavior is especially dangerous, because it is designed to provide the façade the client is improving when, in fact, the eating disorder is continuing. Specific deceptive behaviors associated with eating disorders are wide ranging and include the following: • Hiding food in order to engage in bingeing behavior. • Secretly exercising, even when under close observation. • Stealing laxatives and diet supplements to avoid the embarrassment of purchasing them. • Traveling long distances to use drugstores outside of one’s typical neighborhood to avoid family and friends while purchasing diet aids. • Lying about weight gain and minimizing weight loss. • Wearing clothes to hide weight loss or trying to provide the appearance of gaining weight. In DSM-5, deception is not listed as part of the formal diagnostic criteria, because the description is focused mostly focused on body weight and selfimage. However, studies of individuals with eating disorders consistently revealed deceptive behaviors. Lacey and Evans (1986; also see Lacey, 1993) evaluated 112 individuals with bulimia and diag-

nosed with other impulse control disorders. In this study, 21% of the patients repeatedly stole. These results confirm how deception is frequently manifested in individuals with eating disorders. Family deceptiveness has been identified as a factor in development of eating disorders (Dalzell, 2000). Notably, this factor is frequently relevant in adolescents engaging in illegal sexual behavior. Deceptive family environments, potentially related to overcontrolled and rigid parenting styles, are implicated as an etiological factor for eating disorders. Such families have a history of nondisclosure and frequently deny basic flaws in their families and in each other. These families have the desire to project an outward appearance of perfection. Not surprisingly, individuals with eating disorders often employ deception to maintain control of their eating behavior and to deceive parents and treatment providers into believing that treatment is working. From an explanatory model, the pathogenic model appears to be apt in the attempt to understand deceptive behaviors associated with eating disorders. Substance Abuse andOther AddictiveBehaviors

Deception in the context of substance abuse and addictive behaviors is commonplace. The use of alcohol and illicit drugs has long been associated with denial and misrepresentation. Yet professional athletes and sports participants are now also frequently engaging in elaborate schemes to avoid detection of banned substances. Lance Armstrong, the then seven-time champion of the Tour de France, admitted to Oprah Winfrey that he utilized a concoction of banned substances during his victory streak. Also revealed was the extensive system he utilized to defeat drug tests and sophisticated detection systems. Baseball has also been implicated in the performance-enhancing drug culture. The years between 1994 and 2003 are referred to as the “steroid area,” in which there was a dramatic increase in the number of home runs hit. With increased attention and implementation of more stringent drug testing, home run production dropped precipitously (Erickson, Yanke, Monson, & Romeo, 2015). Professional leagues now severely sanction players with positive drug tests, but new technologies designed to counteract drug tests seem to always be in development. In considering the criteria of a substance abuse disorders and deception, DSM-5 (American Psychiatric Association, 2013) indicates that an “in-

90

II .   D i a g n o s t ic I ss u e s

dividual may spend a great deal of time obtaining the substance, using the substance, or recovering from its effects. In some instances of more severe substance abuse disorders, virtually all of the individual’s daily activities revolve around the substance” (p.483). In order to continue to use substances, most individuals must justify their time away from work and home. Individuals abusing prescribed medications often need to fabricate new ailments or deceive new physicians in order to maintain their access to desired medications. One way for substance abusers to gain a more accurate perception of their behavior is through a 12-step program. Ferrari, Groh, Rulka, Jason, and Davis (2008) reported that participation in formal 12-step programs serves to minimize denial and minimization. Not surprisingly, extant research has identified a relationship between substance abuse and deception. Klein (2000) found a significant overlap between scores on the SIRS and substance abuse identified by the Substance Abuse Subtle Screening Inventory (SASSI; Miller, 1994a, 1994b). College students who self-report more substance abuse also report cheating on examination and having lied to avoid taking an examination (Blankenship & Whitley, 2000). In corrections, Richards and Pai (2003) evaluated 312 inmates and found varyious response styles, secondary to what the inmate believed would work best for his or her situation. Specifically, 22% of the sample faked good (denial and minimization), and almost 15% faked bad (exaggerating psychopathology). Another addictive disorder with links to deception is gambling disorder. The crux of this disorder is recurrent and problematic gambling behaviors leading to significant distress. DSM-5 lists nine specific diagnostic criteria for gambling disorder, and several directly involve deception, such as concealing gambling activities. Similar to other addictive disorders, deception is central to understanding the proposed Internet gaming disorder as lying to avoid taking responsibility and about time spent on the Internet. Application of Rogers’s explanatory models to addictive disorders provides useful information for understanding motivations. The pathogenic model may apply to persons with substance dependence on highly addictive drugs. This categorization is especially salient given the current problems many cities are having with heroin and methamphetamine. A criminogenic model might be applicable if addiction were part of a broader part of antisocial behavior (e.g., stealing to pay gambling debts

or to purchase illicit substances). Finally, the adaptational model may explain the use of substances (or gambling) as a method to cope with adversity or to escape. Awareness of motivations for addictive disorders is useful in the development of effective treatment and intervention models. Paraphilias andSexualAbuse

DSM-5 (American Psychiatric Association, 2013) refers to paraphilias as sexual deviations or perversions, with behaviors or sexual urges focusing on unusual objects, activities, or situations. These deviations and behaviors can present in many forms (i.e., voyeurism, exhibitionism, frotteurism, masochism, pedophilism, fetishism, and transvestism). According to DSM-5, there are dozens of other paraphilias, but these appear to be the most common ones (American Psychiatric Association, 2013, p.685). In many cases, these conditions lead to illegal behavior and lengthy prison sentences. In some states, convicted sex offenders can, after completing their criminal sentences, be civilly committed for the rest of their lives to receive mandated treatment to decrease dangerousness (Hoberman & Jackson, 2016; Phenix & Jackson, 2016). Given the potential consequences, it is relatively obvious why defensiveness, minimization, and lying are frequently observed in sex offenders. Defensiveness and lying refer to voluntary actions typically used to achieve a desired objective. Defensiveness is often manifested in treatment settings in which individuals with paraphilias are reluctant to discuss their sexually based behavior. Lying often occurs in treatment settings, but it may also manifest itself as part of a larger antisocial strategy to dupe unsuspecting victims. For instance, some offenders lie about their age to make an otherwise taboo sexual relationship appear normal. Other individuals who engage in sexual abuse lie (e.g., fabricate a career) to gain sexual access to potential victims. In contrast, cognitive distortions, often in the form of rationalizations, represent patterns of thinking, often referred to as schemas that are potentially less voluntary, yet occur quite frequently in sexual abusers. Carvalho and Nobre (2014) found that sex offenders who offend against children have several maladaptive schemas (e.g., pessimism) compared to non-sex offenders and sex offenders with adult victims. Treatment protocols should explore both deliberate acts of deception and automatic schemas that, although deceptive, appear less under the offender’s control (Carvalho & Nobre, 2014).

5.  Syndromes Associated with Deception 91

The previous discussion emphasizes the complexity of motivations regarding deception among sex offenders. The prevalence of deception in inpatient treatment settings is one of the primary reasons polygraphs are often employed in conjunction with intensive treatment. Although they are not used in court to prove guilt or innocence, polygraphs are combined with intensive treatment to monitor the patient’s report of his or her history and progress in treatment (Branaman & Gallagher, 2005; Kokish, Levenson, & Blasingame, 2005). Despite the use of the polygraph to monitor the veracity of the patient’s report, the use of the polygraph is highly controversial as an effective instrument to monitor patients. Personality Disorders andPsychopathy

Personality disorders are defined as “an enduring pattern of inner experience and behavior that deviates markedly from the expectations of the individual’s culture” (American Psychiatric Association, 2013, p.646). These deviations can be found in cognitions, affectivity, interpersonal functioning, and impulse control. As it concerns deception, ASPD is frequently associated with lying and deception. However, other personality disorders also have been implicated. Borderline, narcissistic, and histrionic (referred to as Cluster B disorders) personality disorders are linked to deception (Rose & Wilson, 2014). Deception is also observed in dependent, schizotypal, and avoidant personality disorder. In each of these diagnoses, the deception is manifested differently. In ASPD, lying is often part of a general predilection to engage in criminal behavior. Specific deceptive behaviors include repeated lying, use of aliases, or conning others (American Psychiatric Association, 2013, p.659). The deception in ASPD is instrumental and generally is designed to advance the agenda of the individual. In contrast, deception in borderline personality disorder frequently occurs in the context of identity disturbance (Engle & O’Donohue, 2012). Deception may occur in narcissism and histrionic personality disorders to compensate for low self-esteem by presenting a facade or to create excitement (see Mandal & Kocur, 2013). In contrast, individuals with avoidant personality disorder may engage in deception to disengage from and circumvent social situations. In dependent situations, individuals deceive in order to present as helpless, with an inability to fend for themselves, so that others can assume responsibility for them.

Psychopathy is defined by Hare (1996) as a “socially devastating personality disorder defined by a constellation of affective, interpersonal, and behavioral characteristics, including egocentricity, manipulativeness, deceitfulness, lack of empathy, guilt or remorse, and a propensity to violate social and legal expectations and norms” (p.105). The Psychopathy Checklist—Revised (Hare, 2003) consists of four interrelated facets: Interpersonal, Affective, Lifestyle, and Antisocial tendencies. As described by Gillard (Chapter 9, this volume), the interpersonal features of psychopathy are most associated with deception, because that facet includes various traits and behaviors of glibness, superficiality, lying, and conning behavior. The Interpersonal Measure of Psychopathy (Kosson, Steuerwald, Forth, & Kirkhart, 1997) has been developed to assess these traits during clinical interviews. With a well-defined factor structure (Vitacco & Kosson, 2010), interpersonal behaviors associated with psychopathy can be identified even in relatively brief periods of time (Fowler, Lilienfeld, & Patrick, 2009). For individuals with high levels of psychopathy, professionals should expect they would engage in deceptive behaviors across a variety of settings. Psychopathy has been linked to deceptive behaviors at work (Mathieu, Hare, Jones, Babiak, & Neumann, 2013; Ragatz, Fremouw, & Baker, 2012) and malingering in forensic settings (Gacono, Meloy, Sheppard, Speth, & Roske, 1995). Although many psychopaths engage in deception, it does not appear that psychopaths are skilled at deception or successful at malingering (Rogers & Cruise, 2000). However, not all research has found a link between psychopathy and malingering. Consistent with Kropp and Rogers (1993), several studies (Kucharski, Duncan, Egan, & Falkenbach, 2006; Ray et al., 2013) indicated a lack of a relationship between psychopathic traits and feigning. Moreover, individuals with psychopathic traits did not appear incrementally effective at feigning mental health disorders. Robinson and Rogers (2015) found that many inmates, irrespective of psychopathy, demonstrate the ability to feign empathy, which may lead to disruptions in specific types of therapeutic communities. Overall, there appears to be an interesting conundrum related to psychopathy and deception. On one hand, those who score high on interpersonal features of psychopathy lie frequently and easily. On the other hand, psychopaths do not appear to be particularly adept at malingering and are as likely to be detected as other offenders. At no point should the

92

II .   D i a g n o s t ic I ss u e s

term psychopathy be substituted for a classification of malingering or interpreted as evidence of the presence of deception.

OTHER CLINICAL PHENOMENA ASSOCIATED WITHDECEPTION A remarkable variety of conditions and situations are often associated with deceptive behavior. This chapter cannot possibly provide a comprehensive account of all disorders and situations associated with deception. Instead, this section focuses on factors of child custody that lead to deception and memory issues, including false-memory syndrome. In the third edition of this volume, I included chronic fatigue syndrome; however, research does not support its continued inclusion in this section on deception. The goal is that this section provide specific situations in which malingering is common and apply explanatory models to explain why deception occurs.

Contested Child CustodyEvaluations Contested child custody evaluations provide several opportunities for deception, including a unique balance of denying one’s own shortfalls and exaggerating the faults of one’s ex-partner. This dynamic may be akin to an awkward dance as one makes one’s previous partner appear unfit and incapable, even despite contrary evidence (see Gould, Rappaport, & Flens, Chapter 25, this volume). The first response style is consistent with defensiveness and/or minimization. A seemingly better term for this (Strong, Greene, Hoppe, Johnston, & Olesen, 1999) is self-deceptive positivity (SDP), which refers to the fact that many individuals undergoing an evaluation to assist in the determination of custodial placement of a child act in a manner to augment their strengths and purposely minimize or deflect negative traits. Each construct is discussed separately in the next two paragraphs. In contested custody evaluations, the examinees’ goal is to create an image of themselves as excellent parents who place the need to their children first. The conscious goal is to appear well-adjusted by minimizing psychopathology and maladaptive traits (Erickson, Lilienfeld, & Vitacco, 2007). Like malingering, SDP is considered a dimensional construct (Strong et al., 1999) and should not be considered “all or none,” and it varies as a function of the individual and the custody situation. Defensive responding has been

evident in research employing multiscale inventories in contested child custody evaluations (see Ackerman & Ackerman, 1997; McCann et al., 2001; Medoff, 1999), including the Minnesota Multiphasic Personality Inventory–2 (Butcher, Dahlstrom, Graham, Tellegen, & Kaemmer, 1989) and the Millon Clinical Multiaxial Inventory–III (MCMI-III; Millon, 1994). Regarding the MCMIIII, Stolberg and Kauffman (2015) note the importance of evaluators being aware of the implications of positive impression management on test results. As a final note, the MCMI-IV has been published, but there is no research on it with parents undergoing custody evaluations. The final area to address is relevant to how some custody litigants have been alleged to attempt to alienate children from the other parent. This phenomenon, which has a lengthy history in the child custody field, has been aptly referred to as parental alienation syndrome (PAS). In PAS cases, one parent uses deception to cause a rift between the child and the other parent by belittling or even lying about the other parent to the child. However, PAS, has remained a controversial term, with limited scientific support. Emery (2005) suggested that in the absence of objective standards, PAS should be considered a hypothetical construct. In the same vein, Thomas and Richardson (2015) referred to PAS as “junk science” due to its lack of general acceptability in the field. Finally, the National Council of Juvenile and Family Court Judges found PAS to be lacking in scientific merit and encouraged the court to reject testimony about PAS (see Thomas & Richardson, 2015, p.22). Whether PAS is a syndrome remains an empirical question, but the use of deception to influence a child in a contested child custody evaluation does occur.

False-MemorySyndrome The idea of false-memory syndrome (FMS; see McNally, Chapter 20, this volume) was first advanced in 1992 by an organization of the same name (False Memory Syndrome Foundation, 2016). Since its creation, research has been appropriately devoted to determining the reliability and validity of FMS. Much of the fanfare has stemmed from high-profile cases in which memories seemingly emerged from nowhere to implicate a person or group of people in a sensational case. FMS describes a cluster of symptoms in which traumatic memories remain in the absence of external evidence. From this perspective, FMS acts like psy-

5.  Syndromes Associated with Deception 93

chodynamic concepts of regression and repression. Far from harmless, such memories have led to criminal prosecutions and civil litigation (Wakefield & Underwager, 1992). Memory itself remains problematic in legal settings, due in large part to the fallibility of human encoding and memory. Dr. Elizabeth Loftus (Loftus & Ketcham, 1994) challenged many techniques endorsed by providers who claimed to be able to recover memories. These techniques include hypnosis (for a review, see Loftus & Davis, 2006). Given the high stakes (e.g., criminal charges) that may be associated with retrieved memories, these cases are quite contentious, with both sides advocating for their specific positions. Moreover, memories have been shown to be substantially altered by police through stressful interrogative techniques. False confessions often are associated with devastating consequences (e.g., conviction of an innocent person for a serious offense). Persinger (1992) reported on six cases of questionable recall in which patients were subjected to hypnosis as part of their treatment. Williams (1994) found that approximately 38% of women brought to the hospital for sexual abuse as children did not recall the abuse many years later. Other scholars are not convinced about problems with repressed memories. Whitfield (2001) believes that perpetrators claim FMS in order to undermine memories and minimize their involvement in criminal behavior. Raitt and Zeedyk (2003) suggested that trauma-based memories are problematic because it is often females who report the recovered memories and, according to the authors, females have traditionally been viewed as less credible. In summary, the research in support of retrieved memories do not view them as stemming from deception, and instead focus on supporting the individuals recalling the memories. Clinicians should continue to consider multiple possibilities when dealing with sudden-onset memories (see McNally, Chapter 20, this volume). A careful review of history should include multiple collateral interviews. In forensic evaluations, the evaluator should determine whether the memories were recovered with the assistance of a therapist and attempt to obtain permission to speak to the therapist to determine the reliability of his or her methods. As noted by Cunnien (1997), dissimulation should be considered in litigated cases of alleged abuse. He wrote, “Factitious claims should be suspected when psychological evaluation reveals a substantial likelihood that revenge, displacement of anger, or recent abandonment triggered the abuse allegations” (p.40).

CONCLUSIONS This chapter has focused on several critical areas of which clinicians should be cognizant when evaluating the potential for deception. This final section underscores several important findings as I discuss the context of clinical and forensic evaluations. This chapter demonstrates the broad array of child and adult diagnoses in which deception is associated with mental disorders. Key points are enumerated: 1. Deception is explicitly listed as criteria for psychopathy and ASPD. 2. Deception can subtle and motivated by selfprotection in RAD. 3. Deception occurs for a variety of reasons and may be adaptive (Rogers, 1990a, 1990b) or arise from a mental disorder; clinicians should not assume a malicious intent for the deception. 4. Deception should not be dichotomized as “all or none.” Instead, clinicians should realize that deception is a dimensional construct with gradients and is used selectively. Evaluators must ethically evaluate information from multiple sources when rendering opinions about response styles. This point is especially salient in forensic evaluations in which data from multiple collateral individuals are critical to reaching a well-developed and appropriate forensic opinion. When it is determined that an individual is engaging in deception, a thorough evaluation must consider various motivations, as well as mental disorders. As noted by Rogers (Chapter 1, this volume), malingering and genuine mental disorders are not mutually exclusive. Finally, clinicians should use standardized methods to evaluate malingering and deception (Vitacco & Tabernik, 2017). Clinicians should be aware of two useful models when evaluating deception: threshold and clinical decision models. The threshold model serves as a screen for identifying potential cases of deception that require further evaluation. The bar is intentionally set low to minimize missing persons engaged in a specific response style. In contrast, the clinical decision model represents the process whereby a definite conclusion is reached about the presence or absence of malingering (or any clinical condition for that matter). The bar for reaching a definitive diagnosis must be substantially higher

94

II .   D i a g n o s t ic I ss u e s

compared to that for the threshold model. In the case of malingering and other response styles, such a definitive conclusion should not be reached without substantial evidence to support the conclusion.

REFERENCES Ackerman, M., & Ackerman, M. (1997). Custody evaluation practices: A survey of experienced professionals (revisited). Professional Psychology: Research and Practice, 28, 137–145. American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: Author. American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: Author. Ayoub, C. C., Alexander, R., Beck, D., Bursch, B., Feldman, K. W., Libow, J., et al. (2002). Position paper: Definitional issues in Munchausen by proxy. Child Maltreatment, 7, 105–111. Baker, A. L., Tabacoff, R., Tornusciolo, G., & Eisenstadt, M. (2001). Calculating number of offenses and victims of juvenile sexual offending: The role of posttreatment disclosures. Sexual Abuse: Journal of Research and Treatment, 13, 79–90. Barbaree, H. E., Marshall, W. L., & Hudson, S. M. (1993). The juvenile sex offender. New York: Guilford Press. Berthelson, L., Mulchan, S. S., Odland, A. P., Miller, L. J., & Mittenberg, W. (2013). False positive diagnosis of malingering due to the use of multiple effort tests. Brain Injury, 27, 909–916. Blankenship, K. L., & Whitley, B. J. (2000). Relation of general deviance to academic dishonesty. Ethics and Behavior, 10, 1–12. Branaman, T. F., & Gallagher, S. N. (2005). Polygraph testing in sex offender treatment: A review of limitations. American Journal of Forensic Psychology, 23, 45–64. Butcher, J. N., Dahlstrom, W. G., Graham, J. R., Tellegen, A. M., & Kaemmer, B. (1989). MMPI-2: Manual for administration and scoring. Minneapolis: University of Minnesota Press. Caldwell, M. F. (2002). What we do not know about juvenile sexual re-offense risk. Child Maltreatment, 7, 291–302. Caldwell, M. F., & Dickinson, C. (2009). Sex offender registration and recidivism risk in juvenile sexual offenders. Behavioral Sciences and the Law, 27, 941–956. Carvalho, J., & Nobre, P. J. (2014). Early maladaptive schemas in convicted sexual offenders: Preliminary findings. International Journal of Law and Psychiatry, 37, 210–216. Cunnien, A. J. (1997). Psychiatric and medical syndromes associated with deception. In R. Rogers (Ed.),

Clinical assessment of malingering and deception (2nd ed., pp.23–46). New York: Guilford Press. Dalzell, H. J. (2000). Whispers: The role of family secrets in eating disorders. Eating Disorders: The Journal of Treatment and Prevention, 8, 43–61. Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. 579 (1994). Denney, R. L. (2005). Criminal responsibility and other criminal forensic issues. In G. J. Larrabee (Ed.), Forensic neuropsychology: A scientific approach (pp.425– 465). New York: Oxford University Press. Drob, S. L., Meehan, K. B., & Waxman, S. E. (2009). Clinical and conceptual problems in the attribution of malingering in forensic evaluations. Journal of the American Academy of Psychiatry and the Law, 37, 98–106. Emery, R. (2005). Parental alienation syndrome: Proponents bear the burden of proof. Family Court Reviews, 43, 8–13. Engle, J., & O’Donohue, W. (2012). Pathways to false allegations of sexual assault. Journal of Forensic Psychology Practice, 12(2), 97–123. Erickson, B. J., Yanke, A., Monson, B., & Romeo, A. (2015). The effect of the steroid era on major league baseball hitters: Did it enhance hitting? Journal of Sport Medicine Doping Studies, 5, 161–163. Erickson, S. K., Lilienfeld, S. O., & Vitacco, M. J. (2007). Failing the burden of proof: The science and ethics of projective tests in custody evaluations. Family Court Review, 45, 185–192. False Memory Syndrome Foundation. (2016, October 5). Retrieved from www.fmsfonline.org. Ferrari, J. R., Groh, D. R., Rulka, G., Jason, L. A., & Davis, M. I. (2008). Coming to terms with reality: Predictors of self-deception within substance abuse recovery. Addictive Disorders and Their Treatment, 7(4), 210–218. Forth, A., Kosson, D., & Hare, R. D. (2003). The Hare Psychopathy Checklist—Youth Version. Toronto: Multi-Health Systems. Fowler, K. A., Lilienfeld, S. O., & Patrick, C. J. (2009). Detecting psychopathy from thin slices of behavior. Psychological Assessment, 21, 68–78. Frick, P. J., Lahey, B. B., Loeber, R., Stouthamer-Loeber, M., Green, S., Hart, E. L., et al. (1991). Oppositional defiant disorder and conduct disorder in boys: Patterns of behavioral covariation. Journal of Clinical Child Psychology, 20, 202–208. Frye v. United States, 293 F. 1013 (D.C., Cir 1923). Gacono, C. B., Meloy, J. R., Sheppard, K., Speth, E., & Roske, A. (1995). A clinical investigation of malingering and psychopathy in hospitalized insanity acquittees. Bulletin of the American Academy of Psychiatry and the Law, 23, 387–397. Gervais, J., Tremblay, R. E., & Héroux, D. (1998). Boys’ lying and social adjustment in pre-adolescence: Teachers’, peers’ and self-reports. Criminal Behaviour and Mental Health, 8(2), 127–138. Gostisha, A. J., Vitacco, M. J., Dismukes, A. R., Brie-

5.  Syndromes Associated with Deception 95

man, C., Merz, J., & Shirtcliff, E. A. (2014). Beyond physiological hypoarousal: The role of life stress and callous unemotional traits in incarcerated adolescent males. Hormones and Behavior, 65, 469–479. Gregory, R. J., & Jindal, S. (2006). Factitious disorder on an inpatient psychiatry ward. American Journal of Orthopsychiatry, 76, 31–36. Guyton, M. R., & Jackson, R. L. (2016). Violence risk assessment. In R. Jackson & R. Roesch (Eds.), Learning forensic assessment: Research and practice (2nd ed., pp.131–161). New York: Routledge/Taylor & Francis Group. Hare, R. D. (1996). Psychopathy: A clinical construct whose time has come. Criminal Justice and Behavior, 23(1), 25–54. Hare, R. D. (2003). The Hare Psychopathy Checklist— Revised (2nd ed.). Toronto: Multi-Health Systems. Hoberman, H. M., & Jackson, R. L. (2016). Civil commitment of select recidivistic sexual offenders deemed likely to sexually reoffend. In A. Phenix & H. M. Hoberman (Eds.), Sexual offending: Predisposing antecedents, assessments and management (pp.779–821). New York: Springer Science and Business Media. Hughes, J. N., & Hill, C. R. (2006). Lying. In G. G. Bear, K. M. Minke, G. G. Bear, & K. M. Minke (Eds.), Children’s needs III: Development, prevention, and intervention (pp.159–169). Washington, DC: National Association of School Psychologists. Klein, S. (2000). Deception and substance abuse. Dissertation Abstracts International, 60, 5348. Kokish, R., Levenson, J. S., & Blasingame, G. D. (2005). Post-conviction sex offender polygraph examination: Client-reported perceptions of utility and accuracy. Sexual Abuse: Journal of Research and Treatment, 17, 211–221. Kosson, D. S., Steuerwald, B. L., Forth, A. E., & Kirkhart, K. J. (1997). A new method for assessing the interpersonal behavior of psychopathic individuals: Preliminary validation studies. Psychological Assessment, 9, 89–101. Kropp, P. R., & Rogers, R. (1993). Understanding malingering: Motivation, method, and detection. In M.Lewis & C. Saarni (Eds.), Lying and deception ineveryday life (pp.201–216). New York: Guilford Press. Kucharski, L. T., Duncan, S., Egan, S. S., & Falkenbach, D. M. (2006). Psychopathy and malingering of psychiatric disorder in criminal defendants. Behavioral Sciences and the Law, 24, 633–644. Lacey, J. H. (1993). Self-damaging and addictive behavior in bulimia nervosa: A catchment area study. British Journal of Psychiatry, 163, 190–194. Lacey, J. H., & Evans, C. D. H. (1986). The impulsivist: A multi-impulsive personality disorder. British Journal of Addiction, 81, 641–649. Larrabee, G. J. (2012). Performance validity and symptom validity in neuropsychological assessment. Journal of the International Neuropsychological Society, 18, 625–631.

Libow, J. A., & Schreier, H. A. (1986). Three forms of factitious illness in children: When is it Munchausen syndrome by proxy? American Journal of Orthopsychiatry, 56, 602–611. Loftus, E., & Davis, D. (2006). Recovered memories. Annual Review of Psychology, 2, 469–498. Loftus, E., & Ketcham, K. (1994). The myth of repressed memory: False memory and allegations of sexual abuse. New York: St. Martin’s Press. Lynam, D. R., Charnigo, R., Moffitt, T. E., Raine, A., Loeber, R., & Stouthamer-Loeber, M. (2009). The stability of psychopathy across adolescence. Development and Psychopathology, 21, 1133–1153. Lynam, D. R., & Gudonis, L. (2005). The development of psychopathy. Annual Review of Clinical Psychology, 1, 381–407. Mandal, E., & Kocur, D. (2013). Psychological masculinity, femininity and tactics of manipulation in patients with borderline personality disorder. Archives of Psychiatry and Psychotherapy, 15, 45–53. Mart, E. G. (2002). Munchausen’s syndrome (factitious disorder) by proxy: A brief review of its scientific and legal status. Scientific Review of Mental Health Practice, 1, 55–61. Mathieu, C., Hare, R. D., Jones, D. N., Babiak, P., & Neumann, C. S. (2013). Factor structure of the BScan 360: A measure of corporate psychopathy. Psychological Assessment, 25, 288–293. McCann, J. T., Flens, J. R., Campagna, V., Collman, P., Lazzaro, T., & Connor, E. (2001). The MCM-III in child custody evaluations: A normative study. Journal of Forensic Psychology Practice, 1, 27–44. Medoff, D. (1999). MMPI-2 validity scales in child custody evaluations: Clinical versus statistical significance. Behavioral Sciences and the Law, 17, 409–411. Miller, G. A. (1994a). The Substance Abuse Subtle Screening Inventory (SASSI) manual. Bloomington, IN: SASSI Institute. Miller, G. A. (1994b). Supplement to the Substance Abuse Subtle Screening Inventory (SASSI) manual. Bloomington, IN: SASSI Institute. Miller, H. A. (2001). Miller Forensic Assessment of Symptoms Test manual. Lutz, FL: Psychological Assessment Resources. Miller, H. A. (2005). The Miller Forensic Assessment of Symptoms Test (M-FAST): Test generalizability and utility across race, literacy, and clinical opinion. Criminal Justice and Behavior, 32, 591–611. Millon, T. (1994). Millon Clinical Multiaxial Inventory– III manual. Minneapolis, MN: National Computer Systems. Mittenberg, W., Patton, C., Canyock, E. M., & Condit, D. C. (2002). Base rates of malingering and symptom exaggeration. Journal of Clinical and Experimental Neuropsychology, 24, 1094–1102. Moffitt, T. E. (1993). Adolescence limited and life course persistent antisocial behavior: A developmental taxonomy. Psychological Review, 100, 674–701. Morey, L. C. (1991). Personality Assessment Inventory

96

II .   D i a g n o s t ic I ss u e s

professional manual. Lutz, FL: Psychological Assessment Resources. Persinger, M. A. (1992). Neuropsychological profiles of adults who report ‘sudden remembering’ of early childhood memories: Implications for claims of sex abuse and alien visitation/abduction experiences. Perceptual and Motor Skills, 75(1), 259–266. Phenix, A., & Jackson, R. L. (2016). Evaluations for sexual offender civil commitment. In R. Jackson & R. Roesch (Eds.), Learning forensic assessment: Research and practice (2nd ed., pp.162–201). New York: Routledge/Taylor & Francis Group. Pope, H. G., Jonas, J. M., & Jones, B. (1982). Factitious psychosis: Phenomenology, family history, and longterm outcome of nine patients. American Journal of Psychiatry, 139, 1480–1483. Ragatz, L. L., Fremouw, W., & Baker, E. (2012). The psychological profile of white-collar offenders: Demographics, criminal thinking, psychopathic traits, and psychopathology. Criminal Justice and Behavior, 39, 978–997. Raitt, F. E., & Zeedyk, M. S. (2003). False memory syndrome: Undermining the credibility of complainants in sexual offences. International Journal of Law and Psychiatry, 26(5), 453–471. Ray, J. V., Hall, J., Rivera-Hudson, N., Poythress, N. G., Lilienfeld, S. O., & Morano, M. (2013). The relation between self-reported psychopathic traits and distorted response styles: A meta-analytic review. Personality Disorders: Theory, Research, and Treatment, 4, 1–14. Reich, P., & Gottfried, L. A. (1983). Factitious disorders in a teaching hospital. Annals of Internal Medicine, 99, 240–247. Resnick, P. J., West, S., & Payne, J. W. (2008). Malingering of posttraumatic disorders. In R. Rogers, (Ed.), Clinical assessment of malingering and deception (3rd ed., pp.109–127). New York: Guilford Press. Richards, H., & Pai, S. (2003). Deception in prison assessment of substance abuse. Journal of Substance Abuse Treatment, 24, 121–128. Robinson, E. V., & Rogers, R. (2015). Empathy faking in psychopathic offenders: The vulnerability of empathy measures. Journal of Psychopathology and Behavioral Assessment, 37, 545–552. Rogers, R. (1990a). Development of a new classificatory model of malingering. American Academy of Psychiatry and Law, 18, 323–333. Rogers, R. (1990b). Models of feigned mental illness. Professional Psychology, 21, 182–188. Rogers, R. (2004). Diagnostic, explanatory, and detection models of Munchausen by proxy: Extrapolations from malingering and deception. Child Abuse and Neglect, 28(2), 225–238. Rogers, R., Bagby, R. M., & Dickens, S. E. (1992). Structured Interview of Reported Symptoms (SIRS) and professional manual. Odessa, FL: Psychological Assessment Resources. Rogers, R., & Cruise, K. R. (2000). Malingering and de-

ception among psychopaths. In C. B. Gacono (Ed.), The clinical and forensic assessment of psychopathy: A practitioner’s guide (pp.269–284). Mahwah, NJ: Erlbaum. Rogers, R., Jackson, R. L., Sewell, K. W., & Salekin, K. L. (2005). Detection strategies for malingering: A confirmatory factor analysis of the SIRS. Criminal Justice and Behavior, 32, 511–525. Rogers, R., & Reinhardt, V. (1998). Conceptualization and assessment of secondary gain. In G. P. Koocher, J. C. Norcross, & S. S. Hill (Eds.), Psychologists’ desk reference (pp.57–62). New York: Oxford University Press. Rogers, R., Sewell, K. W., & Gillard, N. D. (2010). Structured Interview of Reported Symptoms, 2nd Edition, test manual. Lutz, FL: Psychological Assessment Resources. Rogers, R., & Vitacco, M. J. (2002). Forensic assessment of malingering and related response styles. In B. Van Dorsten (Ed.), Forensic psychology: From classroom to courtroom (pp.83–104). New York: Kluwer Academic Press. Rose, C., & Wilson, M. S. (2014). Perceptions of everyday deceptions: Individual differences in narcissism and psychopathy associated with black and white untruths. In A. Besser (Ed.), Handbook of the psychology of narcissism: Diverse perspectives (pp.229–248). Hauppauge, NY: Nova Science. Schimmenti, A., Di Carlo, G., Passanisi, A., & Caretti, V. (2015). Abuse in childhood and psychopathic traits in a sample of violent offenders. Psychological Trauma: Theory, Research, Practice, and Policy, 7, 340–347. Schraft, C. V., Kosson, D. S., & McBride, C. K. (2013). Exposure to violence within home and community environments and psychopathic tendencies in detained adolescents. Criminal Justice and Behavior, 40, 1027–1043. Shirtcliff, E. A., Vitacco, M. J., Graf, A. R., Gostisha, A. J., Merz, J. L., & Zahn-Waxler, C. (2009). Neurobiology of empathy and callousness: Implications for the development of antisocial behavior. Behavioral Sciences and the Law, 27, 137–171. Stirling, J., & the Committee on Child Abuse and Neglect. (2007). Beyond Munchausen syndrome by proxy: Identification and treatment of child abuse in a medical setting. Pediatrics, 119, 1026–1030. Stolberg, R., & Kauffman, C. (2015). The utility of the MCMI-III in child custody evaluations: The impact of clinical norms. Journal of Child Custody, 12, 325– 336. Strong, D. R., Greene, R. L., Hoppe, C., Johnston, T., & Olesen, N. (1999). Taxometric analysis of impression management and self-deception on the MMPI-2 in child-custody litigants. Journal of Personality Assessment, 73, 1–18. Thomas, R. M., & Richardson, J. T. (2015). Parental alienation syndrome: 30 years and on and still junk science. The Judge’s Journal, 54. Retrieved April

5.  Syndromes Associated with Deception 97

15, 2016, from www.americanbar.org/publications/ judges_journal/2015/summer/parental_alienation_syndrome_30_years_on_and_still_junk_science.html. Uniform Code of Military Justice. (2006–2015). Subchapter 10: Punitive articles. Article 115. U.S. v. Greer, 158 F.3d 228 (5th Cir. 1998). Vitacco, M. J., Jackson, R. L., Rogers, R., Neumann, C. S., Miller, H. A., & Gabel, J. (2008). Detection strategies for malingering with the Miller Forensic Assessment of Symptoms test: A confirmatory factor analysis of its underlying dimensions. Assessment, 15, 97–103. Vitacco, M. J., & Kosson, D. S. (2010). Understanding psychopathy through an evaluation of interpersonal behavior: Testing the factor structure of the inter­ personal measure of psychopathy in a large sample ofjail detainees. Psychological Assessment, 22, 638– 649. Vitacco, M. J., & Rogers, R. (2005). Assessment of malingering in correctional settings. In C. Scott (Ed.). Clinical handbook of correctional psychiatry (pp.133– 150). Washington, DC: American Psychiatric Publishing Press. Vitacco, M. J., Rogers, R., Gabel, J., & Munizza, J. (2007). An evaluation of malingering screens with competency to stand trial patients: A known groups comparison. Law and Human Behavior, 31, 249–260. Vitacco, M. J., & Tabernik, H. E. (2017). Structured approaches to assessing malingering. In R. Rosner & C. L. Scott (Eds.), Principles and practice of forensic psychiatry (3rd ed., pp.729–738). Boca Raton, FL: CRC Press/Taylor & Francis.

Wakefield, H., & Underwager, R. (1992). Recovered memories of alleged sexual abuse: Lawsuit against parents. Behavioral Sciences and the Law, 10, 483–507. Waller, D. A. (1983). Obstacles to the treatment of Munchausen by proxy syndrome. Journal of the American Academy of Child and Adolescent Psychiatry, 22, 80–85. Walters, G. D., Berry, D. R., Rogers, R., Payne, J. W., & Granacher, R. J. (2009). Feigned neurocognitive deficit: Taxon or dimension? Journal of Clinical and Experimental Neuropsychology, 31(5), 584–593. Walters, G. D., Rogers, R., Berry, D. R., Miller, H. A., Duncan, S. A., McCusker, P. J., et al. (2008). Malingering as a categorical or dimensional construct: The latent structure of feigned psychopathology as measured by the SIRS and MMPI-2. Psychological Assessment, 20(3), 238–247. Whitfield, C. (2001). The “false memory” defense: Using disinformation and junk science in and out of court. Journal of Child Sexual Abuse, 9, 53–78. Williams, L. M. (1994). Recall of childhood trauma: A prospective study of women’s memories of child sexual abuse. Journal of Consulting and Clinical Psychology, 62(6), 1167–1176. Wilson, S. L. (2001). Attachment disorders: Review and current status. Journal of Psychology: Interdisciplinary and Applied, 135, 37–51. Zilberstein, K. (2006). Clarifying core characteristics of attachment disorders: A review of current research and theory. American Journal of Orthopsychiatry, 76, 55–64.

CHAPTER 6

Malingered Psychosis PhillipJ.Resnick,MD JamesL.Knoll,IV,MD

Though this be madness, yet there is method in it. —William Shakespeare, Hamlet (Act II, Scene 2)

The detection of malingered psychosis demands significant time and effort. It is a complex endeavor, requiring the clinician to take a specialized, systematic approach and consider multiple sources of data. The degree of difficulty involved depends on the skill and knowledge of the malingerer. Malingerers with relatively poor understanding of the phenomenology of genuine psychotic symptoms may be readily detected. In contrast, malingerers possessing shrewdness and detailed knowledge of psychosis may deceive even seasoned forensic clinicians. For example, reputed Mafia leader Vincent “The Chin” Gigante was alleged to have deceived “the most respected minds in forensic psychiatry” by malingering, among other things, schizophrenia (Newman, 2003). Gigante ultimately admitted to deceiving multiple psychiatrists during evaluations of his competency to stand trial from 1990 to 1997. One psychiatrist, who concluded that Gigante was malingering, observed, “When feigning is a consideration, we must be more critical and less accepting of our impressions when we conduct and interpret an examination than might otherwise be the case in a typical clinical situation” (Brodie, personal communication, May 17, 2005). Nonetheless, many clinicians are reluctant to label malingering when it is suspected (Yates, Nordquist, & Schultz-Ross, 1996). Reasons for 98

this reluctance include fears of litigation and of being assaulted. In addition, adverse consequences of an incorrect classification of malingering can include denial of needed care and stigma that may be difficult to shed (Kropp & Rogers, 1993). Thus, it is particularly important for clinicians to use a systematized approach to detect malingering, as opposed to merely forming a global impression (Kucharski, Ryan, Vogt, & Goodloe, 1998) The true prevalence of malingering is not known; however, it is reasonable to conclude that prevalence and incentives to malinger vary significantly depending on the context. A small number of studies have provided estimates and suggest that base rates vary according to setting and circ*mstances. Rogers (1986) established a 4.5% prevalence rate of definite malingering and a 20% rate of moderate deception or suspected malingering among defendants being evaluated for insanity, who were judged “sane.” Cornell and Hawk (1989) found that 8% of defendants referred for pretrial assessments attempted to feign psychosis. Gottfried, Hudson, Vitacco, and Carbonell (2017) have estimated that the base rates of malingering in criminal forensic evaluations range from 20 to 30%. Finally, Pollock, Quigley, Worley, and Bashford (1997) found that 32% of prisoners referred to forensic mental health services fabricated or

6. Malingered Psychosis 99

exaggerated symptoms of mental illness. The accuracy of such prevalence estimates is highly questionable, since examinees who successfully fake psychosis are never included in the statistics. The apparent low prevalence of malingering in nonforensic populations may be due to a combination of low clinician suspicion and the relative lack of motivation to malinger.

DEFINING THEPROBLEM In the early 1900s, persons who malingered insanity were viewed as likely to be mentally “degenerate” and afflicted with an unsound mind. In 1905, a Massachusetts pool hall owner and bartender, known locally as “Joe the Snake,” appeared to be malingering insanity after committing murder. A reviewing psychiatrist concluded: “It seems questionable whether there ever occurs an occasion for simulation of insanity in those who are mentally completely sound” (Drew, 1908, p.680). During the height of psychoanalytic influence, malingering was believed to be a form of mental illness. Eissler (1951, p.252), for example, stated, “It can be rightly claimed that malingering is always a sign of disease often more severe than a neurotic disorder because it concerns an arrest of development at an early phase.” Others have found less merit to this view. For example Wertham (1949, p.49) noted, “There is a strange, entirely unfounded superstition even among psychiatrists that if a man simulates insanity there must be something mentally wrong with him in the first place. As if a sane man would not grasp at any straw if his life were endangered by the electric chair.” At the present time, the Diagnostic and Statistical Manual of Mental Disorders (DSM) model has largely supplanted the pathogenic model of malingering (Rogers, 1990). Malingering is a condition not attributable to a mental disorder. DSM-5 (American Psychiatric Association, 2013) describes malingering in terms of a false presentation (fabrication or severe exaggeration). Malingering often requires differentiation from factitious disorder, which is listed in DSM-5 in the section Somatic Symptom and Related Disorders. It uses a similar description for false presentation, but the motivation is present even in the absence of obvious external rewards. In factitious disorder, a patient usually simulates illness with a motive to assume the sick role, which can be thought of as an internal (i.e., psychological) incentive (Kanaan & Wessely, 2010). Diagnosing factitious disorder

can also be challenging, and a traditional clinical education does not often provide clinicians with the training to understand and deal with patients whose symptoms appear to be simulated (Bass & Halligan, 2014). As a separate term, feigning is defined as the deliberate fabrication or gross exaggeration of symptoms, without any assumptions about its goals (McDermott, 2012; Rogers & Bender, 2013). Malingering can be further categorized into (1) pure malingering, (2) partial malingering, and (3) false imputation (Resnick, 1997). When an individual feigns a disorder that does not exist at all, this is referred to as pure malingering. When an individual has actual symptoms but consciously exaggerates them, it is called partial malingering. False imputation refers to the attribution of actual symptoms to a cause consciously recognized by the individual as having no relationship to the symptoms. For example, a claimant suffering from posttraumatic stress disorder (PTSD) due to an earlier trauma may falsely ascribe the symptoms to a car accident in order to receive compensation. Wooley and Rogers (2014) tested this three-subcategory model of malingering by assessing whether various types of malingerers were able to fake PTSD without being classified as feigning. The partial malingering group proved to be the best at feigning. The term pseudo malingering has been used to describe a prodromal phase of genuine psychosis in which the individual allegedly feigns psychosis in a desperate attempt to ward off decompensation into genuine psychosis. A number of authors have suggested that one should carefully consider pseudo malingering before labeling it as malingering (Berney, 1973; Bustamante & Ford, 1977; Folks & Freeman, 1985; Hay 1983; Pope, Jonas, & Jones, 1982; Schneck, 1970). In the novel The Dilemma, by Leonid Andreyev (1902), a physician committed murder with a premeditated plan to feign insanity. When the physician later began to have true hallucinations, he realized that he was genuinely psychotic. While the idea that someone might become mentally ill after avoiding criminal responsibility by malingering insanity makes for a gripping mystery story, it is extremely rare in forensic practice. Hay (1983) concluded that simulated schizophrenia was a prodromal phase of genuine psychosis that occurred in extremely deviant personalities. In this study, five patients originally thought to have feigned psychosis were evaluated after a lengthy period. Upon reevaluation, four of the five patients were believed to have developed genuine

100

II .   D i a g n o s t ic I ss u e s

schizophrenia. An alternative explanation is that with time and practice, these four individuals improved their ability to malinger psychosis to the point that they were undetectable.

MOTIVATIONS TOMALINGER The motives to malinger psychosis fall into two general categories: (1) to avoid difficult situations or punishment (avoiding pain), and (2) to obtain compensation or medications (seeking pleasure). Table 6.1 lists common motives for malingering. Criminals may seek to avoid punishment by feigning incompetence to stand trial or insanity at the time of the offense. Psychotic illness may also be malingered in an effort to mitigate sentencing. Malingerers may seek to avoid military duty, undesirable military assignments, or combat. Financial gain from Social Security Disability, veteran’s benefits, worker’s compensation, or alleged psychological damages may also be motives to malinger psychosis. In nonforensic settings, malingerers may seek a psychiatric admission to secure social services, “pave the way” for future disability claims, or simply to obtain free room and board. In the correctional setting, inmates may malinger mental illness to do “easier time” or obtain prescription drugs. In certain circ*mstances, malingering may be an adaptive coping strategy. For example, a 14-year-old girl feigned hallucinations in order to be hospitalized to escape from sexual harassment by her mother’s new boyfriend (Greenfield, 1987). She had previously observed an older cousin’s genuine psychosis. When her family situation became intolerably chaotic, she was institutionalized on the basis of a feigned psychosis. She eventually acknowledged to hospital staff that she had faked her psychotic symptoms. TABLE 6.1.  Common Motives of Malingerers

Avoid pain •• Avoid arrest •• Avoid criminal prosecution •• Avoid conscription into the military Seek pleasure •• •• •• ••

Obtain controlled substances Obtain free room and board Obtain disability or worker’s compensation Obtain compensation for alleged psychological injury

RESEARCH ONMALINGEREDPSYCHOSIS If sanity and insanity exist, how shall we know them? —David Rosenhan (1973, p.250)

No research has examined the ability of clinicians to accurately detect malingered psychosis in their daily practice. There is a large body of research on detecting malingered psychosis with psychometric testing such as the Structured Interview of Reported Symptoms (SIRS; Rogers, Bagby, & Dickens, 1992), the SIRS-2 (Rogers, Sewell, & Gillard, 2010), the Stuctured Inventory of Malingered Symptoms (SIMS; Smith & Burger, 1997), and the Personality Assessment Inventory (PAI) and the Minnesota Multiphasic Personality Inventory–2 (MMPI-2; Blanchard, McGrath, Pogge, & Khadivi, 2003). These instruments are addressed in other chapters. However, researchers have yet to address the effectiveness of clinical assessment alone, without the use of these specialized methods. In Rosenhan’s (1973) classic study, eight pseudopatients were admitted to psychiatric hospitals, all alleging only that they heard very atypical auditory hallucinations. Immediately upon admission, they ceased simulating any symptoms, yet all were diagnosed with schizophrenia and remained hospitalized from 9 to 52 days. Interestingly, it was asserted that genuinely disordered inpatients commonly detected the pseudopatients’ “sanity.” On the basis of his study, Rosenhan concluded that mental health professionals were unable to distinguish normality from mental illness. However, this conclusion has been challenged. Furthermore, the criteria for schizophrenia were much more elastic in DSM-II than in DSM-5, which likely resulted in less reliable diagnoses. Anderson, Trethowan, and Kenna (1959) asked 18 normal participants to simulate mental disease in order to study the phenomenon of vorbeirden (giving an approximate answer), a central symptom in Ganser syndrome. The simulators were compared with normal controls and two comparison groups: (1) subjects with organic dementia and (2) patients with pseudodementia (primary diagnosis of hysteria with conversion symptoms). The simulators most often chose to feign depression or paranoid disorders; however, their efforts did not closely resemble well-defined mental disorders. Two feigned mental retardation by maintaining an air of obtuseness, vagueness, and poverty of content. The simulators experienced difficulty, however, in suppressing correct answers to questions.

6. Malingered Psychosis 101

This was attributed to a “pull of reality” that they felt throughout the interviews. Fatigue or difficulty sustaining the response style apparently caused simulators to become increasingly normal during the prolonged interviews. Many simulators in the Anderson et al. (1959) study gave approximate answers, because they believed they should not give the “right” answers. In an effort to avoid the impression of spuriousness, they gave nearly correct answers, in contrast to the more obvious errors by patients with actual dementia. The subjects with genuine organic dementia demonstrated substantial perseveration, which was evident when progressing from one question to the next and during serial-seven subtractions. In contrast, malingerers demonstrated no significant perseveration. The authors concluded that perseveration is a strong indication of true, rather than simulated, organic impairment. The fact that the simulators gave approximate answers lends indirect support to the theory that Ganser syndrome is a form of malingering. Ganser syndrome remains controversial and understudied. In 1897, Sigbert Ganser described three inmates who exhibited symptoms characterized by approximate answers to simple questions, dulling of consciousness, hysterical neurological changes, and hallucinations (De Dios Francos, Sanz Granado, & de la Gándara Martín, 1995). The onset and remission of these symptoms were abrupt and followed by amnesia and bewilderment. The symptom of giving approximate answers has also been described as vorbeireden (talking around). However, this symptom is not pathognomonic, as it can appear in other disorders, such as dementia and schizophrenia. Ganser himself did not view Ganser syndrome as a form of malingering. Rather, he viewed it as a hysterical dissociative reaction, resulting from an unconscious effort to escape an intolerable situation (Jiménez Gómez, & Quintero, 2012). While there are similarities and differences among Ganser syndrome, factitious disorders, and malingering, little scientific research is available to reliably clarify its underpinnings. Cornell and Hawk (1989) studied 39 criminal defendants classified by experienced forensic psychologists as malingering psychotic symptoms. The prevalence of malingering was 8.0% for 314 consecutive evaluations in a forensic hospital. The authors acknowledged that a barrier to developing diagnostic criteria was the lack of an unequivocal “gold standard” for determining malingering. Individuals classified as malingering were more likely to claim bogus symptoms, suicidal ideas,

visual hallucinations, and memory problems. Furthermore, their symptoms did not cluster into any known diagnostic entities. Powell (1991) compared 40 mental health facility employees instructed to malinger symptoms of schizophrenia and 40 inpatients with genuine schizophrenia. Both groups were administered the Mini-Mental Status Examination (MMSE; Folstein, Folstein, & McHugh, 1975), which screens for basic cognitive function. Malingerers were significantly more likely than patients with genuine schizophrenia to give one or more approximate answers on the MMSE. Malingerers also reported a higher occurrence of visual hallucinations, particularly with dramatic and atypical content (e.g., not ordinary human beings). Finally, the malingerers more often called attention to their delusions.

Hallucinations Persons with atypical hallucinations should be questioned about them in great detail. Before discussing the impact of hallucinations on current functioning, examinees should be asked to describe past hallucinations and their responses to them. Specifically, patients should be questioned about content, vividness, and other characteristics of the hallucinations (Seigel & West, 1975). Hallucinations are usually (88%) associated with delusions (Lewinsohn, 1970). Table 6.2 lists potential topics of clinical inquiry when malingered auditory hallucinations are suspected. The detection of malingered mental illness can be conceptualized as an advanced clinical skill, due to the fact that the clinician must already possess a detailed knowledge about the phenomenology of genuine psychiatric symptoms. In the case of malingered psychosis, a thorough understanding of how actual psychotic symptoms present themselves is the clinician’s greatest asset in recognizing simulated hallucinations. AuditoryHallucinations

As a first step, clinicians need to differentiate between psychotic and nonpsychotic hallucinations. About 10–15% of members of the healthy population sometimes experience auditory hallucinations (Sommer et al., 2010). Nonpsychotic hallucinations usually have a childhood onset, with a median age of 12, whereas psychotic hallucinations begin at a median age of 21. Nonpsychotic hallucinations are often attributed to family members, spirits of dead people, or guardian angels

102

II .   D i a g n o s t ic I ss u e s

TABLE 6.2.  Topics of Inquiry in Suspect Auditory Hallucinations

•• Gender

Male or female

•• Age

Child or adult

•• Vocal characteristics

Single or multiple voices, clear/ vague/inaudible, loudness

•• Frequency/ timing

Continuous or intermittent, time of day, during sleep

•• Familiarity

Known/unknown person, familiar/ unfamiliar

•• Type of language

Commands, stilted language, speaking in second or third person

•• Response

Degree of insight, ability to disregard, emotional response, converses with them

•• Associated characteristics

Hallucinations in other sensory modalities, delusions, other psychotic symptoms

rather than real people, such as a member of the Secret Service, or malevolent neighbors (Larøi, 2012). Nonpsychotic hallucinations contain very little negative content, whereas almost all schizophrenic patients report some negative content. Nonpsychotic voices do not tend to cause distress or disturbance in the daily life of the individual (Larøi, 2012). Goodwin, Alderson, and Rosenthal (1971), who studied 116 hallucinating patients, provide helpful data on the characteristics of genuine hallucinations. Both male and female voices were heard by 75% of their patients. For the majority of the time (88%), both familiar and unfamiliar voices were heard. About two-thirds of hallucinating subjects could identify the person speaking (Goodwin et al., 1971; Kent & Wahass, 1996; Leudar, Thomas, McNally, & Glinski, 1997; McCarthy-Jones et al., 2014). The content of hallucinations was accusatory in over one-third of the cases. In persons with genuine auditory hallucinations, 71% could recall the first time they heard voices (Hoffman, Varanko, Gilmore, & Mishara, 2008). It is important for clinicians to pay attention to the temporal course of onset and resolution of suspected auditory hallucinations. Some malingerers may allege that their hallucinations ceased after 1 or 2 days of treatment with antipsychotic medication. However, the first time a psychotic

patient is given antipsychotics, the median length of time for hallucinations to completely clear is 27 days (Gunduz-Bruce et al., 2005). Thus, individuals who allege complete clearing after 1 or 2 days of treatment should be viewed with suspicion. In patients treated for schizophrenic hallucinations, after 1 month, the voices became less loud and less distressing (Schneider, Jelinek, Lincoln, & Moritz, 2011). After 6 months of antipsychotics, they heard the voices less frequently and felt that they had more control of them. Many also recognized that the voices were self-generated (Schneider et al., 2011). The majority (82%) of patients with genuine hallucinations describe having them in more than one modality (McCarthy-Jones et al., 2014). Atypical features of auditory hallucinations were reported in less than 5% of patients with auditory hallucinations (McCarthy-Jones et al., 2014). Atypical voices included (1) a voice whose normal speaking tone is yelling, (2) only female voices, (3) only children’s voices, and (4) never hearing the same voice twice. Stephane, Pellizzer, Roberts, and McCalannahan (2006) identified atypical content of auditory hallucinations, such as claiming to hear the voices of animals, voices sounding robotic, voices referring to them as “Mr.” or “Mrs.” and voices occurring whenever they open a window. Hallucinated voices were most often perceived as originating from outside of the head (86%) in the Goodwin et al. (1971) study. However, Junginger and Frame (1985) reported that only 50% of patients with schizophrenia reported auditory hallucinations as originating from outside of the head. Further research has suggested that many patients with psychosis hear voices both internally and externally, and there appears to be no consistent differential effect of internal versus external auditory hallucinations (McCarthy-Jones et al., 2014; Copoloy, Trauer, & Mackinnon, 2004). A growing body of research theorizes that persons with genuine schizophrenia who experience auditory hallucinations have difficulty determining the source (self vs. other) of verbal information (Arguedas, Stevenson, & Langdon, 2012). Such source monitoring errors (e.g., external misattribution of self-generated information) appear to correlate with positive symptoms, in particular, auditory hallucinations, thought intrusion, and alien control symptoms (Ferchiou et al., 2010). Thus, questions about whether voices originate internally or externally have limited utility. Furthermore, it is possible that persons with chronic schizophrenia may have gained enough insight

6. Malingered Psychosis 103

into their illness to attribute hallucinated voices to an internal source. Baethge et al. (2005) studied hallucinations in 4,972 psychiatrically hospitalized patients. At admission, the prevalence of hallucinations by disorder was the following: schizophrenia (61.1%), bipolar mixed (22.9%), bipolar manic (11.2%), bipolar depressed (10.5%), and unipolar depressed (5.9%). Across all diagnoses, hallucinations—particularly olfactory—were significantly associated with delusions. Clinicians should be aware that the content of auditory hallucinations may vary with the individual’s culture. For example, Kent and Wahass (1996) found that the auditory hallucinations of Saudi patients were of a religious and superstitious nature, whereas instructional themes and running commentary were common in U.K. patients. Some patients view their hallucinated voices as omnipotent and omniscient (Chadwick & Birchwood, 1994). Evidence of omniscience is based on the voices knowing that person’s thoughts and being able to predict the person’s future. Hallucinated voices are often perceived as malevolent (Chadwick & Birchwood, 1994). Patients with genuine malevolent hallucinations usually develop some strategies to decrease them (McCarthy-Jones & Resnick, 2014). Patients commonly said that evil commands were evidence that the voice was “bad.” Malevolent voices evoke negative emotions (anger, fear, depression, anxiety). Patients often respond by arguing, shouting, noncompliance, and avoidance of cues that trigger voices. Benevolent voices often use kind, protective words. They usually provoke positive emotions (amusem*nt, reassurance, calm, happiness). Patients often respond to benevolent voices by elective listening, willing compliance, and doing things to bring on the voices. Most persons with genuine auditory hallucinations (81%) report that they are worried or upset by their hallucinations (Carter, Mackinnon, & Copoloy, 1996). The major themes of auditory hallucinations in schizophrenia are usually persecutory or instructive (Small, Small, & Andersen, 1966). Auditory hallucinations in schizophrenia tend to consist of personal insults, abuse, and derogatory comments about the patient or the activities of others (Goodwin et al., 1971; Oulis, Mavrea, Mamounas, & Stefanis, 1995; Leudar et al., 1997; Nayani & David, 1996). Larøi (2012) also found that auditory hallucinations in a clinical population were associated with significant distress and negative emotional content. Nayani and David (1996) found that some female subjects describe

terms of abuse conventionally directed at women (e.g., “slu*t”), while men describe male insults such as those imputing a slur about hom*osexuality (e.g., “queer”). Such voices are not likely to be faked, because they are unflattering and fail to exculpate criminal defendants. About one-third of persons with auditory hallucinations reported that voices asked them questions such as “Why are you smoking?” or “Why didn’t you do your essay?” (Leudar et al., 1997). Genuine hallucinated questions tend to chastise rather than seek information. The classic running commentary or voices conversing with each other is sometimes reported (Andreasen, 1987). However, hallucinations of music are rare in psychotic disorders (Fischer, Marchie, & Norris, 2004). Their onset is often related to organic brain pathology, aging, and sensory impairment. If the origin is due to brain disease, insight is more common than if the hallucination is the result of a mental disorder (Berrios, 1991). COMMAND AUDITORYHALLUCINATIONS

Command hallucinations are auditory hallucinations that instruct a person to act in a certain manner. McCarthy-Jones et al. (2014) found that 76% of their patients said they were able to resist their command hallucinations. Junginger (1990) reported that patients with hallucination-related delusions and hallucinatory voices were more likely to comply with the commands. In contrast, Kasper, Rogers, and Adams (1996) found that 84% of psychiatric inpatients with command hallucinations had obeyed them within the last 30 days. Among those reporting command hallucinations in a forensic population, 74% indicated that they acted in response to some of their commands during the episode of illness (Thompson, Stuart, & Holden, 1992). Junginger (1995) studied the relationship between command hallucinations and dangerousness. He found that 43% reported full compliance with their most recent command hallucination. People are more likely to obey their command hallucinations if (1) the voice is familiar, (2) there are hallucination-related delusions (Junginger, 1990), and (3) the voice is perceived as powerful (Fox, Gray, & Lewis, 2004; Shawyer et al., 2008). Compliance is less likely if the commands are dangerous (Junginger, 1995; Kasper et al., 1996). Noncommand auditory hallucinations (85%) and delusions (75%) are usually present with command hallucinations (Thompson et al., 1992).

104

II .   D i a g n o s t ic I ss u e s

The majority of commands to commit dangerous acts are not obeyed; therefore, the examiner must be alert to the possibility that a defendant may (1) fake an exculpatory command hallucination, or (2) lie about an inability to refrain from a hallucinatory command. Hellerstein, Frosch, and Koenigsberg (1987), in a retrospective chart review, found that 38% of all patients with auditory hallucinations reported commands. Studies of schizophrenic auditory hallucinations found that 30–67% included commands (Small et al., 1966; Goodwin et al., 1971; Hellerstein et al., 1987; Mott, Small, & Andersen, 1965; McCarthy-Jones et al., 2014). Command hallucinations occurred in 30 (Goodwin et al., 1971) to 40% (Mott et al. 1965) of alcoholic withdrawal hallucinations. Patients with affective disorders reported that 46% of their hallucinations were commands (Goodwin et al., 1971). Hellerstein et al. (1987) reported that the content of command hallucinations included 52% suicide, 5% homicide, 12% nonlethal injury of self or others, 14% nonviolent acts, and 17% unspecified. However, their research involved reviewing charts rather than making direct inquiries, which probably increased the relative proportion of violent commands, since these are likely to be charted. Furthermore, command hallucinations are unlikely to produce action without other psychological variables (e.g., beliefs about the voices, coexisting delusions) mediating the process (Braham, Trower, & Birchwood, 2004). Hence, someone alleging an isolated command hallucination in the absence of other psychotic symptoms should be viewed with suspicion. Leudar and colleagues (1997) found that most patients engage in an internal dialogue with their hallucinations. Many cope with chronic hallucinations by incorporating them into their daily life as a kind of internal advisor. Interestingly, sometimes patients report that their hallucinated voices insist on certain actions after patients refuse to carry them out. The voices rephrase their requests, speak louder, or curse the patient for being noncompliant. In contrast, malingerers are more likely to claim that they were compelled to obey commands without further consideration. Most patients (98%) reported experiencing significant adverse effects of their hallucinations (Miller, O’Connor, & DiPasquale, 1993), such as difficulty holding a job, emotional distress, and feeling threatened. Yet approximately half of patients also reported some positive effects of their hallucinations, such as companionship, finding

them relaxing, and making it easier to receive disability benefits. Therefore, attitudes toward hallucinations appear unhelpful in discriminating between genuine and malingered hallucinations. Persons with genuine schizophrenia typically develop a variety of coping strategies to deal with their hallucinations. Genuine hallucinations of schizophrenia tend to diminish when patients are involved in activities (Falloon & Talbot, 1981; Goodwin et al., 1971). Carter et al. (1996) found that 66% of patients with auditory hallucinations reported ways of managing the voices, and 69% described at least some success using one or more strategies. Coping strategies may involve engaging in activities (e.g., working, watching TV), changes in posture (e.g., lying down, walking), seeking out interpersonal contact, or taking medications (Falloon & Talbot, 1981; Kanas & Barr, 1984). The most common strategy for dealing with dangerous command hallucinations is prayer. Therefore, persons suspected of malingered auditory hallucinations should be asked what they do to make the voices go away or diminish in intensity. Patients with malevolent voices are likely to have developed a strategy, whereas those with benevolent voices may not be motivated to reduce their voices. The suspected malingerer may also be asked what makes the voices worse. Eighty percent of persons with genuine hallucinations reported that being alone worsened their hallucinations (Nayani & David, 1996). Voices may also be made worse by listening to the radio and watching television (Leudar et al., 1997). Genuine auditory hallucinations are characterized by a wide range of intensity, from whispers to shouting. This range is sometimes experienced within the same patient; however, the cadence of the speech is typically normal. In contrast, malingerers may report auditory hallucinations that consist of stilted or implausible language. For example, a malingerer charged with attempted rape alleged that voices said, “Go commit a sex offense.” Malingerers may also allege implausible or far-fetched commands, such as a bank robber who alleged that voices kept screaming, “Stick up, stick up, stick up!” Both examples contain language that is very questionable for genuine hallucinations, in addition to providing “psychotic justification” for an illegal act that has a rational alternative motive. VisualHallucinations

The incidence of visual hallucinations in psychotic individuals is estimated at only 24 (Mott et

6. Malingered Psychosis 105

al., 1965) to 30% (Kanas & Barr, 1984). However, Duncan (1995) found that 54.4% of inpatients with symptoms of active schizophrenia experienced visual hallucinations during their most recent acute episode. Persons with genuine visual hallucinations report that they are humanoid about 70% of the time (Goodwin et al., 1971). A minority of visual hallucinations include animals or objects. About 95% of the time, the visions are not something that the hallucinator has actually seen before. Over 80% of persons with visual hallucinations report that their response to their first visual hallucination was to be overwhelmed or fearful (Gaunlett-Gilbert & Kuipers, 2003). Occasionally, small (Lilliputian) people are seen in alcoholic, organic (Cohen, Adler, Alfonso, & Haque, 1994), or toxic psychoses (Lewis, 1961), especially anticholinergic drug toxicity (Asaad, 1990; Contardi et al., 2007). The little people are sometimes 1 or 2 inches tall; at other times, they are up to 4 feet in height. Only 5% of visual hallucinations in the study by Goodwin et al. (1971) consisted of miniature or giant figures. Psychotic visual hallucinations do not typically change if the person’s eyes are closed or open. In contrast, drug-induced hallucinations are more readily seen with the eyes closed, or in darkened surroundings (Assad & Shapiro, 1986). Unformed hallucinations, such as flashes of light, shadows, or moving objects are typically associated with neurological disease and substance use (Cummings & Miller, 1987; Mitchell & Vierkant, 1991). Visual hallucinations occurring in persons over age 60 may suggest eye pathology, particularly cataracts (Beck & Harris, 1994). Dramatic or atypical visual hallucinations should arouse suspicion of malingering (Powell, 1991). For example, one defendant charged with bank robbery was evaluated for competence to stand trial. During the evaluation, he calmly reported experiencing visual hallucinations consisting of a “thirty foot tall, red giant smashing down the walls” of the interview room. When he was asked further detailed questions about his hallucinations, he frequently replied, “I don’t know.” He subsequently admitted to malingering. DistinctiveHallucinations

DSM-5 diagnostic criteria for an alcohol-induced psychotic disorder (AIPD) include the presence of delusions and/or hallucinations, with evidence that these symptoms resulted from intoxication or withdrawal (American Psychiatric Association,

2013). Individuals with AIPD are often brought to the emergency room or to an acute-care setting. However, the psychotic symptoms are typically temporary and resolve when the substance is discontinued. AIPD is most often observed after severe intoxication (Perala et al., 2010). The symptoms improve without formal treatment in a matter of days to weeks after cessation of severe intoxication and/or withdrawal. AIPD can be clinically distinguished from schizophrenia. Persons with AIPD had higher levels of depressive and anxiety symptoms, fewer disorganized symptoms, better insight and judgment, and less functional impairment compared to patients with schizophrenia (Jordaan, Nel, Hewlett, & Emsley, 2009). In a general population study of over 8,000 subjects, increased risk of AIPD was associated with young age at onset of alcohol dependence, family alcohol problems, and multiple hospital treatments (Perala et al., 2010). Most individuals had multiple episodes of AIPD, with full recovery between episodes. A majority of individuals with AIPD have comorbid psychiatric disorders. It is not clear whether this finding indicates sensitivity to AIPD or is simply an indicator of a history of more intense substance misuse. In AIPD, auditory hallucinations are most common, but the likelihood of noise, music, or unintelligible voices is greater than that with schizophrenia. The auditory hallucinations of AIPD typically consist of accusatory, threatening, or insulting voices directed at the patient (Ali, Patel, Avenido, Jabeen, & Riley, 2011; Cummings & Mega, 2003). Persons with alcohol-induced hallucinations discuss them more easily than do persons with hallucinations due to schizophrenia (Alpert & Silvers, 1970). Mott et al. (1965) found that persons hospitalized due to alcohol misuse had an 84% prevalence of hallucinations (75% auditory and 70% visual). The major themes of the alcoholic auditory hallucinations were spirituality, persecution, and instructions concerning the management of everyday affairs. The majority of patients thought the hallucinations were real at the time but later recognized their unreality. While alcoholics are typically frightened by their hallucinations, persons with schizophrenia often become comfortable with them over the course of their illness. Hallucinations due to a medical or neurological disorder may often be distinguished from schizophrenia due to the higher prevalence of prominent visual hallucinations, and the lower prevalence of thought disorders, bizarre behavior, negative symptoms, and rapid speech (Cornelius

106

II .   D i a g n o s t ic I ss u e s

et al., 1991). Tactile hallucinations are frequently seen in cocaine-induced psychosis (e.g., cocaine bugs), and involve sensations of cutaneous or subcutaneous irritation, sometimes leading the individual to excoriate the skin with excessive scratching (­ Ellinwood, 1972). Unlike persons with schizophrenia, those with cocaine-induced psychosis do not generally report delusions of identity, grandiosity, or beliefs that their families are imposters (i.e., Capgras syndrome; Mitchell & Vierkant, 1991). Certain neurological syndromes can produce striking and relatively stereotyped complex visual hallucinations that often involve animals and human figures in bright colors and dramatic settings. The most common causes of complex visual hallucinations are epileptic disorders, brain stem lesions, and visual pathway lesions (Manford & Andermann, 1998). Peduncular hallucinosis is a syndrome of hallucinations and neurological symptoms due to a brain stem lesion. In this rare disorder, most patients were unable to discriminate their hallucinations from reality (Benke, 2006). Olfactory hallucinations are present in 13–17% of persons with schizophrenia (Langdon, McGuire, Stevenson, & Catts, 2011). Olfactory hallucinations and hallucinations of taste, touch, and bodily sensation frequently co-occur. Olfactory hallucinations often involve unpleasant odors, and self-smells may be associated with self-deprecating thought content. Olfactory and tactile hallucinations are commonly associated with general medical causes; however, persons with late-onset schizophrenia (onset after age 45) may also have visual, tactile, and olfactory hallucinations (Pearlson et al., 1989). Olfactory and gustatory hallucinations are likely to be of unpleasant odors and tastes (Goodwin et al., 1971). Olfactory hallucinations may also be associated with cerebral ischemia and epilepsy (Beume, Klingler, Reinhard, & Niesen, 2015). Persons with schizophrenia who experience olfactory hallucinations have difficulty determining whether the odor is real or imagined (Arguedas et al., 2012).

Delusions Genuine delusions vary in content, theme, degree of systemization, and relevance to the person’s life. Most delusions involve the following general themes: disease (somatic delusions), grandiosity, jealousy, love (erotomania), persecution, religion,

and being possessed (Spitzer, 1992). Grandiose delusions occur across a range of psychiatric illnesses. About two-thirds of persons diagnosed with bipolar disorder, and about 50% of those diagnosed with schizophrenia have grandiose delusions (Knowles, McCarthy-Jones, & Rowse, 2011). Slightly over 10% of the healthy general population experience grandiose thoughts that do not rise to the level of a grandiose delusion. Green et al. (2006) studied 70 individuals with persecutory delusions; the majority described the threat as severe and enduring. Depression was higher in those who felt less powerful than their persecutors. Persecutory delusions are more likely to be acted on than other types of delusions (Wessely et al., 1993). Generally, delusional systems reflect the intelligence level of the individual in terms of complexity. Delusions of nihilism, poverty, disease, and guilt are commonly seen in depression. Higher levels of psychom*otor retardation, guilt, feelings of worthlessness, and increased suicidal ideation are found more commonly in psychotic than in nonpsychotic depression (Thakur, Hays, Ranga, & Krishnan, 1999). Delusions of technical content (e.g., computer chips, telephones, telepathy) occur seven times more frequently in men than in women (Kraus, 1994). Malingerers may claim the sudden onset or disappearance of a delusion. In reality, systematized delusions usually take weeks to develop and much longer to disappear. As with suspected auditory hallucinations, the examiner should pay attention to the time course, onset, and resolution of the alleged delusion. The median length of time for delusions to clear fully after the first initiation of antipsychotic medication is reported to be 73 days (Gunduz-Bruce et al., 2005). Thus, malingering should be suspected if a person claims that a delusion suddenly appeared or disappeared. Typically, genuine delusions become somewhat less relevant, and the individual will gradually relinquish its importance after adequate treatment (Sachs, Carpenter, & Strauss, 1974). Most individuals with schizophrenia and other psychotic disorders who demonstrate disordered speech also have odd beliefs (Harrow et al., 2003). Furthermore, the more bizarre the content of the reported delusions, the more disorganized the individual’s thinking is likely to be. Therefore, when suspect delusions are alleged, the clinician should carefully consider the associated behavior and speech patterns. Table 6.3 lists suspect hallucinations and delusions.

6. Malingered Psychosis 107 TABLE 6.3.  Suspect Hallucinations andDelusions

Atypical auditory hallucinations •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• ••

Always unbearably distressing Sound mechanical or robotic Sound like voices of animals Come from inside parts of body other than head Refer to person as “Mr.” or “Mrs.” Change gender midsentence Never hearing the same voice twice Voice only yells Always vague, inaudible or mumbling Only female or only children’s voices Allegation that all command hallucinations were obeyed Hallucinations not associated with delusions No coping strategies for malevolent voices (i.e., never having any control over voices) Only auditory verbal (i.e., never any music, clicks, bangs, visual, tactile) Being alone does not increase frequency Ask questions seeking information (“What time is it?”, “What is the weather like?” versus “Why are you smoking?”, “Why didn’t you clean your room?”) Never affected by context (e.g., mood, place, circ*mstances) Stilted language (“go commit a sex offense”) Unable to recall first time hearing voices No behavioral evidence of distraction

Atypical visual hallucinations •• Black and white rather than color •• Dramatic, atypical visions •• “Schizophrenic” hallucinations that change when the eyes are closed •• Only visual hallucinations in “schizophrenia” •• Miniature or giant figures •• Visions unrelated to delusions or auditory hallucinations Atypical olfactory/gustatory hallucinations •• Pleasant odors or tastes Atypical delusions •• •• •• •• ••

Abrupt onset or termination Conduct inconsistent with delusions Bizarre content without disorganization Eagerness to discuss High conviction without adverse effects on daily functioning

With genuine delusions, the individual’s behavior usually conforms to the content of the delusions. For example, Russell Weston made a deadly assault on the U.S. Capitol building in Washington, DC, in 1998. He suffered from schizophrenia and had a delusional belief that Washington, DC was being destroyed by “cannibalism.” Allegations of persecutory delusions without any corresponding paranoid behaviors should arouse the clinician’s suspicion of malingering. One exception to this principle is persons with long-standing delusions, who have grown accustomed to their delusion and may no longer behave in a corresponding manner. Harrow et al. (2004) found that patients with schizophrenia and affective disorders with a high emotional commitment to their delusions showed poor work functioning and were likely to be hospitalized. Thus, persons who enthusiastically allege a firm conviction of their delusion should be carefully assessed for work performance and community functioning. The delusions seen in Alzheimer’s dementia frequently involve paranoid beliefs about caregivers stealing or being deceitful (Trabucchi & Bianchetti, 1997). In a study of 771 patients with Alzheimer’s dementia, Mizrahi and Starkstein (2006) found delusions in about 33% and hallucinations in approximately 7%. Delusions were significantly associated with depression, anosognosia (unawareness of illness), overt aggression, and agitation.

MalingeredMutism Mutism and mental illness have had a long-standing historical relationship with the issue of competence to stand trial (Daniel & Resnick, 1987). During the early Colonial period, persons who refused to enter a plea were considered to be either “mute by malice,” or mute by “visitation of God.” If defendants remained mute and did not put forth a plea, they were “pressed” for a plea by gradually increasing poundage placed on their chest. This is the origin of the phrase—to be “pressed for an answer.” Malingered mutism may occur as a solitary symptom or as part of a malingered psychosis. It is a difficult task to give up speech for a lengthy period, and it is not usually attempted unless the individual is facing a severe penalty or anticipating a large reward. Genuine mutism may occur in patients with or without catatonia. Mutism with catatonic stupor is recognized by the presence of posturing, negativism, automatic obedience, and waxy flexibility.

108

I I .   D i a g n o s t ic I ss u e s

Mutism without catatonia may also be seen in patients with paranoid schizophrenia who are unwilling to communicate due to paranoid distrust. In addition, corticosteroids (Kalambokis, Konitsiotis, Pappas, & Tsianos, 2006) and antihypertensive agents (Altshuler Cummings, & Mills, 1986) have been identified as possibly producing mutism without catatonia. Catatonic mutism can also occur in severe depression, mania, phencyclidine (PCP) use, and brief dissociative states. Medical etiologies include neurological disease (e.g., head injury), herpes encephalitis, tertiary syphilis, frontal lobe lesions, postictal states, akinetic mutism, and Wernicke’s encephalopathy (Altshuler et al., 1986). It is extremely common for mutism due to stroke to produce other neurological impairments. Only one case report of stroke-induced mutism without other neurological findings was published (Evyapan, 2006). Because mutism and/or catatonia are difficult states to simulate for long periods, observation of the suspected malingerer should ideally take place in an inpatient setting. Comprehensive evaluation may include the following: neurological examination, repeat interviews, observation at unsuspected times for communicative speech with peers, handwriting samples, and collateral nursing documentation. Feigned mutism may sometimes be exposed by suddenly arousing the individual from a deep sleep and immediately asking a simple question. A malingerer may reflexively reply before remembering to feign mutism (Davidson, 1952). Daniel and Resnick (1987) reported the case of a defendant who remained mute for 10 months in an effort to malinger incompetence to stand trial. The 53-year-old defendant was charged with raping and murdering an 11-year-old girl. The day after the crime, he was admitted voluntarily to a state hospital and complained of hearing voices. He stopped talking completely when told he was charged with murder. When he did not know he was being observed, he appeared to initiate conversations with fellow patients. No signs of catatonia or depression were observed. A thorough neurological workup and laboratory studies were negative. With the permission of the defendant and his attorney, a sodium amobarbital interview was conducted; the defendant described the offense and spoke for about 90 minutes but did not utter a word afterward. A careful review of collateral data revealed a pattern of voluntary admissions to psychiatric hospitals after several prior offenses, with the charges against him being dismissed.

CLINICAL INDICATORS OFMALINGEREDPSYCHOSIS The best liar is he who makes the smallest amount of lying go the longest way. —Samuel Butler

Malingerers may be detected because they have inadequate or incomplete knowledge of the illness they are faking. Indeed, malingerers are like actors who can only portray a role as best they understand it (Ossipov, 1944). However, they often overact their part (Wachpress, Berenberg, & Jacobson, 1953), or mistakenly believe that the more bizarre their behavior, the more convincing they will be. Conversely, “successful” malingerers are more likely to endorse fewer symptoms and avoid endorsing bizarre or unusual symptoms (Edens et al., 2001). Jones and Llewellyn (1917, p.80) observed that the malingerer “sees less than the blind, he hears less than the deaf, and he is more lame than the paralyzed. Determined that his insanity shall not lack multiple and obvious signs, he, so to speak, crowds the canvas, piles symptom upon symptom and so outstrips madness itself, attaining to a but clumsy caricature of his assumed role.” Malingerers are more likely to volunteer their putative symptoms in contrast to genuine patients with schizophrenia, who are often reluctant to discuss them (Ritson & Forest, 1970; Powell, 1991). For example, one male malingerer in a forensic evaluation proffered that he was an “insane lunatic” when he killed his parents at the behest of hallucinations that “told me to kill in my demented state.” Another malingering defendant who stabbed his wife to death claimed to have mistaken her for an intruder who appeared to be a dark, evil presence with a “huge skeleton head.” Without prompting, he asserted that at the time of the stabbing, he had been “crazy insane with mental illness.” A malingerer may go so far as to accuse the psychiatrist of implying that he or she is faking. Such accusations are rarely seen in genuinely psychotic individuals. DSM-5 states that the presence of antisocial personality disorder should arouse suspicions of malingering, but some studies have failed to show a relationship. Psychopathic traits were associated with malingering in one study (Edens, Buffington, & Tomicic, 2000). Yet a number of studies have suggested that antisocial or psychopathic persons are no more adept than others at malingering (Poythress, Edens, & Watkins, 2001; Rogers & Cruise, 2000) and have questioned the

6. Malingered Psychosis 109

rationale for including antisocial personality disorder as a DSM screening indicator for malingering. For example, in a study of coached malingerers, subjects who scored higher on psychopathy and intelligence did not fake symptoms more successfully when compared with other participants (Demakis, Rimland, Reeve, & Ward, 2015).

Malingered Cognitive Deficits intheSetting ofMalingeredPsychosis Some malingerers believe that faking intellectual deficits, in addition to psychotic symptoms, will make them more believable (Bash & Alpert, 1980; Edens et al., 2001; Powell, 1991; Schretlen, 1988). Thus, malingerers may give incorrect answers to patently obvious questions that an individual with a genuine, serious cognitive disorder could answer correctly. Some malingerers may also believe that they must demonstrate serious memory deficits. They may claim impairment discordant with typical patterns seen in genuine memory impairment, such as claiming long-term memory impairment worse than short-term impairment (Soliman & Resnick, 2010). Malingerers give more evasive answers than patients with genuine schizophrenia (Powell, 1991). Malingerers may also repeat questions or answer questions slowly to give themselves time to think about how to successfully deceive the evaluator. Malingerers are more likely to give vague or hedging answers to straightforward questions. For example, when asked the gender of an alleged voice, a malingerer may reply, “It was probably a man’s voice.” Malingerers may also answer, “I don’t know” to detailed questions about psychotic symptoms. The malingerer who has never experienced the symptoms “doesn’t know” the correct answer. Malingerers have more difficulty imitating the form and process of psychotic thinking than the content of a bogus delusion (Sherman, Trief, Sprafkin, 1975). Psychotic symptoms such as derailment, neologisms, loose associations, and word salad are difficult to simulate. If malingerers are asked to repeat an idea, they may do it exactly, whereas genuine patients with schizophrenia often become tangential. As previously noted, malingerers rarely show perseveration, which is more likely to suggest brain pathology, or conversely, an extremely well-prepared malingerer. Also, malingerers are unlikely to imitate the subtle signs of schizophrenia, such as negative symptoms (e.g., flat affect, alogia, avolition), impaired relatedness, digressive speech, or peculiar thinking. In con-

trast, they find the positive symptoms (e.g., hallucinations, delusions) easier to feign because of their more obvious nature. Malingerers’ alleged symptoms may not fit into any known diagnostic category, instead representing symptoms from various psychoses. Therefore, malingering should always be considered before making a diagnosis of “unspecified schizophrenia spectrum and other psychotic disorder.” Table 6.4 lists a number of clinical factors suggestive of malingering. Persons with true schizophrenia may also malinger additional symptoms to avoid criminal responsibility or seek an increase in disability compensation. For example, a man with genuine schizophrenia in partial remission killed his mother because she would not give him money to purchase cocaine. He then alleged that the murder was due to a command hallucination from God. Such cases are very difficult to assess accurately for several reasons. First, clinicians usually have a lower index of suspicion because of the individual’s documented history of genuine psychotic symptoms. Second, these malingerers are able to draw on their own previous experience with psychotic symptoms and observations of other patients while in a psychiatric hospital. In essence, they know what questions to expect from the clinician and may be better equipped to successfully provide deceptive answers. Finally, some clinicians have a tendency to dichotomize forensic patients into either “mad or bad.” Therefore, it is important that clinicians not conceptualize malingering and genTABLE 6.4.  Clinical Factors Suggestive ofMalingering

•• Absence of active or subtle signs of psychosis •• Marked inconsistencies or contradictions •• Improbable psychiatric symptoms ||Mixed symptom profile: Endorse depressive symptoms plus euphoric mood ||Overly dramatic ||Extremely unusual: “Do you believe that cars are members of organized religion?”, “Do you ever see words coming out of people’s mouths spelled out?” •• Evasiveness or noncooperation ||Excessively guarded or hesitant ||Frequently repeats questions ||Frequently replies, “I don’t know” to simple questions ||Hostile, intimidating: Seeks to control interview or refuses to participate •• Psychological testing indicates feigning: SIRS-2, M-FAST, MMPI-2

110

II .   D i a g n o s t ic I ss u e s

uine psychosis as an “either–or” situation (Rogers, Sewell, & Goldstein, 1994). The following case illustrates clinical indicators of feigned psychosis. Mr. B was charged with murder and referred by his defense attorney for evaluation of a possible insanity defense. Mr. B reported that he had his brother killed because he believed his brother had betrayed the family and “violated his honor.” Upon evaluation, Mr. B was found to have no psychiatric history, and nothing that suggested insanity at the time of his offense. Mr. B’s attorney referred him for a second psychiatric evaluation. Mr. B told the second evaluator about persons wanting to destroy the family business and replacing his brother with a robot. The second defense psychiatrist accepted the robot story at face value. Interviews with family members did not confirm the robot story. Upon arrest, Mr. B had denied any involvement in his brother’s killing and made no mention of the robot story. Mr. B showed the following clinical indicators of malingering: (1) absence of past psychotic symptoms, (2) an atypical delusion, (3) nonpsychotic alternative motives of anger and greed, (4) absence of behavior consistent with his alleged delusion, and (5) contradictions in his story.

CLINICAL INTERVIEWAPPROACH Because of the complexities involved in conclusions of malingering with reasonable professional certainty, a comprehensive malingering evaluation is recommended, particularly in difficult cases (Rogers, Vitacco, & Kurus, 2010; Drob, Meehan, & Waxman, 2009). An outline for the comprehensive evaluation of malingering is given in Table 6.4. Any information that might assist in supporting or refuting alleged symptoms should be carefully reviewed (e.g., prior treatment records, insurance records, police reports, collateral interviews). Clinicians should utilize multiple sources of data, including interviews, collateral sources, and psychometric tests in detecting malingering (Zapf & Grisso, 2012). Reliance on clinical interviews alone does not allow the examiner to classify malingering in any but the most obvious cases. Good interviewing techniques are critical to accurately detecting malingering. When malingering is suspected, the clinician should refrain from showing suspicion, and proceed in conducting an objective evaluation. A clinician’s annoyed or incredulous response is likely to result in the examinee becoming more defensive, thus decreasing

the ability of the clinician to detect malingering. It is important to begin the evaluation by asking open-ended questions that allow examinees to report symptoms in their own words. In the initial stage, inquiries about symptoms should be carefully phrased to avoid asking leading questions that give clues to the nature of genuine psychotic symptoms. Later in the interview, the clinician can proceed to more detailed questions of specific symptoms, as discussed below. The clinician should also attempt to ascertain whether each examinee has ever had the opportunity to observe persons with psychosis (e.g., during employment or in prior hospitalizations). The interview may be prolonged, since fatigue may diminish a malingerer’s ability to maintain fake symptoms (Anderson et al., 1959). In very difficult cases, inpatient assessment should be considered, because feigned psychotic symptoms are extremely difficult to maintain 24 hours a day (Broughton & Chesterman, 2001). Clinicians should be on alert for rare or improbable symptoms. Improbable symptoms are almost never reported, even in severely disturbed patients (Thompson, LeBourgeois, & Black, 2004). Malingerers may be asked about improbable symptoms: for example, “When people talk to you, do you see the words they speak spelled out?” (Miller, 2001), or “Have you ever believed that automobiles are members of an organized religion?” or “Do you ever have periods when you experience upside down vision?” (Rogers, 1987). Finally, some clinical strategies are controversial but may assist the clinician in very difficult cases. For example, one strategy is to mention, within earshot of the suspected malingerer, some easily imitated symptom that is not present. The rapid endorsem*nt of that symptom would suggest malingering. Another strategy is to ask examinees a relevant question at the moment that they have “let down their guard,” such as during a break or a relaxed moment during the evaluation. In particularly difficult cases, a clinician may consider a clinical technique called set shifting, which involves interspersing questions for malingering detection within history taking, or even in informal conversation. This shifting may cause the defendant to momentarily forget or abandon the malingering role, and important data may be revealed. This strategy was used in the Gigante case to conclude that he was malingering a severe dementia. Gigante claimed that he was disoriented with regard to time. During a break in the examination, Gigante was asked in an offhand manner whether he knew

6. Malingered Psychosis 111

what time it was. Without hesitation, Gigante reflexively responded with the correct time (J. D. Brodie, personal communication, May 17, 2005).

Inconsistencies Lying is like alcoholism. You are always recovering. —Steven Soderbergh

The clinician should pay close attention to evidence of inconsistency or contradiction in evaluating a suspected malingerer. Inconsistencies may be classified as either internal or external to the individual’s presentation. Table 6.5 lists examples of important internal and external inconsistencies. Internal inconsistencies are evident when malingerers report severe symptoms, such as mental confusion or memory loss, but are able to clearly articulate multiple examples of confusion or memory loss. Another type of internal inconsistency occurs when malingerers give markedly conflicting versions of their own history to the same evaluator. External inconsistency occurs between what the subject reports and the symptoms that are observed. For example, a malingerer may allege active auditory and visual hallucinations, yet show no evidence of being distracted. However, not all patients who hallucinate show external evidence of this. External inconsistency may be apparent between the examinee’s self-reported level of functioning and observations of his or her functioning by others. For example, a malingerer may behave in a disorganized or confused manner around the clinician, yet play excellent chess with other pa-

TABLE 6.5.  Inconsistencies Seen inMalingerers

Internal •• In subject’s own report of symptoms: giving a clear and articulate explanation of being confused •• In subject’s own reported history: giving conflicting versions External •• Between subject’s reported and observed symptoms •• Between subject’s reported level of functioning and observed level of functioning •• Between subject’s reported symptoms and the nature of genuine symptoms •• Between subject’s reported symptoms and psychological testing

tients. There may also be inconsistency between the subject’s reported symptoms and how genuine symptoms actually manifest themselves. For example, a malingerer may report seeing visual hallucinations in black and white, whereas genuine visual hallucinations are generally seen in color.

CLINICALAPPLICATIONS In a clinical setting, malingering should always be considered when there is the possibility of an external incentive for the patient. Otherwise, small separate clues of feigning may be overlooked that would lead to a more detailed investigation. In cases of suspected malingered psychosis, the clinician should inquire about the specific details of hallucinations and delusions, since the typical characteristics of these symptoms have been well researched. A major focus of this book is the development of clinical decision models to establish malingering. Table 6.6 offers such a model for malingered psychosis. The classification of malingering requires feigning plus the following: (1) the motivation is conscious, and (2) the motivation is an external incentive as opposed to a desire to be in the sick role. Furthermore, to reach a firm conclusion of malingered psychosis, the clinician must observe TABLE 6.6.  Model Criteria fortheAssessment of Malingered Psychosis

A. Clear external incentive to malinger B. Marked variability of presentation as evidenced by at least one of the following: 1. Marked discrepancies in interview and noninterview behavior 2. Gross inconsistencies in reported psychotic symptoms 3. Blatant contradictions between reported prior episodes and documented psychiatric history C. Improbable psychiatric symptoms as evidenced by one or more of the following: 1. Elaborate psychotic symptoms that lack common paranoid, grandiose or religious themes 2. Sudden emergence of alleged psychotic symptoms to explain criminal behavior 3. Atypical hallucinations or delusions (see Table 6.3) D. Confirmation of malingered psychosis by either: 1. Admission of malingering 2. Strong corroborative data, such as psychometric testing or past proof of malingering

112

II .   D i a g n o s t ic I ss u e s

(1) inconsistent or contradictory presentations, (2) improbable or incongruous clinical presentations, and (3) supportive collateral data. The following case illustrates the importance of detecting malingering in a clinical setting. Mrs. C, a 43-year-old woman, successfully malingered chronic schizophrenia over a 12-year period in order to receive Social Security Disability payments. Whenever Mrs. C went to her disability evaluations or a disability hearing, she dressed bizarrely and gave absurd answers to questions. She alleged confusion so severe that she was unable to drive or to meet her own basic needs. At her disability hearing, she testified that her entire family was dead, and refused to give any telephone numbers of third parties. Mr. and Mrs. C worked together on a number of other disability frauds using false Social Security numbers. Mrs. C was ultimately caught when it was discovered that she was in graduate school. Subsequently, she was videoed by federal agents appropriately dressed and attending graduate classes.

FORENSICAPPLICATIONS There were no real demons, no talking dogs, no satanic henchmen. I made it all up via my wild imagination so as to find some form of justification for my criminal acts against society. —“Son of Sam” serial killer David Berkowitz (in Samenow, 1984, p.130)

Concern about defendants faking mental illness to avoid criminal responsibility dates back at least to the 10th century (Brittain, 1966; Collinson, 1812; Resnick, 1984). By the 1880s, many Americans considered physicians to be a generally impious, mercenary, and cynical lot who might participate in the “insanity dodge” (Rosenberg, 1968). After the Hinckley verdict, columnist Carl Rowan (1982, p.10B) stated, “It is about time we faced the truth that the ‘insanity’ defense is mostly the last gasp legal maneuvering, often hoaxes, in cases where a person obviously has done something terrible.” In cases that capture national attention, a finding of insanity often results in a public outcry that forensic mental health professionals are fanciful, paid “excuse makers.”

Clinical Assessment ofCriminalDefendants When evaluating criminal defendants in a forensic setting, the clinician should always consider

malingering (Glancy et al., 2015). Particularly in federal courts, there has been increased attention on the issue of malingering. Several cases have upheld sentencing enhancements for defendants who have malingered incompetence to stand trial (U.S. v. Batista. 448 F.3d 237, 238, 3rd Cir., 2007; U.S. v. Binion. 132 Fed Appx. 89, 8th Circ., 2005, U.S. v. Greer. 158 F.3d 228, 5th Cir., 1998). Prior to evaluating a defendant, the clinician should be equipped with as much background information as possible, such as police reports, witness statements, autopsy findings, past psychiatric records, statements of the defendant, and observations of correctional staff. Consultations with family members, social contacts, or witnesses are often helpful prior to the clinician’s examination. The clinician should also attempt to learn some relevant information about the defendant or crime that the defendant does not know the clinician knows. This approach provides a method of assessing veracity, in that the information can be compared to the defendant’s self-report upon questioning. For example, will the defendant honestly report past criminal activity as recorded on his “rap sheet”? Or (see Hall, 1982), how does the defendant’s version of the offense compare to victim or witness accounts? An attempt should be made to evaluate the defendant who raises psychiatric issues as a defense as soon as possible after the crime. An early evaluation reduces the likelihood that defendants will have been coached or will have observed genuine psychotic symptoms in a forensic hospital setting. Defendants will have less time to plan deceptive strategy, craft a consistent story, and rehearse fabrications. Normal memory distortions are also less likely to occur. Moreover, prompt examination enhances the clinician’s credibility in court. Defendants who present with a mixed picture of schizophrenia and antisocial features may pose difficulties for the clinician due to negative countertransference feelings. Such a scenario may cause the clinician to focus on the antisocial traits to the exclusion of a genuine comorbid illness (Travin & Protter, 1984). The clinician must guard against accepting the psychotic version at face value, or dismissing it out of hand. Any facile attempt to dichotomize a defendant into “mad” (assuming the credibility of the psychotic symptoms) or “bad” (assuming the fabrication of psychotic symptoms) may reduce the accuracy of the forensic assessment. The farsighted clinician will record in detail the defendant’s early account of the crime, even if he

6. Malingered Psychosis 113

is not competent to stand trial. Once defendants are placed in a jail or forensic hospital, they may learn how to modify their story to avoid criminal responsibility (Samenow, 1984). Recording the early version also reduces the likelihood of being misled later by a defendant’s unconscious memory distortions. The clinician should take a careful history of past psychiatric illnesses, including details of prior hallucinations, before eliciting an account of the current crime. Malingerers are less likely to be on guard because they infrequently anticipate the relevance of such information to the current insanity issue. If defendants should subsequently fake hallucinations to explain his criminal conduct at the time of the offense, it will be too late to falsify their past symptoms to lend credence to the deception. Whenever possible, a defendant’s report of prior hallucinations and delusions should be confirmed by review of past hospital records. Kucharski et al. (1998) found that malingerers with no history of psychiatric treatment were likely to evidence current psychiatric presentations inconsistent with their recent Global Assessment of Functioning (GAF) as well as atypical hallucinatory experiences. Jaffe and Sharma (1998) found malingering defendants exhibited more uncommon psychiatric symptoms such as coprophagia, eating co*ckroaches, and seeing “little green men.”

Malingered Incompetence toStandTrial In a review of 8,416 forensic evaluations of competence to stand trial, Warren et al. (2006) found that clinicians opined incompetence in 19% of the cases. Findings of incompetence were strongly associated with clinical findings of psychosis and organic/intellectual disorders. Criminal defendants may seek to malinger psychosis in an effort to be found incompetent to stand trial. The defendant who successfully malingers incompetence to stand trial and is found “unrestorable” may have the benefit of less scrutiny and more freedom than an insanity acquittee. In some cases, defendants’ charges are dropped altogether. Defendants adjudicated incompetent to stand trial may malinger mental illness to avoid trial and/or pave the way for an insanity defense. McDermott, Dualan, and Scott (2013) found a rate 17.5% of malingering in patients found incompetent to stand trial who had been sent to a hospital for restoration. Understandably, the highest rates of malingering were observed in patients found incompetent for more serious offenses (e.g., murder).

A systematized, multimodal approach to detecting feigned incompetence to stand trial should consider the Evaluation of Competency to Stand Trial—Revised (ECST-R), a standardized interview designed to assess competence to stand trial, as well as serve as a screening tool for feigned incompetence to stand trial (Rogers, Jackson, Sewell, & Harrison, 2004). The ECST-R contains “atypical presentation” (ATP) scales which screen for feigned incompetency. The ATP scale intersperses distractor questions (e.g., “Do you find it hard to cope with the overcrowding and noise in the jail?”) with questions that screen for feigned incompetency (e.g., “Have you ever felt like the court reporter is someone from your family, but in disguise?”). Studies of jail detainees and inpatient competency restoration patients provide evidence that the ECST-R can be a valuable tool for screening for feigned incompetence to stand trial (Fogel, Schiffman, Mumley, Tillbrook, & Grisso, 2013; Norton & Ryba, 2010; Vitacco, Rogers, Gabel, & Munizza, 2007). The following example provides an unusual window into the thinking of a defendant who repeatedly feigned psychosis for the purpose of avoiding trial. Mr. K was charged with aggravated robbery. Observations of Mr. K by correctional officers revealed no abnormal behavior. During his evaluation, Mr. K rocked back and forth and sang songs. He spoke rapidly and repeatedly interrupted the evaluator. He reported that he had ESP powers and was being tormented by the government as a political prisoner. He answered nearly all questions with questions. He alleged that all courtroom personnel were against him due to a government plot. When the clinician left the room, Mr. K stopped rocking and was quiet. Several letters that Mr. K had written to his girlfriend (an incarcerated codefendant) were available for review. The following excerpts advise his girlfriend on how also to malinger incompetence (K. Quinn, personal communication, June 8, 1985). 1.  “When the doctors see you, they only hold you for a little while. All the time you are with them, don’t hold a normal conversation with them. When they start asking you a question, interrupt them before they can finish asking. You can always use scriptures from the Bible to interrupt them with; make up your own scriptures, stare a lot at the floor, turn your head away from them and mumble to yourself.”

114

II .   D i a g n o s t ic I ss u e s

2. “Start talking about any- and everything. Keep changing subjects. Don’t complete sentences with them. You don’t know the judge from the bailiff or prosecutor. You don’t fully understand what you are charged with. Accuse your lawyer of being a communist. You don’t understand the regular courtroom procedures; the voices told you that the courtroom was like a circus or zoo.... Talk stupid, dumb, and crazy to even your social worker.” Several clues were observed in ascertaining that Mr. K was malingering. He overacted his part and was eager to call attention to his illness. He did not maintain his psychotic behavior for 24 hours a day. He answered many questions with “I don’t know” and refused to give details. He pretended to be both psychotic and to have low intelligence about the criminal justice system. The finding of malingering was confirmed by his letters to his girlfriend.

MALINGEREDINSANITY I am essentially not in madness, but mad in craft —Shakespeare, Hamlet (Act 3, Scene 4)

In Shakespeare’s play, Hamlet could be said to have assumed the role of a “madman.” With his simulated madness as a cover, Hamlet was at liberty to exact his revenge. Malingerers are likely to have nonpsychotic motives for their criminal behavior, such as revenge, jealousy, greed, or anger. In contrast, a crime without an apparent motive (e.g., random killing of a stranger) may lend some credence to the presence of true mental illness. Genuine psychotic explanations for rape, robbery, fraud, and check forging are extremely unusual. Warren, Murrie, Chauhan, Dietz, and Morris (2004) reviewed 5,175 insanity evaluations. Forensic evaluators opined that 12% of the defendants were insane. Opinions supporting insanity were associated with psychotic disorders, organic disorders, affective disorders, and a past history of psychiatric treatment. Opinions supporting sanity were associated with drug charges, personality disorders, intoxication at the time of the offense, and prior criminal history. In assessing defendants for criminal responsibility, clinicians must determine whether they are malingering psychosis at the time of the offense only or are continuing to malinger at the time of the examination (Hall, 1982; see Table 6.7). The

TABLE 6.7.  Malingered Psychosis DuringaCrime

A. Faking psychosis while actually committing the crime (very infrequent) B. Faking in the evaluation of “psychosis during the crime” and either: 1. Claiming to be well now 2. Still faking psychosis C. Actually psychotic during the crime, but superimposing faked exculpatory symptoms at the evaluation.

importance of the differentiation was demonstrated by Rogers et al. (1984) using the Schedule of Affective Disorders and Schizophrenia (SADS) diagnostic interview. Although the SADS summary scales successfully differentiated between sane and insane defendants at the time of their crimes, no significant differences were found at the time of their evaluations. Some malingerers mistakenly believe that they must show ongoing symptoms of psychosis in order to succeed with an insanity defense. When defendants present with current psychiatric symptoms, the clinician has the opportunity to see whether these alleged symptoms are consistent with contemporaneous psychological testing. Several clues can assist in the detection of fraudulent insanity defenses (see Table 6.8). A psychotic explanation for a crime should be ques-

TABLE 6.8.  Model Criteria fortheAssessment of Malingered Psychosis inDefendants Pleading Insanity

Malingering should be suspected if any two of the following are present: 1. A nonpsychotic, alternative motive for the crime 2. Suspect hallucinations or delusions (see Table 6.3) 3. Current offense fits a pattern of prior criminal conduct 4. Absence of negative symptoms of psychosis during evaluation 5. Report of a sudden, irresistible impulse 6. Presence of a partner in the crime 7. “Double denial” of responsibility (disavowal of crime + attribution to psychosis) 8. Alleged illness inconsistent with documented level of functioning 9. Alleged intellectual deficit coupled with alleged psychosis

6. Malingered Psychosis 115

tioned if the crime fits the same pattern as in previous criminal convictions. Gacono, Meloy, Sheppard, Speth, and Roske (1995) compared legitimate insanity acquittees with individuals who had successfully malingered insanity. Malingerers were significantly more likely to have a history of murder or rape, be diagnosed with antisocial personality disorder or sexual sadism, and produce higher PCL-R Factor 1, Factor 2, and total scores than insanity acquittees who did not malinger. Postacquittal, malingerers were also significantly more likely to be verbally or physically assaultive, have sexual relations with female staff, deal drugs, and be considered an escape risk within the forensic hospital. Malingerers may tell a far-fetched story in an attempt to “retro-fit” the facts of the crime into an insanity defense. For example, one malingerer with prior armed robbery convictions claimed that he robbed only upon command hallucinations and gave away all the stolen money to homeless people. Malingering of insanity should be suspected if a partner was involved in the crime because most accomplices of normal intelligence will not participate in a crime that is motivated by psychotic beliefs. In such cases, the clinician may assess the validity of the alleged insanity by questioning the codefendant. Thompson et al. (1992) found 98% of successful insanity acquittees in Michigan acted alone. Malingering defendants may present themselves as doubly blameless within the context of their feigned illness. For example, a male defendant pled insanity to a charge of stabbing a 7-year-old boy 60 times with an ice pick. He stated that he was sexually excited and intended to force hom*osexual acts on the victim but abandoned his plan after the boy began to cry. When he started to leave, he alleged that “10 faces in the bushes” began chanting, “Kill him, kill him, kill him.” He yelled, “No,” and repeatedly struck out at the faces with an ice pick. He alleged the next thing he knew, “the victim was covered with blood.” The defendant’s version of the offense demonstrates a double avoidance of responsibility: (1) The faces told him to kill, and (2) he claimed to have attacked the faces and not the victim. After his conviction for the offense, he confessed to six unsolved sexually sad*stic murders.

Malingering inCorrectionalSettings The rate at which inmates in correctional settings feign mental illness is unclear. Pollock et al. (1997),

using the SIRS and MMPI-2, reported a 32% rate of malingering among prisoners referred to a medium secure unit. Mentally ill inmates must sometimes exaggerate their symptoms simply to ensure that needed treatment will be provided (Kupers, 2004). Thus, commonly used detection strategies may not be reliable in the postconviction correctional setting. Indeed, Vitacco and Rogers (2005) have noted that DSM screening indices for suspected malingering do not apply in a correctional setting. An incorrect classification of malingering may have seriously detrimental and long-lasting effects for a genuinely mentally ill inmate. A label of malingering will be extremely difficult for the inmate to overcome and may subject him to a variety of adverse outcomes. In addition to denial of needed treatment, an improper classification of malingering can actually result in disciplinary actions against the inmate in many prisons. The stigma of the malingering label may encourage correctional staff to disregard all the inmate patient’s future complaints. Many correctional facilities remain dangerous and under-resourced, causing difficulty distinguishing malingering from adaptive coping strategies. Inmates with serious mental illness may exaggerate symptoms to avoid toxic and stressful environments such as punitive isolation. Some inmates may malinger to seek the relatively protected environment of a mental health unit, particularly if they are harassed by general population inmates. Certainly, not all inmate patients malinger for adaptive reasons. Some have illicit motives such as obtaining medications to abuse or sell, avoiding appropriate disciplinary actions, or gaining transfer to other living situations. Inmates nearing the end of their sentences may malinger to establish a “documented” history of mental illness in order to obtain disability benefits in the community (Knoll, 2015). The finding that an inmate patient has malingered one or more symptoms of psychosis does not rule out the presence of true mental illness. Therefore, a determination of malingering should not exclude the inmate from receiving further mental health services. Kupers (2004) and Knoll (2015) have suggested some clinical indicators for when correctional mental health staff should use caution before making a classification of malingering. Table 6.9 lists clinical factors to consider prior to a conclusion of malingering by an inmate patient.

116

II .   D i a g n o s t ic I ss u e s

TABLE 6.9.  Clinical Factors Warranting Caution before Diagnosing Malingering inaCorrectional Setting

  1.  No clear external incentive

symptoms voluntarily, while others lacked insight into their illness. Those lacking insight were older and more likely to be psychotic than intentional dissimulators.

  2.  Extensive history of psychiatric treatment  3. Not self-referred   4.  Defensive, minimizes illness, or opposes treatment   5.  Shows improvement on psychiatric medications   6.  Does well or improves in a mental health unit  7. Intellectual disability   8.  Frequently requires special observations   9.  High number of objective suicide risk factors 10.  History of serious traumatic brain injury

Defensiveness andDenial ofPsychotically MotivatedCrimes Forensic clinicians are usually trained to be vigilant for defendants who may be malingering during a forensic evaluation. However, the opposite can occur with defendants who deny psychotic symptoms or otherwise “simulate” sanity (Rogers, 2008). Defensiveness is the concealment of genuine symptoms of mental illness in an effort to portray psychological health. To successfully dissimulate a psychological disorder, the individual must possess adequate self-control to simulate a healthy state or engage in impression management (Martino et al., 2016). The denial of psychiatric symptoms has been reported anecdotally in persons who have committed crimes (Diamond, 1994). One reason that defendants may deny psychotic symptoms is avoid the stigma and consequences of being labeled with a mental illness. For some defendants with mental illness, to admit their actions “were motivated by delusions, rather than reality... is a public humiliation destructive to one’s self-esteem” (Diamond, 1994, p.166). For example, in the case of Theodore Kaczynski (i.e., the Unabomber), some experts suggested that he suffered from paranoid schizophrenia. However, he was highly averse to raising mental illness as a defense because he believed it would undermine the credibility of his antitechnology “manifesto” (Knoll, 2016). When a defendant conceals psychotic symptoms, the potential exists for a miscarriage of justice. Caruso, Benedek, Auble, and Bernet (2003) found that the minimization of symptoms could be classified as either intentional or lacking insight. Intentional dissimulators concealed their

CONCLUSIONS Identifying malingered psychosis is necessary to bring accuracy to forensic assessments and to prevent miscarriages of justice and misuse of limited healthcare resources. The detection of malingered psychosis can be difficult and requires a systematic approach. To confidently conclude that an individual is malingering psychotic symptoms, the clinician must have a detailed understanding of genuine psychotic symptoms and review data from multiple sources. The clinician must assemble clues from a thorough evaluation, clinical records, collateral data, and especially psychological testing. Although substantial effort is required, the clinician bears considerable responsibility to assist society in differentiating true psychosis from malingering.

REFERENCES Ali, S., Patel, M., Avenido, J., Jabeen, S., & Riley, W. (2011). Hallucinations: Common features and causes. Current Psychiatry, 10(11), 22–29. Alpert, M., & Silvers, K. (1970). Perceptual characteristics distinguishing auditory hallucinations in schizophrenia and acute alcoholic psychoses. American Journal of Psychiatry, 127(3), 298–302. Altshuler, L., Cummings, J., & Mills, M. (1986). Mutism: Review, differential diagnosis and report of 22 cases. American Journal of Psychiatry, 143(11), 1409– 1414. Anderson, E. W., Trethowan, W. H., & Kenna, J. (1959). An experimental investigation of simulation and pseudo-dementia. Acta Psychiatrica Neurologica Scandanavia, 34(132). Andreasen, N. C. (1987). The diagnosis of schizophrenia. Schizophrenia Bulletin, 13(1), 9–22. Andreyev, L. (1902). The dilemma. In L. Hamalian & V. Von Wiren-Garczyski (Eds.), Seven Russian short novel masterpieces. New York: Popular Library. Arguedas, D., Stevenson, R. J., & Langdon, R. (2012). Source monitoring and olfactory hallucinations in schizophrenia. Journal of Abnormal Psychology, 121(4), 936–943. Asaad, G. (1990). Hallucinations in clinical psychiatry: A guide for mental health professionals. Philadelphia: Brunner/Mazel. Assad, G., & Shapiro, B. (1986). Hallucinations: Theo-

6. Malingered Psychosis 117

retical and clinical overview. American Journal of Psychiatry, 143(9), 1088–1097. Baethge, C., Baldessarini, R. J., Freudenthal, K., Streeruwitz, A., Bauer, M., & Bschor, T. (2005). Hallucinations in bipolar disorder: Characteristics and comparison to unipolar depression and schizophrenia. Bipolar Disorders, 7(2), 136–145. Bash, I. Y., & Alpert, M. (1980). The determination of malingering. Annals of the New York Academy of Science, 347, 86–99. Bass, C., & Halligan, P. (2014). Factitious disorders and malingering: Challenges for clinical assessment and management. Lancet, 383(9926), 1422–1432. Beck, J., & Harris, M. (1994). Visual hallucinations in non-delusional elderly. International Journal of Geriatric Psychiatry, 9(7), 531–536. Benke, T. (2006). Peduncular hallucinosis: A syndrome of impaired reality monitoring. Journal of Neurology, 253(12), 1561–1571. Berney, T. (1973). A review of simulated illness. South African Medical Journal, 47(31), 1429–1434. Berrios, G. (1991). Musical hallucinations: A statistical analysis of 46 cases. Psychopathology, 24(6), 356–360. Beume, L., Klingler, A., Reinhard, M., & Niesen, W. (2015). Olfactory hallucinations as primary symptom for ischemia in the right posterior insula. Journal of the Neurological Sciences, 354(1–2), 138–139. Blanchard, D. D., McGrath, R. E., Pogge, D. L., & Khadivi, A. (2003). A comparison of the PAI and MMPI-2 as predictors of faking bad in college students. Journal of Personality Assessment, 80(2), 197–205. Braham, L. G., Trower, P., & Birchwood, M. (2004). Acting on command hallucinations and dangerous behavior: A critique of the major findings in the last decade. Clinical Psychology Review, 24(5), 513–528. Brittain, R. (1966). The history of legal medicine: The assizes of Jerusalem. Medicolegal Journal, 34, 72–73. Broughton, N., & Chesterman, P. (2001). Malingered psychosis. Journal of Forensic Psychiatry, 12(2), 407– 422. Bustamante, J., & Ford, C. (1977). Ganser’s syndrome. Psychiatric Opinion, 14, 39–41. Carter, D. M., Mackinnon, A., & Copoloy, D. (1996). Patients’ strategies for coping with auditory hallucinations. Journal of Nervous and Mental Disease, 184(3), 159–164. Caruso, K. A., Benedek, D. M., Auble, P. M., & Bernet, W. (2003). Concealment of psychopathology in forensic evaluations: A pilot study of intentional and uninsightful dissimulators. Journal of the American Academy of Psychiatry and the Law Online, 31(4), 444–450. Chadwick, P., & Birchwood, M. (1994). The omnipotence of voices: A cognitive approach to auditory hallucinations. British Journal of Psychiatry, 164(2), 190–201. Cohen, M., Adler, A., Alfonso, C. A., & Haque, M. M. (1994). Lilliputian hallucinations and medical illness. General Hospital Psychiatry, 16(2), 141–143.

Collinson, G. (1812). A treatise on the law concerning idiots, lunatics, and other persons non compos mentis. London: Reed. Contardi, S., Rubboli, G., Giulioni, M., Michelucci, R., Pizza, F., Gardella, E., et al. (2007). Charles Bonnet syndrome in hemianopia, following antero-mesial temporal lobectomy for drug-resistant epilepsy. Epileptic Disorders, 9(3), 271–275. Copoloy, D., Trauer, T., & Mackinnon, A. (2004). On the non-significance of internal versus external auditory hallucinations. Schizophrenia Research, 69(1), 1–6. Cornelius, J., Mezzich, J., Fabrega, H., Cornelius, M., Myer, J., & Ulrich, R. (1991). Characterizing organic hallucinosis. Comprehensive Psychiatry, 32(4), 338–344. Cornell, D. G., & Hawk, G. L. (1989). Clinical presentation of malingerers diagnosed by experienced forensic psychologists. Law and Human Behavior, 13(4), 375–383. Cummings, J. L., & Mega, M. S. (2003). Hallucinations. In J. L. Cummings & M. S. Mega (Eds.), Neuropsychiatry and behavioral neuroscience (pp.187–199). New York: Oxford University Press. Cummings, J. L., & Miller, B. L. (1987). Visual hallucinations: Clinical occurrence and use in differential diagnosis. Western Journal of Medicine, 146(1), 46–51. Daniel, A. E., & Resnick, P. J. (1987). Mutism, malingering, and competency to stand trial. Bulletin of the American Academy of Psychiatry and the Law, 15(3), 301–308. Davidson, H. (1952). Malingered psychosis. Bulletin of the Menninger Clinic, 14, 157–163. de Dios Francos, A., Sanz Granado, O., & de la Gándara Martín, J. J. (1995). Síndrome de Ganser. Psiq Biol, 2(4), 151–156. Demakis, G., Rimland, C., Reeve, C., & Ward, J. (2015). Intelligence and psychopathy do not influence malingering. Applied Neuropsychology: Adult, 22(4), 262–270. Diamond, B. (1994). The psychiatrist in the courtroom: Selected papers of Bernard L. Diamond, M.D. Hillsdale, NJ: Analytic Press. Drew, C. A. (1908). An insane (?) malingerer. American Journal of Psychiatry, 64(4), 669–684. Drob, S., Meehan, K., & Waxman, S. (2009). Clinical and conceptual problems in the attribution of malingering in forensic evaluations. Journal of the American of Psychiatry and the Law, 37(1), 98–106. Duncan, J. (1995). Medication compliance in schizophrenia. Unpublished doctoral dissertation, University of North Texas, Denton, TX. Edens, J. F., Buffington, J. K., & Tomicic, T. L. (2000). An investigation of the relationship between psychopathic traits and malingering on the psychopathic personality inventory. Assessment 2000, 7(3), 281– 296. Edens, J. F., Guy, L. S., Otto, R. K., Buffington, J. K., Tomicic, T. L., & Poythress, N. G. (2001). Factors dif-

118

II .   D i a g n o s t ic I ss u e s

ferentiating successful versus unsuccessful malingerers. Journal of Personality Assessment, 77(2), 333–338. Eissler, K. (1951). Malingering. In G. B. Wilbur, & W. Muensterberger (Eds.), Psychoanalysis and culture (pp.218–253). New York: International Universities Press. Ellinwood, E. (1972). Amphetamine psychosis: Individuals, settings, and sequences. In E. H. Ellinwood, Jr., & S. Cohen (Eds.), Current concepts on amphetamine abuse (pp.145–157). Rockville, MD: National Institute of Mental Health. Evyapan, A. (2006). Pure mutism due to simultaneous bilateral lenticulostriate artery territory infarction. CNS Spectrums, 11(4), 257–259. Falloon, I. R. H., & Talbot, R. E. (1981). Persistent auditory hallucinations: Coping mechanisms and implications for management. Psychological Medicine, 11(2), 329–339. Ferchiou, A., Schürhoff, F., Bulzacka, E., Mahbouli, M., Leboyer, M., & Szöke, A. (2010). Source monitoring: General presentation and review of literature in schizophrenia. Encephale, 36(4), 326–333. Fischer, C., Marchie, A., & Norris, M. (2004). Musical and auditory hallucinations: A spectrum. Psychiatry and Clinical Neurosciences, 58(1), 96–98. Fogel, M. H., Schiffman, W., Mumley, D., Tillbrook, C., & Grisso, T. (2013). Ten year research update (2001–2010): Evaluations for competence to stand trial (adjudicative competence). Behavioral Sciences & the Law, 31(2), 165–191. Folks, D., & Freeman, A. (1985). Munchausen’s syndrome and other factitious illness. Psychiatric Clinics of North America, 8, 263–278. Folstein, M., Folstein, S., & McHugh, P. (1975). Minimental state: A practical method of grading cognitive state of patients for the clinician. Journal of Psychiatric Research, 12(3), 189–198. Fox, J. R. E., Gray, N. S., & Lewis, H. (2004). Factors determining compliance with command hallucinations with violent content: The role of social rank, perceived power of the voice and voice malevolence Journal of Forensic Psychiatry and Psychology, 15(3), 511–531. Gacono, C. B., Meloy, J. R., Sheppard, K., Speth, E., & Roske, A. (1995). A clinical investigation of malingering and psychopathy in hospitalized insanity acquittees. Bulletin of the American Academy of Psychiatry and the Law, 23(3), 387–397. Gaunlett-Gilbert, J., & Kuipers, E. (2003). Phenomenology of visual hallucinations in psychiatric conditions. The Journal of Nervous and Mental Disease, 191(3), 203–205. Glancy, G., Ash, P., Bath, E. P., Buchanan, A., Fedoroff, P., Frierson, R. L., et al. (2015). AAPL Practice Guideline for the Forensic Assessment. Journal of the American Academy of Psychiatry and the Law Online, 43(2), S3–S53. Goodwin, D. W., Alderson, P., & Rosenthal, R. (1971). Clinical significance of hallucinations in psychiatric

disorders: A study of ll6 hallucinatory patients. Archives of General Psychiatry, 24(1), 76–80. Gottfried, E., Hudson, B., Vitacco, M., & Carbonell, J. (2017). Improving the detection of feigned knowledge deficits in defendants adjudicated incompetent to stand trial. Assessment, 24(2), 232–243. Green, C., Garety, P., Freeman, D., Fowler, D., Bebbington, P., Dunn, G., et al. (2006). Content and affect in persecutory delusions. British Journal of Clinical Psychology, 45(4), 561–577. Greenfield, D. (1987). Feigned psychosis in a 14-yearoldgirl. Hospital and Community Psychiatry, 38(1), 73–75. Gunduz-Bruce, H., McMeniman, M., Robinson, D. G., Woerner, M. G., Kane, J. M., Schooler, N. R., et al. (2005). Duration of untreated psychosis and time to treatment response for delusions and hallucinations. American Journal of Psychiatry, 162(10), 1966–1969. Hall, H. (1982). Dangerous predictions and the maligned forensic professional. Criminal Justice and Behavior, 9, 3–12. Harrow, M., Herbener, E. S., Shanklin, A., Jobe, T. H., Rattenbury, F., & Kaplan, K. J. (2004). Follow-up of psychotic outpatients: Dimensions of delusions and work functioning in schizophrenia. Schizophrenia Bulletin, 30(1), 147–161. Harrow, M., O’Connell, E. M., Herbener, E. S., Altman, A. M., Kaplan, K. J., & Jobe, T. H. (2003). Disordered verbalizations in schizophrenia: A speech disturbance or thought disorder? Comprehensive Psychiatry, 44(5), 353–359. Hay, G. (1983). Feigned psychosis: A review of the simulation of mental illness. British Journal of Psychiatry, 143, 8–10. Hellerstein, D., Frosch, W., & Koenigsberg, H. (1987). The clinical significance of command hallucinations. American Journal of Psychiatry, 144(2), 219–221. Hoffman, R., Varanko, M., Gilmore, J., & Mishara, A. (2008). Experiential features used by patients with schizophrenia to differentiate ‘voices’ from ordinary verbal thought. Psychological Medicine, 38, 1167–1176. Jaffe, M. E., & Sharma, K. K. (1998). Malingering uncommon psychiatric symptoms among defendants charged under California’s “three strikes and you’re out” law. Journal of Forensic Science, 43(3), 549–555. Jiménez Gómez, B., & Quintero, J. (2012). Ganser syndrome: Review and case report. Actas Españolas de Psiquiatría, 40(3), 161–164. Jones, A., & Llewellyn, J. (1917). Malingering. London: Heinemann. Jordaan, G., Nel, D., Hewlett, R., & Emsley, R. (2009). Alcohol-induced psychotic disorder: A comparative study on the clinical characteristics of patients with alcohol dependence and schizophrenia. Journal of Studies on Alcohol and Drugs, 70(6), 870–876. Junginger, J. (1990). Predicting compliance with command hallucinations. American Journal of Psychiatry 147(2), 245–247. Junginger, J. (1995). Command hallucinations and the

6. Malingered Psychosis 119

prediction of dangerousness. Psychiatric Services, 46, 911–914. Junginger, J., & Frame, C. (1985). Self-report of the frequency and phenomenology of verbal hallucinations. Journal of Nervous and Mental Disease, 173(3), 149–155. Kalambokis, G., Konitsiotis, S., Pappas, D., & Tsianos, E. V. (2006). Akinetic mutism followed by a manic reaction on introduction of steroid replacement for Addison’s disease. Journal of Endocrinological Investigation, 29(3), 257–260. Kanaan, R., & Wessely, S. (2010). The origins of factitious disorder. History of the Human Sciences, 23(2), 68–85. Kanas, N., & Barr, M. A. (1984). Letter to the editor: Self-control of psychotic productions in schizophrenics. Archives of General Psychiatry, 41(9), 919–920. Kasper, E., Rogers, R., & Adams, P. (1996). Dangerousness and command hallucinations: An investigation of psychotic inpatients. Bulletin of the American Academy of Psychiatry and the Law, 24, 219–224. Kent, G., & Wahass, S. (1996). The content and characteristics of auditory hallucinations in Saudi Arabia and the UK: A cross-cultural comparison. Acta Psychiatra Scandinavica, 94(6), 433–437. Knoll, J. L. (2015). Evaluation of malingering in corrections. In R. L. Trestman, K. L. Appelbaum, & J. L. Metzner (Eds.), The Oxford textbook of correctional psychiatry (pp.117–122). New York: Oxford University Press. Knoll, J. L. (2016). Every last tie: The story of the unabomber and his family. Durham, NC: Duke University Press. Knowles, R., McCarthy-Jones, S., & Rowse, G., (2011). Grandiose delusions: A review and theoretical integration of cognitive and affective perspectives. Clinical Psychology Review, 31(4), 684–696. Kraus, A. (1994). Phenomenology of the technical delusion in schizophrenia. Journal of Phenomenological Psychology, 25, 51–69. Kropp, P. R., & Rogers, R. (1993). Understanding malingering: Motivation, method, and detection. In M. Lewis & C. Saarini (Eds), Lying and deception (pp.201–216). New York: Guilford Press. Kucharski, L. T., Ryan, W., Vogt, J., & Goodloe, E. (1998). Clinical symptom presentation in suspected malingerers: An empirical investigation. Journal of the American Academy of Psychiatry and the Law, 26(4), 579–585. Kupers, T. (2004). Malingering in correctional settings. Correctional Mental Health Report, 5(6), 81. Langdon, R., McGuire, J., Stevenson, R., & Catts, S. V. (2011). Clinical correlates of olfactory hallucinations in schizophrenia. British Journal of Clinical Psychology, 50(2), 145–163. Larøi, F. (2012). How do auditory verbal hallucinations in patients differ from those in non-patients? Frontiers in Human Neuroscience, 6(25). Leudar, I., Thomas, P., McNally, D., & Glinski, A.

(1997). What voices can do with words: Pragmatics of verbal hallucinations. Psychological Medicine, 27(4), 885–898. Lewinsohn, P. M. (1970). An empirical test of several popular notions about hallucinations in schizophrenic patients. In W. Keup (Ed), Origin and mechanisms of hallucinations (pp.401–404). New York: Springer. Lewis, D. J. (1961). Lilliputian hallucinations in the functional psychoses. Canadian Psychiatric Association Journal, 6, 177–201. Manford, M., & Andermann, F. (1998). Complex visual hallucinations: Clinical and neurobiological insights. Brain: A Journal of Neurology, 121(10), 1819–1840. Martino, V., Grattagliano, I., Bosco, A., Massaro, Y., Lisi, A., Campobasso, F., et al. (2016). A new index for the MMPI-2 test for detecting dissimulation in forensic evaluations: A pilot study. Journal of Forensic Sciences, 61(1), 249–253. McCarthy-Jones, S., & Resnick, P. (2014). Listening to voices: The use of phenomenology to differentiate malingered from genuine auditory verbal hallucinations. International Journal of Law and Psychiatry, 37(2), 183–189. McCarthy-Jones, S., Trauer, T., Mackinnon, A., Sims, E., Thomas, N., & Copolov, D. (2014). A new phenomenological survey of auditory hallucinations: Evidence for subtypes and implications for theory and practice. Schizophrenia Bulletin, 40(1), 231–235. McDermott, B. E. (2012). Psychological testing and the assessment of malingering. Psychiatric Clinics of North America, 35(4), 855–876. McDermott, B., Dualan, I., & Scott, C. (2013). Malingering in the correctional system: Does incentive affect prevalence? International Journal of Law and Psychiatry, 36(3–4), 287–292. Miller, H. A. (2001). M-Fast Interview booklet. Lutz, FL: Psychological Assessment Resources. Miller, L., O’Connor, E., & DiPasquale, T. (1993). Patients’ attitudes toward hallucinations. American Journal of Psychiatry, 150, 584–588. Mitchell, J., & Vierkant, A. D. (1991). Delusions and hallucinations of cocaine abusers and paranoid schizophrenics: A comparative study. Journal of Psychology, 125(3), 301–310. Mizrahi, R., & Starkstein, S. E. (2006). Phenomenology and clinical correlates of delusions in Alzheimer disease. American Journal of Geriatric Psychiatry, 14(7), 573–581. Mott, R. H., Small, I. F., & Andersen, J. M. (1965). Comparative study of hallucinations. Archives of General Psychiatry, 12, 595–601. Nayani, T. H., & David, A. S. (1996). The auditory hallucination: A phenomenological survey. Psychological Medicine, 26(1), 177–189. Newman, A. (2003, April). Analyze This: Vincent Gigante, Not Crazy After All Those Years. New York Times. Norton, K., & Ryba, N. (2010). An investigation of the

120

I I .   D i a g n o s t ic I ss u e s

ECST-R as a measure of competence and feigning. Journal of Forensic Psychology Practice, 10(2), 91–106. Ossipov, V. P. (1944). Malingering: The simulation of psychosis. Bulletin of the Menninger Clinic, 8, 31–42. Oulis, P. G., Mavreas, V. G., Mamounas, J. M., & Stefanis, C. N. (1995). Clinical characteristics of auditory hallucinations. Acta Psychiatrica Scandanavia, 92(2), 97–102. Pearlson, G., Kreger, L., Rabins, R., Chase, G., Cohen, B., & Wirth, J., et al. (1989). A chart review study of late-onset and early-onset schizophrenia. American Journal of Psychiatry, 146(12), 1568–1574. Perala, J., Puoppasalmi, K., Pirkola, S., Härkänen, T., Saami, S., Tuulio-Henriksson, A., et al. (2010). Alcohol-induced psychotic disorder and delirium in the general population. British Journal of Psychiatry, 197(3), 200–206. Pollock, P., Quigley, B., Worley, K., & Bashford, C. (1997). Feigned mental disorder in prisoners referred to forensic mental health services. Journal of Psychiatric and Mental Health Nursing, 4(1), 9–15. Pope, H., Jonas, J., & Jones, B. (1982). Factitious psychosis: Phenomenology, family history, and longterm outcome of nine patients. American Journal of Psychiatry, 139(11), 1480–1483. Powell, K. E. (1991). The malingering of schizophrenia. Unpublished doctoral dissertation, University of South Carolina, Columbia, SC. Poythress, N. G., Edens, J. F., & Watkins, M. M. (2001). The relationship between psychopathic personality features and malingering symptoms of major mental illness. Law and Human Behavior, 25(6), 567–582. Resnick, P. (1984). The detection of malingered mental illness. Behavioral Science and the Law, 2(1), 20–38. Resnick, P. (1997). Malingering of posttraumatic stress disorders. In R. Rogers (Ed.), Clinical assessment of malingering and deception, 2nd edition (pp.130–152). New York: Guilford Press. Ritson, B., & Forest, A. (1970). The simulation of psychosis: A contemporary presentation. British Journal of Psychology, 43(1), 31–37. Rogers, R. (1986). Conducting insanity evaluations. New York: Van Nostrand Reinhold. Rogers, R. (1987). Assessment of malingering within a forensic context. In D. W. Weisstub (Ed.), Law and psychiatry: International perspectives (3rd ed., pp.209– 238). New York: Plenum Press. Rogers, R. (1990). Development of a new classification of malingering. American Academy of Psychiatry and the Law, 18(3), 323–333. Rogers, R. (2008). Current status of clinical methods. In R. Rogers (Ed.), Clinical assessment of malingering and deception (3rd ed., pp.391–410). New York: Guilford Press. Rogers, R., Bagby, R. M., & Dickens, S. E. (1992). Structured Interview of Reported Symptoms. Lutz, FL: Psychological Assessment Resources. Rogers, R., & Bender, S. (2013). Evaluation of malinger-

ing and deception. In R. K. Otto (Ed.), Comprehensive handbook of psychology: Forensic psychology (2nd ed., Vol. 11, pp.517–539). New York: Wiley. Rogers, R., & Cruise, K. (2000). Malingering and deception among psychopaths. In C. B. Gacono (Ed.), The clinical and forensic assessment of psychopathy: A practitioner’s guide (pp.269–284). Mahwah, NJ: Erlbaum. Rogers, R., Jackson, R., Sewell, K., & Harrison, K. (2004). An examination of the ECST-R as a screen for feigned incompetency to stand trial. Psychological Assessment, 16(2), 139–145. Rogers, R., Sewell, K. W., & Gillard, N. D. (2010). SIRS2: Structured Interview of Reported Symptoms: Professional manual. Odessa, FL: PsychologicalAssessment Resources. Rogers, R., Sewell, K., & Goldstein, A. M. (1994). Explanatory models of malingering: A prototypical analysis. Law and Human Behavior, 18(5), 543–552. Rogers, R., Vitacco, M., & Kurus, S. (2010). Assessment of malingering with repeat forensic evaluations: Patient variability and possible misclassification on the SIRS and other feigning measures. Journal of the American Academy of Psychiatry and the Law, 38(1), 109–14. Rosenberg, C. (1968). The trial of the assassin of Guiteau. Oxford, UK: University of Chicago Press. Rosenhan, D. (1973). On being sane in insane places. Science, 172(4070), 250–258. Rowan, C. (1982, June 21). Cleveland Plain Dealer, p.10B. Sachs, M. H., Carpenter, W. T., & Strauss, J. S. (1974). Recovery from delusions. Archives of General Psychiatry, 30(1), 117–120. Samenow, S. (1984). Inside the criminal mind. New York: Times Books. Schneck, J. (1970). Pseudo-malingering and Leonid Andreyev’s the dilemma. Psychiatric Quarterly, 44(1), 49–54. Schneider, S. D., Jelinek, L., Lincoln, T. M., & Moritz, S. (2011). What happened to the voices?: A finegrained analysis of how hallucinations and delusions change under psychiatric treatment. Psychiatry Research, 188(1), 13–17. Schretlen, D. J. (1988). The use of psychological tests to identify malingered symptoms of mental disorder. Clinical Psychology Review, 8(5), 451–476. Seigel, R., & West, L. (Eds.). (1975). Hallucinations: Behavior, experience and theory. New York: Wiley. Shawyer, F., Mackinnon, A., Farhall, J., Sims, E., Blaney, S., Yardley, P., et al. (2008). Acting on harmful command hallucinations in psychotic disorders: An integrative approach. Journal of Nervous and Mental Disease, 196(5), 390–398. Sherman, M., Trief, P., Sprafkin, R. (1975). Impression management in the psychiatric interview: Quality, style and individual differences. Journal of Consulting and Clinical Psychology, 43(6), 867–871.

6.  Malingered Psychosis 121

Small, I. F., Small, J. G., & Andersen, J. M. (1966). ­Clinical characteristics of hallucinations of schizophrenia. Diseases of the Nervous System, 27(5), 349– 353. Smith, G. P., & Burger, G. O. (1997). Detection of malingering: Validation of the structured inventory of malingered symptomatology. Journal of the American Academy of Psychiatry and the Law, 25(2), 183–189. Soliman, S., & Resnick, P. J. (2010). Feigning in adjudicative competence evaluations. Behavioral Sciences and the Law, 28(5), 614–629. Sommer, I. E., Daalman, K., Rietkerk, T., Diederen, K. M., Bakker, S., Wijkstra, J., et al. (2010). Healthy individuals with auditory verbal hallucinations; who are they?: Psychiatric assessments of a selected sample of 103 subjects. Schizophrenia Bulletin, 36(3), 633–641. Spitzer, M. (1992). The phenomenology of delusions. Psychiatric Annals, 22(5), 252–259. Stephane, M., Pellizzer, G., Roberts, S., & McCalannahan, K. (2006). The computerized binary Scale of Auditory Speech Hallucinations (cbSASH). Schizophrenia Research, 88(1–3), 73–81. Thakur, M., Hays J., Ranga, K., & Krishnan, R. (1999). Clinical, demographic and social characteristics of psychotic depression. Psychiatry Research, 86(2), 99–106. Thompson, J. S., Stuart, G. L., & Holden, C. E. (1992). Command hallucinations and legal insanity. Forensic Reports, 5(1), 29–42. Thompson, J. W., LeBourgeois, H. W., & Black, F. W. (2004). Malingering. In R. Simon & L. Gold (Eds.), Textbook of forensic psychiatry (pp.427–448). Arlington, VA: American Psychiatric Publishing. Trabucchi, M., & Bianchetti, A. (1997). Delusions. International Psychogeriatrics, 8(S3), 383–385. Travin, S., & Protter, B. (1984). Malingering and malingering-like behavior: Some clinical and conceptual issues. Psychiatric Quarterly, 56(3), 189–197.

Vitacco, M. J., & Rogers, R. R. (2005). Malingering in corrections. In C. Scott (Ed.), Handbook of correctional mental health. Washington, DC: American Psychiatric Publishing. Vitacco, M. J., Rogers, R., Gabel, J., & Munizza, J. (2007). An evaluation of malingering screens with competency to stand trial patients: A known-groups comparison. Law and Human Behavior, 31(3), 249– 260. Wachpress, M., Berenberg, A. N., & Jacobson, A. (1953). Simulation of psychosis. Psychiatric Quarterly, 27, 463–473. Warren, J., Murrie, D., Chauhan, P., Dietz, P., & Morris, J. (2004). Opinion formation in evaluating sanity at the time of the offense: An examination of 5,175 pre-trial evaluations. Behavioral Sciences and the Law, 22(2), 171–186. Warren, J., Murrie, D., Stejskal, W., Colwell, L., Morris, J., Chauhan, P., et al. (2006). Opinion formation in evaluating the adjudicatory competence and restorability of criminal defendants: A review of 8,000 evaluations. Behavioral Sciences and the Law, 24(2), 113–132. Wertham, F. (1949). The show of violence. Garden City, NY: Doubleday. Wessely, S., Buchanan, A., Reed, A., Cutting, J., Everitt, B., Garety, P., et al. (1993). Acting on delusions: I. Prevalence. British Journal of Psychiatry, 163, 69–76. Wooley, C., & Rogers, R. (2014). The effectiveness of the Personality Assessment Inventory with feigned PTSD: An initial investigation of Resnick’s model of malingering. Assessment, 22(4), 449–458. Yates, B. D., Nordquist, C. R., & Schultz-Ross, R. A. (1996). Feigned psychiatric symptoms in the emergency room. Psychiatric Services, 47(9), 998–1000. Zapf, P. A., & Grisso, T. (2012). Use and misuse of forensic assessment instruments. In D. Faust (Ed.), Coping with psychiatric and psychological testimony (6th ed., pp.488–510). New York: Oxford University Press.

CHAPTER 7

Malingered Traumatic Brain Injury ScottD.Bender,PhD

Within the fields of psychological and neuropsychological assessment, research on malingering continues to outpace most other research topics. As noted in the equivalent chapter in the previous edition, neuropsychology saw a threefold increase in the number of publications on malingering between 1990 and 2000 (Sweet, King, Malina, Bergman, & Simmons, 2002). And more recent survey data (Martin, Shroeder, & Odland, 2015) indicate that almost 25% of all articles published in The Clinical Neuropsychologist and Archives of Clinical Neuropsychology between 2009 and 2014 involved effort and malingering. This equates to approximately 1,400 articles. Also, though precise numbers continue to be difficult to calculate, the vast majority of forensic cases seem to involve issues of contested mild traumatic brain injury (mTBI), a major focus of malingering research. It is now generally accepted that all forensic neuropsychological assessments should include methods for the detection of malingering (American Academy of Clinical Neuropsychology [AACN] Consensus Conference Statement; Heilbronner et al., 2009). In fact, the onus has shifted to justifying why effort testing was not conducted. Despite this increase in awareness, very little attention has been given to the precise meanings and causes of poor effort and/or symptom exaggeration. As I discuss in this chapter, this imbalance of attention and awareness has potentially far-reaching (and problematic) implications for forensic practice.

Forensic neuropsychologists have an increased awareness that a broad array of dissimulated behaviors may masquerade as neurocognitive dysfunction. In addition, symptom overlap is commonly observed among healthy populations, certain mental disorders, and mTBI. This lack of symptom specificity poses a major challenge for clinicians asked to render differential diagnoses involving brain injuries, particularly when faced with the rigorous standards of legal admissibility. In this chapter I describe genuine traumatic brain injury (TBI) subtypes and their typical courses, and also review malingering, both practically and conceptually. It is important to note at the outset that malingering is not the same construct as other response-style constructs such as deception, feigning, faking, incomplete/suboptimal/poor effort, uncooperativeness, exaggeration, or dissimulation. Most cognitive tests of “malingering” are actually measures of effort, whereas most psychiatric “malingering” tests measure symptom fabrication or gross exaggeration. This chapter provides detection strategies for the assessment of malingering. Importantly, poor effort is neither necessary nor sufficient for a determination of malingering in TBI cases. Rather, a causal relationship between poor effort (and/or exaggeration) and deliberate intention in the context of a substantial external incentive must be established. As already alluded to, this apparently simple task is actually often very challenging, primarily be-

122

7. Malingered Traumatic Brain Injury 123

cause motivation—central to the concept of malingering—cannot be directly measured. Excellent reviews of effort testing in neuropsychological assessment may be found in Carone and Bush (2013), Larrabee (2012), and Nicholson and Martelli (2007). And while many of the concepts and methods may apply to pediatric neuropsychology, this chapter is limited to adults (see Kirkwood, 2015, for a review of validity testing in children). The main topics of this chapter include (1) strategies for detecting malingered TBI; (2) the special problem of postconcussional syndrome; and (3) common pitfalls in differential diagnosis, especially in the forensic setting. Due to its prevalence in forensic settings, special emphasis is placed on mild TBI. The chapter ends with an illustrative case study.

MALINGERING DEFINITION ANDSUBTYPES DSM-5 (American Psychiatric Association, 2013) conceptualizes malingering in terms of feigned presentations and external motivation. It is not a diagnosis, but a classification of behavior, hence its status as a V-code (V65.2). According to seminal work by Lipman (1962), four types of malingering exist: 1. Invention: No genuine symptoms present; patient fabricates symptoms. 2. Perseveration: Genuine symptoms existed but have resolved; the patient alleges their continuance. 3. Exaggeration: Genuine symptoms currently exist but have been magnified beyond their true severity. 4. Transference: Genuine symptoms exist but are not related to the injury in question. The first type of malingering is the only one that involves outright fabrication of symptoms, and it is the rarest form. The other three types involve conscious manipulation of genuine symptoms and probably rightly illustrate the often nondichotomous nature of malingering—which is to say, a finding of malingering does not mean that there are no bona fide symptoms present.

Prevalence The true prevalence of malingering remains unknown, because successful malingerers go undetected and uncounted. Past surveys of forensic

experts suggested that feigning occurred in 7–17% of mental health assessments, with the higher estimate referring to forensic cases (Cornell & Hawk, 1989; Rogers, Salekin, Sewell, Goldstein, & Leonard, 1998; Rogers, Sewell, & Goldstein, 1994). In contrast to these older data, more recent surveys suggest that the rate is much higher, at least for mTBI cases, and approaches 40% when the category is broadened to include symptom exaggeration (e.g., Mittenberg, Patton, Canyock, & Condit, 2002). Though consistent with Larrabee’s (2003a) review of 11 neuropsychological studies involving mTBI, potential problems with these estimates exist, as discussed by Young (2015) and Rogers, Bender, and Johnson (2011a). For instance, many studies established the prevalence of malingering by calculating the percentage of patients who simply performed below generally accepted cutoff scores on tests of feigning in the presence of an incentive; very few studies have proven that the individuals in question were actually motivated by the external gain. Recent research by Ruff, Klopfer, and Blank (2016) suggests that estimates of malingering prevalence greatly depend on the operational definition of malingering being used. Their estimates are much more in line with older estimates of prevalence. This problem is not specific to these studies. The inability to know someone else’s true motivation(s) lies at the heart of the problem of malingering detection. As a result, prevalence rates of “malingering” may better be thought of as the prevalence of “feigning.” Accurate classification of malingering is further compromised by the multiple conditions that can resemble malingering, such as factitious disorder, conversion disorder, anxiety, posttraumatic stress disorder (PTSD), depression, and psychosocial factors involving exaggeration (Bender & Matusewicz, 2013). Given these problems, some researchers (e.g., Boone, 2007; Drob, Meehan, & Waxman, 2009) have argued that psychologists should not classify malingering at all, but just report on the validity (or lack thereof) of test results.

Diagnostic Criteria ofMalingered NeurocognitiveDeficits Slick, Sherman, and Iverson (1999) provided the first comprehensive diagnostic criteria for possible, probable, and definite malingering of cognitive dysfunction. Their publication helped standardize malingering detection in neurocognitive assessment and spurred malingering research dramatically.

124

I I .   D i a g n o s t ic I ss u e s

The Slick and Sherman (2013) update to the malingered neurocognitive dysfunction (MND) criteria addresses several concerns raised by researchers and clinicians in the years since 1999. For example, cogniform disorder (Delis & Wetter, 2007) and the concept of secondary malingering are now considered within the criteria. The Slick and Sherman (2013) classification of definite malingering still requires clear and compelling evidence of volitionally exaggerated or fabricated cognitive dysfunction in the absence of a plausible alternative explanation. Clinical diagnoses that might explain response patterns must be ruled out, and the behavior should not be the result of diminished mental capacity. Probable malingering requires similar features (e.g., volitional production of symptoms) but with comparatively less evidence (e.g., less than clear and compelling but still strongly suggestive evidence). As mentioned, the Slick and Sherman (2013) classification system has produced much needed research, but especially for forensic practice, several problematic issues remain (e.g., the criteria for definite malingering may be too narrow). For the interested reader, Bender and Frederick (Chapter 3, this volume) provide a comprehensive review and critique of the updated Slick criteria (see also Larrabee, Greiffenstein, Greve, & Bianchini, 2007; Rogers & Bender, 2013; and Rogers et al., 2011a). It is almost always easier to detect dissimulation than it is to determine its underlying motivation. No single score (or set of scores) can be considered “diagnostic” of malingering. A determination of malingering must include careful examination of contextual (i.e., nontest) factors. Nevertheless, certain scores and patterns are highly indicative of feigned or purposely biased performance. Importantly, the task of determining whether a certain test performance represents an intentional attempt to obtain material gain requires much more consideration than simply deriving a score within the “malingering range.” For example, the nature and possible meaning of the incentive to the patient must be considered. But before this topic or a full discussion of malingered symptoms of mTBI can be undertaken, a review of the characteristics of genuine TBI is needed.

TRAUMATIC BRAININJURY Based on hospital and emergency department records, the Centers for Disease Control and Pre-

vention (CDC) reported that 2.5 million cases of suspected TBI occurred in the United States in 2010 (Faul, Xu, Wald, & Coronado, 2010). This estimate does not include data from outpatient facilities, the U.S. military, or from those who did not seek medical care. As a result, it likely does not reflect the true burden of TBI in the United States. So, though at least 80% of TBIs are classified as “mild,” this number probably represents a substantial underestimate due to many unreported mTBI cases. The recent increase in public awareness of TBI may result in better prevalence estimates derived from hospitals and in forensic neuropsychological practice, where mTBI already represents the majority of cases and poses significant diagnostic challenges.

mTBI andConcussion The terms concussion and mTBI are often used interchangeably, and indeed, refer to the same general neuropathology. However, the diagnosis of mTBI remains muddy and controversial. One reason is that most definitions of concussion require only that an alteration of consciousness occur in order to meet its diagnostic criterion. Such a subjective phenomenon, often with no other party present to observe associated clinical signs, makes it difficult to establish whether this fundamental indicator of concussion even transpired. Likewise, alterations of consciousness and/or amnesia may actually be longer than those reported by the patient, leading to a misinformed impression of a milder injury. In short, the reliability of self-report in these situations is notoriously poor, which seriously complicates diagnosis of mTBI. Diagnostic agreement has been compromised further by a history of overly complex rating systems, with at least 27 gradation systems for concussion (Bender, Barth, & Irby, 2004). Definitions

In general terms, TBI is defined by the presence of (1) trauma to the brain either through blunt force to the skull or by acceleration–deceleration forces and (2) resultant signs and symptoms. In contrast to moderate and severe TBI, mTBI is usually undetectable on traditional neuroimaging. In fact, structural evidence of brain injury is usually viewed as an exclusion criterion, or as a reason to subclassify the injury (i.e., “complicated mTBI”). Despite the lack of evidence on computerized tomography (CT) and magnetic resonance imaging

7. Malingered Traumatic Brain Injury 125

(MRI), studies have shown that diffuse physiological and chemical changes do occur with mTBI (e.g., Giza & Hovda, 2004; see Lezak, Howieson, Bigler, & Tranel, 2012). Diffuse axonal injury via shear strain effects has also been documented (Bigler, 2008; Bigler & Maxwell, 2012). However, clinical evidence of mTBI sequelae can be wideranging and variable, often with a lack of correlation between imaging and clinical signs. Notably, the use of diffusion tensor imaging (DTI) in research has grown rapidly since the 2008 version of this chapter in the previous edition, and in recent years, DTI studies have targeted mTBI (e.g., Chong & Schwedt, 2015; Meier et al., 2015). DTI, which is sensitive to white-matter damage, can detect small changes in white-matter integrity via fractional anisotropy. However, so far, experts in neuropsychology and neuroradiology (Delouche et al., 2016; Trafimow, 2014; Wortzel, Tsiouris, & Filippi, 2014) concur that DTI lacks specificity in mTBI and might even produce misleading results. The best methods for operationalizing the signs and symptoms of mTBI have been debated, but some consensus has been achieved. For example, mTBI is typically said to have occurred when one or more of the following is present: A score ≥ 13 on the Glasgow Coma Scale (GCS; Jennett, 2002), posttraumatic amnesia (PTA) of under 24 hours, and loss of consciousness (LOC) of less than 30 minutes (as delineated in the American Congress of Rehabilitation Medicine criteria [ACRM; 1993]). Of course, the definition requires that these features occur in the context of brain trauma. Though somewhat dated, most clinicians and researchers agree that the ACRM criteria for mTBI are still the standard both for research and practice (see Table 7.1 for ACRM criteria for mTBI). Natural Course ofmTBI

Early estimates suggested that approximately 85– 90% of patients with mTBI recover and are asymptomatic by about 3 months postinjury (Alexander, 1995; Binder, 1997). However, these estimates are based on cases that were seen in the health care system and do not include individuals whose symptoms evidently did not require assessment or treatment. Therefore, the percentage of patients with mTBI who recover completely is likely higher than originally thought. The symptoms that occur within 3 months after an mTBI have been well-documented, but the symptoms are heterogeneous; thus, no single pro-

TABLE 7.1.  American College of Rehabilitation Medicine’s Definition of mTBI

A patient with mTBI is a person who has had a traumatically induced physiological disruption of brain function, as manifested by at least one of the following: 1. Any period of loss of consciousness 2. Any loss of memory for events immediately before or after the accident 3. Any alteration in mental state at the time of the accident (e.g., feeling dazed, disoriented, or confused) 4. Focal neurological deficit(s) that may or may not be transient, in which the severity of the injury does not exceed the following: a. Loss of consciousness of approximately 30 minutes or less b. After 30 minutes, an initial Glasgow Coma Scale (GCS) of 13–15 c. Posttraumatic amnesia (PTA) not greater than 24 hours

file is available to clinicians trying to distinguish feigned from genuine deficits. The problem of significant heterogeneity of symptoms only worsens beyond the 3-month window, and a unique profile of neuropsychological deficits in this time frame almost certainly does not exist (see Silverberg & Iverson, 2011). Acute symptoms of mTBI (i.e., less than 24 hours postinjury) include confusion and disorientation, dizziness, gait abnormality, visual disturbance (e.g., diplopia), and headache (Barth, Macciocchi, Boll, Giordani, & Rimel, 1983; Landre, Poppe, Davis, Schmaus, & Hobbs, 2006; Levin et al., 1987). Many of these symptoms can persist into the postacute phase (i.e., 7–10 days postinjury) but can also emerge postacutely, after an asymptomatic period. Symptoms that emerge after the acute phase tend to be more cognitive in nature, such as amnesia, attention deficits, and slowed cognitive processing speed, although any of the symptoms can occur during either phase. Pain is often a primary complaint both early and later in the course of recovery, and psychological disturbances are not uncommon, especially later in the course of recovery. A challenging issue facing forensic neuropsychologists is the condition assumed to be the residual phase of mTBI: postconcussion syndrome (PCS). PCS, which refers to the constellation of symptoms that follows concussion, often involves multiple wide-ranging somatic, cognitive, and psychiatric symptoms. Challenges inherent to differ-

126

II .   D i a g n o s t ic I ss u e s

ential diagnoses involving PCS are discussed later in the chapter. The Course ofModerate toSevereTBI

Severe TBI is typically defined by GCS scores of 3–8 and/or posttraumatic amnesia (PTA) for 24 hours or longer. A TBI of moderate severity is usually said to have occurred when the GCS score is 9–12 and/or PTA lasts 1–24 hours. In general, TBI follows a dose–response curve, which means that the more severe the TBI, the longer the recovery period and the poorer the outcome. Penetrating injuries, subdural hematoma, and diffuse axonal injury are more often associated with severe TBI than with milder injuries. In addition, severe TBI often involves frank neurological signs (e.g., nonreactive pupils) and other traumatic injuries (e.g., spinal cord injury). Neurosurgical intervention and higher complication rates are also common in severe cases (Lezak et al., 2012). Individuals with severe brain injury produce highly variable scores on neuropsychological tests. Deficits tend to be wide-ranging and include (1) impaired learning and memory, (2) attention and concentration deficits, (3) slowed mental processing speed, and (4) dysexecutive syndromes (e.g., poor cognitive flexibility). Length of coma, or the time until the patient follows commands, strongly predicts degree of cognitive impairment, especially in severe cases (Dikmen, Machamer, Winn, & Temkin, 1995; Dikmen, Machemer, Fann, & Temkin, 2010; Jennett & Bond, 1975). For severe TBI, between 20 and 40% achieve “good” recovery (Millis et al., 2001). Although cognitive deficits due to moderate to severe TBI can persist indefinitely, they do not worsen over time. However, severe TBI has been linked to an increased incidence of Alzheimer’s disease (Institute of Medicine Committee on Gulf War and Health, 2009; Shively, Sher, Perl, & Diaz-Arrastia, 2012) and secondary psychological disturbances such as depression, which can worsen or prolong cognitive dysfunction. Interested readers should consult Dikmen et al. (2010) and Roebuck-Spencer and Sherer (2008) for excellent reviews of the neuropsychological consequences of non-mTBI. The clinical issue of whether severe TBI can be malingered has received far less scrutiny than feigned mTBI, partly due to two assumptions: (1) that patients with documented brain damage cannot successfully feign cognitive deficits, and (2) that they are not motivated to do so. However, a handful of published case studies involving severe

TBI and malingering appear to challenge this notion (e.g., Bianchini, Greve, & Love, 2003; Boone & Lu, 2003).

PCS andControversy PCS describes the cognitive, somatic, and emotional problems reported by some patients with mTBI well beyond the expected time frame (i.e., 1–3 months). However, this term may also refer to the customary and expected symptoms that occur in the acute and postacute phases (i.e., within 3 months of the injury). Not surprisingly, these two contrasting definitions have caused confusion. It would probably help to distinguish between the two by using PCS to refer to symptoms that occur during the typical course of recovery and persisting PCS (PPCS) when referring to symptoms that persist beyond the expected time frame. However, PCS is already likely too entrenched as the term that refers to persisting symptoms for a change in terminology to be accepted now. Thus, forensic neuropsychologists should remain cognizant of this ongoing and potentially problematic conflation of terms. PCS is a controversial construct (Bender & Matusewicz, 2013). It has been difficult (1) to distinguish acute psychological effects of injury (e.g., shock, anxiety, and other impediments to cognition) from frank neurocognitive deficits, (2) to know how to interpret new neuroimaging techniques in concussion, and (3) to determine whether the disability is due to the neurological injury itself or to disruptions in routines of daily living (Davies & McMillan, 2005). PCS is also difficult to define because its persisting symptoms can be explained equally well by other psychological or neurological problems, or both. Criteria in the 10th edition of the International Statistical Classification of Diseases (ICD-10; World Health Organization, 1992) illustrate the challenges inherent to current definitions of PCS. ICD-10 offers nine diagnostic criteria for PCS, but they appear to have little clinical utility. Kashluba, Casey, and Paniak (2006) found that endorsem*nt of ICD10 criteria correctly classified just 67% of patients 1 month post-mTBI. Only one symptom (increased susceptibility to fatigue) showed superior discrimination 1 month postinjury, and none successfully discriminated at the 3-month postinjury interval. For quite some time, critics have underscored the conceptual limitations of the PCS construct. For instance, Gouvier, Uddo-Crane, and Brown (1988) noted that the diagnostic validity of PCS

7. Malingered Traumatic Brain Injury 127

is suspect due to the high base rate of PCS-type symptoms in the general population. Symptoms such as “difficulty becoming interested” and “often irritable” were endorsed at comparable rates by normal controls and patients with PCS. Only one self-reported symptom (“subjective sense of restlessness”) differed significantly between these groups. Iverson and McCracken (1997) noted a similar problem of symptom overlap when comparing people with symptoms of PCS and those with chronic pain without mTBI. Lees-Haley and Brown (1993; see also Fox, Lees-Haley, Earnest, & Dolezal-Wood, 1995) found a very high incidence of PCS-like symptoms in litigants without a neurological injury or claim. More recently, Zakzanis and Yeung (2011) replicated findings that symptoms of PCS are very common in healthy people. They also noted that certain symptoms are more prevalent than others, depending on the culture. The authors cautioned that certain symptoms may be prone to misattribution to PCS due to their commonness. When compared to trauma controls (i.e., individuals with orthopedic but not brain trauma), PCS symptoms are reported in largely equivalent numbers for 3 or more months after the injury (e.g., 46.8% in mTBI vs. 48.3% in trauma controls; Meares et al., 2011). Dikmen et al. (2010) reported a larger discrepancy in PCS symptom endorsem*nt between their trauma controls and patients with mTBI, but their mTBI group included injuries of greater severity than those found in most of the existing literature. Recent investigations continue to raise concerns about the diagnostic utility of PCS. Iverson (2003) and Ponsford et al. (2012) found that anxiety predicts persistence of PCS symptoms at 3 months postinjury, but presence of mTBI did not. In a striking example of the diagnostic imprecision of PCS, Donnell, Kim, Silva, and Vanderploeg (2012) found that just 32% of mTBI patients met DSM-IV-TR criteria for PCS, whereas 91% of patients with somatization disorder met these criteria. Several other DSM-IV-TR disorders were more diagnostically similar to PCS than was mTBI. Similarly, Iverson (2006) studied 64 patients with a depressive disorder but no TBI, and found that 72% endorsed three or more “moderate to severe” symptoms of PCS. Iverson cautioned that depression, chronic pain, sleep problems, litigation stress, and malingering can cause patients to report PCS symptoms. In short, a rapidly growing body of research indicates that mTBI is not a useful predictor of PCS.

Another long-standing concern regarding PCS involves the biased recall and responding that can occur in some patients after mTBI. Gunstad and Suhr (2001) described systematically lower reporting of preinjury problems relative to postinjury problems as the “good old days” bias. In comparing the self-reported symptoms of mTBI patients and orthopedic controls, Silverberg et al. (2016) found that the “good old days” bias was more pronounced in mTBI patients than in controls when reporting more post-injury symptoms. This was explained by reattribution theory, whereby mTBI patients consider their symptoms to be caused by the mTBI, and as a result do not often recall preinjury PCS-like symptoms. Patients with such an expectation may detect “symptoms” that are actually benign sensations associated with otherwise normal stressors and attribute them to mTBI (see also Lange, Iverson, & Rose, 2010; Silverberg & Iverson, 2011; Sullivan & Edmed, 2012; Vanderploeg, Belanger, & Kaufmann, 2014). Bender and Matusewicz (2013) classify the “good old days” bias as a patient-based source of error, which (along with provider-based errors) may help explain the misattribution of some symptoms following mTBI. Patient-based errors may also occur due to distorted perceptions of illness and anxiety sensitivity. The influence of illness perceptions on outcomes has been described in multiple conditions, and PCS may be particularly vulnerable to this phenomenon. For example, Whittaker, Kemp, and House (2007) discovered that patients with mTBI who believed that their symptoms were serious and long-lasting tended to report more symptoms 3 months later than those who anticipated a less negative outcome. Importantly, injury severity, anxiety, depression, and PTSD symptoms did not add to the predictive model. Hou et al. (2012) found similar results in patients with mTBI; negative perceptions about mTBI, all-or-nothing appraisals, depression, and anxiety all predicted PCS 6 months postinjury (see also Snell, Siegert, HaySmith, & Surgenor, 2011). Suhr and Wei (2013) provide an excellent review of the possible effects of expectancies on neuropsychological test performance and presentation. The term provider-based errors refers to the inadvertent misattributions of PCS symptoms to mTBI made by health care professionals, thereby endorsing the notion that the symptoms represent ongoing brain damage. Patients may then feel more despairing and helpless, which can both perpetuate these symptoms and cause new symptoms. In this scenario, there is no easy way for providers or

128

I I .   D i a g n o s t ic I ss u e s

patients to discern that these symptoms are not attributable to the mTBI itself. An important caution by Andrikopoulos and Greiffenstein (2011) regarding PTSD likely should be heeded for PCS as well, namely, that based on current nosology regarding mTBI and PCS, well-meaning doctors may inadvertently perpetuate a sequence of iatrogenic consequences simply by making the PCS diagnosis. Another source of response bias arises from litigation and other adversarial situations. So-called jurisogenic or lexigenic symptoms arise when the patient’s symptoms are challenged or contested. Decades ago, Weissman (1990) introduced the term jurisogenic to refer to symptoms that plaintiffs may report in increasing numbers and severity stemming from protracted legal proceedings. For example, persistent questioning of the veracity of the injury and symptoms often occurs in litigated mTBI cases (in contrast to cases of stroke, for example), and may contribute to the jurisogenic process (Hall & Hall, 2012; Silver, 2012). Interestingly, perceptions of unfairness seem to affect cognition and decision making (Sanfey, Rilling, Aronson, Nystrom, & Cohen, 2003) and may also play significant explanatory roles in protracted legal cases involving PCS. Despite these concerns, many clinicians continue to use PCS to define persisting symptoms ostensibly arising from mTBI, at least partially because few diagnostic alternatives exist. In fact, no single syndrome effectively captures the wide array of symptoms persisting beyond the expected time frame following mTBI. The use of PCS checklists and questionnaires exacerbates this problem by increasing number of symptoms without improving specificity (Iverson, Brooks, Ashton, & Lange, 2010). In summary, the constellation of symptoms known as PCS still lacks clear diagnostic boundaries and, as cautioned in the previous edition of this volume, PCS has likely become a pejorative term to many professionals. It contributes to the mislabeling of symptoms and may cause iatrogenic problems (Bender & Matusewicz, 2013; Larrabee & Rohling, 2013; Wood, 2004). Forensic neuropsychologists are urged to evaluate other potential symptom etiologies carefully before concluding that they are due to mTBI. However, it is equally important (1) not to rule out the possibility of genuine PCS, especially if within 3 months or so of the injury, and (2) to consider other possible causes of the symptoms that may still require treatment, whether or not they are due to mTBI.

EFFECT SIZES FORmTBI ANDMALINGEREDmTBI The seminal meta-analysis by Binder, Rohling, and Larrabee (1997) compared neuropsychological test scores among patients with mTBI to those of healthy controls; the average effect size of impairment was small (d = 0.12). In a more recent meta-analysis of neuropsychological performance at 3 months postinjury, Rohling et al. (2011) found an effect size for mTBI so small that it did not significantly differ from zero (d = 0.07). As Larrabee and Rohling (2013) pointed out, this effect size is much smaller than those for substance abuse, PTSD, major affective disorder, and probable malingering. Meta-analytic studies have shown larger effect sizes for mTBI in the acute phase of recovery (e.g., d = 0.41 within the first 6 days after injury; Schretlen & Shapiro, 2003). By comparison, moderate to severe TBI yields large to very large effect sizes on neuropsychological tests (e.g., ds from 0.97 to 2.41; Rohling, Meyers, & Millis, 2003; Schretlen & Shapiro, 2003). As I discuss later, the strategy of “severity indexing” uses this type of information to help identify malingering. In contrast to the negligible effect sizes for genuine mTBI, feigners (i.e., suspected malingerers and simulators) have shown a moderately large effect size on neuropsychological tests. In the meta-analysis by Vickery, Berry, Inman, Harris, and Orey (2001), the effect size of malingered performance on testing (mean d = 1.13) appeared to be more commensurate with severe TBI than with mTBI (see also Green, Rohling & Lees-Haley, 2001). As Iverson (2003) has observed, these meta-analytic findings indicate that neuropsychological tests are potentially strong indicators of severe TBI and probable malingering but can be poor indicators of genuine mTBI. Effect sizes associated with patients’ expectancies have also been calculated, although only a few studies have been conducted to date. For example, in patients with mTBI, stereotyped beliefs about the consequences of the injury (“diagnosis threat”) produced a moderate effect size on average (d = 0.68) on tests of intellectual and executive functioning (Suhr & Gunstad, 2005). Subsequent studies of diagnosis threat suggest that the effect may be more dependent on the evaluation context, past history of the patient, and whether the patient received accurate feedback regarding his or her neuropsychological performance (Suhr & Wei, 2013).

7. Malingered Traumatic Brain Injury 129

As mentioned, the process of litigation plus the possibility of substantial financial gain can have profound effects on patients, including effects on their neuropsychological test performances. Binder and Rohling (1996) found that such incentives potentially lower test scores by nearly one-half standard deviation (d = 0.47), which is much more than the decrement associated with mTBI. However, this figure is a potentially misleading statistic, as it may persuade clinicians and attorneys to believe that litigation is a unitary entity that is solely responsible for the patient’s deficits (e.g., “accident neurosis”; see Miller, 1961). Importantly, litigation involves a number of stress-producing components that may be completely unrelated to any attempts at malingering. As Weissman (1990) noted, protracted litigation lowers morale, reduces effectiveness of treatment, and fosters iatrogenic conditions (see also Hall & Hall, 2012). Such effects can lower test performance and should not be mistaken for malingering. Furthermore, the repeated recounting of symptoms during litigation likely leads to some “drift” and at least minor fictionalization of symptoms (see Weissman, 1990; Hall & Hall, 2012). In summary, the complex set of psychological factors that accompany litigation may increase the likelihood of malingering but cannot be equated with it.

MASQUERADING SYNDROMES ANDTHEDIFFERENTIAL DIAGNOSIS OFmTBI Accurate mTBI diagnosis must consider not only malingering but also other, genuine clinical conditions. Neuropsychologists are strongly cautioned against an unsophisticated differential diagnosis that consists of simply mTBI versus malingering. Other diagnostic considerations include depression, anxiety, PTSD, and pain. Attributing poor outcome to the mTBI alone is often overly simplistic and misleading (Mooney, Speed, & Sheppard, 2005). For example, depressive disorders occur in approximately 75% of moderate to severe cases of TBI (Alway, Gould, Johnston, McKenzie, & Ponsford, 2016), but also occur in 14–35% of patients with mTBI (Busch & Alpern, 1998; Deb, Lyons, Koutzoukis, Ali, & McCarthy, 1999). Depressive symptoms can mimic symptoms of mTBI both clinically and on neuropsychological test scores. Depressed patients typically have low motivation, which traditionally is considered a correlate of malingering in the cognitive domain. Decreased attention and concentra-

tion, and slowed cognitive processing speed, are common in patients with depression and in those with brain injury. Further complicating the picture, individuals with depression frequently report symptoms of PCS, such as headache, dizziness, and blurred vision (Silver, McAllister, & Arciniegas, 2009). Meta-analyses indicate that depression is associated with cognitive dysfunction with moderate effect sizes (e.g., d = 0.52; Zakzanis, Leach, & Kaplan, 1998). Differential diagnoses of mTBI and depression rely on multiple indicators. Obviously, an indication of brain trauma is required for mTBI. Beyond that, a preexisting history of depression and cognitive deficits that fluctuate appreciably with mood changes is often indicative of depression. Deficits that resolve with time with less fluctuation are consistent with mTBI. Practitioners should be careful in their diagnosis, because mTBI may include behavioral and emotional dysregulation that lowers cognitive efficiency, especially during times of irritability. Differentiating depression from mTBI is more difficult in the postacute stages, when symptom overlap increases, and the true source of symptoms is blurred. Lange, Iverson, and Rose (2011) found that PCS symptoms were reported significantly more often by outpatients with depression than by patients with mTBI within 8 months of their injury, with a moderate effect size (d = 0.68). No accurate methods exist to determine whether symptoms represent a depressive disorder or a neuropathological component of the mTBI. But the symptoms are less likely due to mTBI itself as the time after injury increases. Mild TBI must also be differentiated from anxiety and PTSD. Anxiety is highly comorbid with depression and, when severe, can diminish cognitive efficiency in its own right (Clarke & MacLeod, 2013). Also, attention and judgment are more vulnerable in anxious individuals; perceived threats and their associated costs are more commonly reported in this group (Cisler & Koster, 2010; Nelson, Lickel, Sy, Dixon, & Deacon, 2010). The essential feature of PTSD involves the development of disabling symptoms following exposure to a potentially life-threatening traumatic event (e.g., being a victim or witnessing violence). Patients with PTSD alone often have neuropsychological symptoms such as slowed processing speed, inattention, and poor verbal learning and memory (see Kay, 1999; Scott et al., 2015). These symptoms overlap with those of mTBI and PCS, and therefore pose obvious difficulties when trying

130

II .   D i a g n o s t ic I ss u e s

to determine the source of deficits in the context of mTBI. Estimates vary appreciably, but studies suggest that PTSD may be more common following mTBI than originally thought, with prevalence rates mediated by the cause of trauma (Harvey & Bryant, 1998; Koren, Norman, Cohen, Berman, & Klein, 2005). For instance, Hoffman, Dikmen, Temkin, and Bell (2012) found that 17% of civilians with mTBI also met diagnostic criteria for PTSD. The presence of bodily trauma appears to increase the risk of PTSD (Koren, Hemel, & Klein, 2006). PTSD symptoms and subclinical features also are more common in patients whose symptoms have been slow to resolve, and PTSD can have a similar course to PCS, leading to the suggestion that PTSD is a variant of PCS (Davidoff, Kessler, Laibstain, & Mark, 1988; Schneiderman, Braver, & Kang, 2008). Cases in which patients have witnessed a death or suffered a personal attack may increase the vulnerability to PTSD. Amnesia for the traumatic event may protect against PTSD but seems to depend on the severity of the TBI, duration of amnesia, and attitudes and attributions of the patient (Al-Ozairi, McCullagh, & Feinstein, 2015; Bryant et al., 2009; Hoffman et al., 2012). Importantly, recent studies suggest that PTSD does not increase the degree of cognitive impairment above that associated with mTBI alone (Gordon, Fitzpatrick, & Hilsabeck, 2011; Soble, Spanierman, & Fitzgerald-Smith, 2013). It is noteworthy that DSM-5 directs special attention to PTSD as a potential target of malingering. Rosen and Taylor (2007) have suggested that this is due to the ease with which one can feign symptoms of PTSD and the difficulty establishing whether a traumatic event actually occurred. They note that the prevalence of malingered PTSD is still unknown. Preliminary estimates from Demakis, Gervais, and Rohling (2008) suggest that prevalence rates may be near 30%. Moreover, the new diagnostic criteria for PTSD have increased the number of symptom combinations to over 636,000 (Galatzer-Levy & Bryant, 2013). The new criteria reduce diagnostic precision of PTSD, thereby also likely hindering accurate malingering detection. Given these diagnostic complexities, Kay’s (1999) four key provisions for forensic neuropsychologists working with patients with PTSD remain important: 1. PTSD can co-occur with genuine symptoms of TBI. 2. Neuropsychological deficits out of proportion

to the injury raise the probability of psychological versus neuropsychological disorder. 3. Regardless of the severity of injury and context, all plausible diagnoses must be considered, including PTSD. 4. Exaggeration by the patient does not rule out legitimate psychological conditions or mTBI. Like depression and PTSD, pain can produce symptoms and deficits similar to those of mTBI. Pain is subjective and can be a method of expressing psychological distress in patients who are not psychologically minded (e.g., patients with somatic symptom disorder). Given its subjectivity, pain is also a potential symptom to feign. Neuropsychologists must be aware of and account for the neuropsychological effects of pain but, unfortunately, neuropsychological evaluations are not designed to detect pain directly. Therefore, the veracity of pain symptoms must be inferred from performance on tests of other constructs, including tests of malingered cognitive impairment, and from behavioral observations. This determination is troubling given that (1) genuine pain is a common comorbid condition with mTBI, (2) chronic pain patients without a history of concussion often have several symptoms of PCS (e.g., over 80% of chronic pain patients without TBI report three or more PCS symptoms; see Iverson & McCracken, 1997; Nampiaparampil, 2008), and (3) the prevalence of malingered pain has been estimated to be as high as 34% (Mittenberg et al., 2002). Chronic pain complaints may be over four times more common in mTBI than in moderate to severe TBI (Uomoto & Esselman, 1993). Bianchini, Greve, and Glynn (2005) outlined their proposed criteria for malingered pain-related disability (MPRD), largely based on Slick and colleagues’ (1999) original criteria for malingered neurocognitive deficits. Interestingly, the criteria include (but are not limited to) poor effort on purely neuropsychological tests. This approach has questionable specificity and benefit, because (1) most neuropsychological tests do not assess pain, and (2) less than one-third of litigating pain patients may be expected to fail neurocognitive “validity checks” (Meyers & Diep.2000). B ­ ianchini et al. (2005) applied the Slick et al. (1999) criteria to approach to this complex diagnostic problem. In their subsequent work, Greve, Ord, Bianchini, and Curtis (2009) found that the prevalence rate of cognitive effort test failure in pain patients is higher than prior estimates. However, the group’s methodology may be problematic. For example,

7. Malingered Traumatic Brain Injury 131

they did not establish false-positive rates for genuine pain patients; rather, they used scores from a TBI group, raising concerns that they assumed that feigning one condition (e.g., pain) is synonymous with feigning any condition. Greve, Ord, Curtis, Bianchini, and Brennan (2008) found that over half of chronic patients and nearly onethird of patients with mTBI could not be confidently categorized as malingering (or not) by three commonly used performance validity tests (PVTs): Test of Memory Malingering (TOMM), Word Memory Test (WMT), and Portland Digit Recognition Test (PDRT). Patients who failed one PVT and one symptom validity test (SVT) were assumed to be malingering, which obscures what exactly is being malingered (i.e., pain or cognitive abilities). It also falls short of more recent recommendations that at least three indicators be used to classify performance as malingered (Slick & Sherman, 2013). Pain is naturally distracting, but it also seems to produce a separate state of psychological distress that interferes with cognitive efficiency (Kewman, Vaishampayan, Zald, & Han, 1991). Clinicians working with patients with mTBI must be aware of pain syndromes, as well as somatic symptom disorders, and carefully consider the role of pain when questions of adequate effort are raised. In complex cases, consultation with a pain specialist may be needed. Chronic pain has been shown to increase attention paid to physical sensations, which may not only increase reports of PCS symptoms (Meares et al., 2011) but may also result in deficits on formal neuropsychological testing. For instance, individuals with MRI evidence of neck pain following mTBI may perform worse on neurocognitive tests than patients with mTBI without such evidence of pain (Fakhran, Qu, & Alhilali, 2016). Moreover, the mTBI was particularly mild for the entire sample (e.g., PTA < 30 minutes, LOC < 1 minute), suggesting that the differences in cognitive test scores were not attributable to mTBI. Resnick, West, and Payne (2008) suggested several characteristics that can differentiate pain and conversion symptoms from malingering. For instance, malingerers tend to be less eager to be examined or treated, whereas patients with pain disorders welcome the opportunity to obtain an explanation for their symptoms. Also, genuine mTBI and pain patients often avoid the details of the accident that caused the symptoms, but malingerers typically go into detail. Such willingness may reveal the lack of genuine emotional salience

of the purported trauma and may reflect a deliberate external goal (i.e., malingering). Although neuropsychological tests are not tests of pain per se, some cognitive measures may assist in determining the genuineness of reported pain. Etherton, Bianchini, Greve, and Ciota (2005) found that the TOMM was not impaired by the effects of laboratory-induced pain. They suggested that failed TOMM scores are caused by something other than pain, such as poor effort. Subsequent investigations by the same group (Etherton, Bianchini, Ciota, Heinly, & Greve, 2006) showed that the Working Memory Index (WMI) and the Processing Speed Index (PSI) from the Wechsler Adult Intelligence Scale–III (WAIS-III) are only modestly affected by either experimentally induced or chronic pain. In contrast, many of those feigning pain had low WMI scores. These cognitive tests are not designed to assess the validity of pain complaints but may hold utility because they measure pain-related complaints. More research is definitely needed. Bianchini et al. (2014) reported that the Pain Disability Index (PDI) and the Modified Somatic Perception Questionnaire (MSPQ) may have merit as measures of malingered pain-related disability. Just over 30% of the criterion group (known chronic pain patients with incentive to feign) met MPRD criteria for probable (27.4%) or definite (4.6%) malingering. When indeterminate scores were excluded, both the PDI and the MSPQ differentiated between pain patients with incentive and those without with acceptable specificities, but sensitivities were quite low. Studies have begun to document effect sizes and utility of certain measures of feigned cognitive impairment in individual populations. The available data (see Table 7.2) suggest that some measures of feigning may have moderate to large effect sizes when compared with these disorders, but further work is needed. Importantly, the comorbidity of mTBI and clinical conditions of depression, PTSD, and pain has not been systematically evaluated. These representative studies underscore the importance of further research for systematically evaluating comorbid conditions. In summary, depression, anxiety and PTSD, and pain represent common differential diagnoses when considering mTBI. In addition, each diagnosis can be feigned in its own right, which means that there are multiple possible outcomes from a neuropsychological evaluation: For example, the patients may (1) have a genuine brain injury, (2) have a genuine mental disorder masquerading as

132

II .   D i a g n o s t i c I s s u e s

TABLE 7.2.  Specialized Measures of Feigning with Comparisons to Specific Conditions

Feigners vs. specific conditions (Cohen’s ds) Measure

Representative studies

TBI

TOMM (retention trial)

Gierok, Dickson, & Cole (2005) a

PTSD

Etherton, Bianchini, Greve, & Ciota (2005)

Weinborn, Orr, Woods, Conover, & Feix

4.65 1.42

(2003) b

0.71

Axelrod, Fichtenberg, Millis, & Wertheimer (2006)

1.21

Babikian, Boone, Lu, & Arnold (2006) c

1.05

Etherton, Bianchini, Greve, & Heinly (2005)

1.27

21-Item Test

Vickery et al. (2001) c

1.46

Category Test

Forrest, Allen, & Goldstein (2004) c

1.40

Greve, Bianchini, & Roberson (2007)d

0.42

(2007) e

0.64

Greve, Bianchini, & Roberson CVLT-II FBS/SVS

Root et al. (2006) c

0.82

Wolfe et al. (2010) f

1.06

Greiffenstein et al. (2004) g

Pain

0.95

Rees, Tombaugh, Gansler, & Moczynski (1998)

RDS

Dep

5.05

Note. Dep, depression; PTSD, posttraumatic stress disorder. aDifferential prevalence design; heterogeneous psychiatric inpatient group. bDifferential prevalence design; forensic psychiatric inpatients. cHeterogeneous clinical group, Trials 1–5 (raw score). dErrors on Subtests 1 and 2 (not MND vs. MND). eEasy items (not MND vs. MND). fTotal Recall Discriminability (standard score). gThis figure represents the mean when men and women were combined. When considered separately, the effects sizes are markedly different (d = 2.9 in men; d = 7.2 in women).

TBI, or (3) be malingering. Combinations of the three alternatives are possible as well. The forensic neuropsychologist must establish the veracity not only of the primary complaint (e.g., mTBI) but also the competing diagnoses. The accuracy of this determination is essential. It is not difficult to imagine how disastrous the consequences would be for a genuine patient who is incorrectly classified as a malingerer.

STRATEGIC IDENTIFICATION OFFEIGNEDTBI The clinical history, context of injury, and other background information are critical parts of the determination of injury severity. Forensic neuropsychologists must use their knowledge of both the symptomatology and the natural course of TBI to identify neuropsychological inconsisten-

cies. Many authors have recommended a systematic assessment of patterned inconsistencies during each evaluation (e.g., Iverson & Binder, 2000; Iverson, 2006; Larrabee & Rohling, 2013; Sweet, Goldman, & Guidotti-Breting, 2013). They are patterned in the sense that they are not random; performance suggestive of malingering is always poorer than expected. Reported symptoms should match observed behavior. If, for example, the patient is reportedly “almost blind” but grasps a pen from the examiner without difficulty, there is an inconsistency between report and observation. If the patient reports severe anterograde amnesia but recalls the route to the clinician’s office, then there is a mismatch between report and observation. The examination of patterned inconsistencies permits many comparisons. Although highly accessible and intuitively appealing, patterned inconsistencies represent a qualitative approach

7. Malingered Traumatic Brain Injury 133

that is subject to biases in clinical judgment (see Iverson, Brooks, & Holdnack, 2008, for a review of common cognitive biases and logical fallacies affecting clinical judgment). The second type of assessment for cognitive feigning, often called syndrome analysis (Walsh, 1985), uses information about patients’ genuineness of presentation. With this method, the clinician compares features of the current presentation with the known syndrome. An exhaustive review is beyond the scope of this chapter, but an example would be a reported cognitive decline in the postacute stages of mTBI in the absence of an evolving bleed or other neurologic cause. Such an exacerbation would be inconsistent with the known course of mTBI and raises questions of feigning or the emergence of a genuine psychological disturbance, the latter of which occurs more commonly in forensic than in clinical cases. Likewise, a newly acquired aphasia in the context of TBI without welldocumented damage to the temporal lobe is highly unusual and deserving of a comprehensive assessment for possible response biases. Neither observation by itself is pathognomonic of malingering; as mentioned, other diagnoses must be ruled out. This approach is only a part of malingering detection, because atypical yet genuine cases do not always fit the syndrome in question. A third method of assessing feigned cognitive abilities involves the assessment of intertest inconsistency. Generally consistent performances are expected on tests of similar abilities (e.g., Arithmetic and Digit Span subtests; Finger Tapping and Grooved Pegboard), and significant deviations are difficult to explain syndromally. However, a word of caution is warranted: As several investigators have noted (e.g., Binder, Iverson, & Brooks, 2009; Schretlen, Munro, Anthony, & Pearlson, 2003), significant scatter among the Wechsler subtests is the norm, not the exception. The base rate of a difference between two tests must be considered before rendering an opinion about the likelihood that the inconsistency represents actual feigning. The fourth and final method, severity indexing, can be used to compare the current patient’s characteristics (including demographic data, injury characteristics, test scores, etc.) with those of previously studied patient groups. This approach capitalizes on the fact that a dose–response curve is associated with TBI, such that symptom load increases as severity of injury increases (Sweet et al., 2013). It is similar to the traditional use of norms but incorporates norms from TBI populations rather than healthy controls.

Rohling et al. (2003) proposed the Overall Test Battery Mean (OTBM) as such a measure of severity inconsistency. Basically, the authors transformed raw scores from neuropsychological tests into Tscores and then calculated an overall average (i.e., the OTBM). They then compared the patients’ OTBM with datasets from samples of neurocognitively impaired individuals. The authors used the outcome data from Dikmen and colleagues (1995) and from Volbrecht, Meyers, and Kaster-Bundgaard (2000) as benchmarks for performances among patients with TBI. The comparison of OTBMs to these datasets allows clinicians to determine the congruity between acute injury and residual cognitive deficits. An OTBM outside of the expected range in light of the patient’s length of loss of consciousness should be considered suspect. The severity indexing approach is promising because it yields an empirical index of the expected severity of neurocognitive deficits given the characteristics of the injury. Symptoms and ostensible deficits beyond the expected course and severity of injury raise the likelihood of dissimulation or psychologically based symptoms. Also, there are limited data regarding the prevalence of genuine but atypical neurocognitive profiles that do not fit with the existing estimates of severity. Clinicians are encouraged to explore this technique conceptually with data from Dikmen and colleagues (1995), Volbrecht et al. (2000), or Rohling et al. (2003) for specific comparisons. However, clinicians are also cautioned not to rely on this approach alone when making determinations of malingering. See Hill, Rohling, Boettcher, and Meyers (2013) for a potential complication involving intraindividual variability when using this approach.

Severe TBI andMalingering Very few case studies have investigated feigned severe TBI. As an exception, Boone and Lu (2003) described noncredible performances in two cases of severe TBI. Both patients underwent numerous multidisciplinary workups, but apparently the possibility of malingering was not considered. In their subsequent evaluations, the authors documented “odd” and “nonsensical” fluctuations in neurocognitive test performance. For instance, one patient had shown good neuropsychological recovery 6 months postinjury and had reenrolled part time in high school, but her neurocognitive scores declined significantly in the next 2 months. No identifiable brain mechanism was identified to explain this unexpected decline.

134

II .   D i a g n o s t i c I s s u e s

Bianchini et al. (2003) reported three cases of litigating patients, each with documented severe TBI. Each patient performed significantly below chance on symptom validity tests and had clear incentives to fake deficits. The investigators’ published classification of “definite malingering” appears to be the first for feigned severe TBI. They reiterated Boone and Lu’s (2003) contention that tests of effort and feigning be included in all cases with potential material gain (see also Ju & Varney, 2000).

SELECTED NEUROPSYCHOLOGICAL TESTS USED INTBI Multiple studies have demonstrated (to varying degrees) that certain neuropsychological tests may be used simultaneously to evaluate both genuine neurocognitive dysfunction and malingering. The chief advantage of these embedded measures involves time efficiency, but they may also be less recognizable as tests of malingering and therefore more resistant to coaching. A brief review of some commonly administered “double-duty” tests is provided here. More comprehensive reviews can be found in Boone (2007) and Larrabee (2007). Van Gorp and colleagues (1999) and Curtiss and Vanderploeg (2000) provide dissenting perspectives about the utility of embedded effort measures.

California Verbal LearningTest The recognition trial of the California Verbal Learning Test (CVLT; Delis, Kramer, Kaplan, & Thompkins, 1987) and CVLT-II (Delis, Kramer, Kaplan, & Ober, 2000) provides an opportunity for forced-choice testing of response bias during a test of true neurocognitive dysfunction. Although several studies have evaluated the utility of the CVLT in detecting dissimulation (e.g., Ashendorf, O’Bryant, & McCaffrey, 2003; Millis, Putnam, Adams, & Ricker, 1995; Slick, Iverson, & Green, 2000), it is difficult to synthesize their findings due to the differing criteria used to define the feigning groups. Among TBI patients, Bauer, Yantz, and Ryan (2005) used the WMT (Green, Iverson, & Allen, 1999) as the partial criterion (see Rogers, Chapter 1, this volume) for malingering (defined as performance below 82.5% on one of three WMT scores). A discriminant function using five CVLT-II variables correctly classified 75.8% of their sample. Specificity was an impressive 95.6%,

but sensitivity was just 13.5%, raising serious questions about the CVLT-II’s overall utility. Root, Robbins, Chang, and van Gorp (2006) evaluated the utility of the critical item analysis (CIA) procedure, as well as the forced-choice recognition indices from the CVLT-II. Basically, the CIA relies on the performance curve strategy (see Rogers, Chapter 2, this volume) in comparing performance on easy versus difficult items. Using a partial criterion design (PCD) they divided the entirely forensic sample into adequate- and inadequate-effort groups as defined by performance on the Validity Indicator Profile (VIP) and/or TOMM. They found that the CIA Recognition Index yielded exceptionally good specificity (100%) and positive predictive power (PPP; 100%). The CVLT-II appears to be a strong indicator of inadequate effort, but it lacks sensitivity, ranging from 4 to 60%. Although combining indices from the CVLT-II seems to improve incremental validity somewhat, clinicians are again advised not to rely on one indicator for malingering detection.

Digit Span andReliable DigitSpan Several investigators (e.g., Greiffenstein, Baker, & Gola, 1994; Iverson & Tulsky, 2003) argued that the Digit Span subtest from the WAIS-III was a useful measure of effort, because only 5% of both healthy and clinical samples score below the 5th percentile. As such, it represents an example of the floor effect strategy. Axelrod, Fichtenberg, Millis, and Wertheimer (2006) reported a cutoff score of 7 on the Digit Span subtest, and correctly classified 75% of probable malingerers and 69% of patients with mTBI. Though studies report specificities ranging from 90 to 100%, the degree of forensic utility is limited by small samples and lack of replication. Greiffenstein and colleagues (1994) introduced Reliable Digit Span (RDS) as a measure of effort, which is now an incorporated measure of effort in the Advanced Clinical Solutions program for the WAIS-IV. At least eight studies have provided estimates of specificity, with results ranging from 57 to 100%, but sensitivity is too low to be useful (e.g., 10–49%; Young, Sawyer, Roper, & Baughman, 2012). Efforts to improve the sensitivity of the RDS (the revised RDS [RDS-R]) have yielded largely similar rates (Young et al., 2012). Moreover, RDS may not be much more effective than standard Digit Span scaled score (Spencer et al., 2013) and may not be appropriate for individuals with

7. Malingered Traumatic Brain Injury 135

borderline IQs or lower. A summary of RDS sensitivities and specificities may be found in Babikian, Boone, Lu, and Arnold (2006).

CategoryTest The Category Test also employs the floor effect to detect feigned responding. Bolter, Picano, and Zych (1985) determined that certain items of the Category Test (i.e., “Bolter items”) were so easy that examinees with brain injuries rarely missed them. Tenhula and Sweet (1996) later developed six validity indicators for the Category Test that, in some combinations, correctly classified 88.9%, with a specificity of 92.6%. Adding more difficult items lowered specificity. Neuropsychologists should remember that whenever cutoff scores are combined, extensive cross-validation is required before they can be applied clinically. Boone (2007) provides comprehensive reviews of the Category Test and recommendations for its use in forensic practice. In general, the Bolter/easy items from the Category Test seem to hold promise for malingering detection. As can be seen, tests of genuine neuropsychological dysfunction may also be used to assess suboptimal effort. However, none should be used in isolation. Multiple independent measures (i.e., not strongly correlated) should be used. Studies have demonstrated the benefit of using multiple measures of relatively independent constructs to increase incremental validity (e.g., Rogers & Bender, 2013; Larrabee, 2014; Nelson et al., 2003).

MultiscaleInventories In this volume, Wygant, Walls, Brothers, and Berry (Chapter 14, this volume) and Boccaccini and Hart (Chapter 15, this volume) examine the use of multiscale inventories to evaluate feigning and other response styles. Rogers and Bender (2013) advanced a strong argument (see also Rogers, Chapter 2, this volume) for addressing feigned cognitive impairment as different from other domains (i.e., mental disorders and medical complaints). In essence, each domain places unique demands on malingerers and has its own detection strategies. However, some feigners may simulate personality changes and psychological impairment as arising from their putative brain traumas. Three multiscale inventories are often used with suspected cases of feigned cognitive impairment: the Minnesota Multiphasic Personality In-

ventory–2 (MMPI-2), MMPI-2-Restructured Form (MMPI-2-RF), and Personality Assessment Inventory (PAI). They are summarized briefly in the next two subsections. MMPI-2 andMMPI-2-RF

The MMPI-2 is among the most commonly administered tests in forensic neuropsychology. As Lees-Haley, Iverson, Lange, Fox, and Allen (2002) noted, the MMPI-2 not only characterizes emotional distress but also yields data regarding effort and exaggeration. It is critical to note that the MMPI-2 validity scales were not designed to evaluate the genuineness of examinees’ cognitive complaints. Nonetheless, its utility for that very purpose has been scrutinized with some potentially surprising results. In a meta-analysis of 19 studies, Nelson, Sweet, and Demakis (2006) summarized the effect sizes of the most commonly used validity scales from the MMPI-2 (e.g., L, F, K, Fp, F-K, O-S, FBS) when used to differentiate “overresponders” from honest groups. Nelson and colleagues found that several validity scales yielded large effect sizes, with the Fake Bad Scale (FBS; Lees-Haley, English, & Glenn, 1991) yielding the largest (d = 0.96). The FBS, now referred to as the SVS (Symptom Validity Scale), has also shown promise in specific neuropsychological studies, as it appears to correlate more with tests of feigned cognitive impairment (the Victoria Symptom Validity Test; VSVT) than with F, F-K, and Fp (Slick, Hopp, Strauss, & Spellacy, 1996). The SVS appears to accurately classify mildly head-injured litigants putting forth poor effort (e.g., Ross, Millis, Krukowski, Putnam, & Adams, 2004). Peck et al. (2013) found that while the SVS correctly classified only 50% of the criterion group (known TBI feigners), it misclassified only 6% of patients with conversion disorder. Sample sizes were small, and their data need to be replicated for relevant groups (e.g., TBI feigners vs. patients with conversion disorder). Larrabee (2003b) found that SVS scores of patients with moderate to severe TBI were far below those of litigants who had failed the PDRT, with an effect size of 1.81. Similarly, Dearth, Berry, and Vickery (2005) found that the SVS yielded an effect size of 1.39 between honest patients with TBI and TBI simulators. Other investigators (Arbisi & Butcher, 2004; Butcher, Gass, Cumella, Kally, & Williams, 2008) questioned the construct validity and potential

136

II .   D i a g n o s t ic I ss u e s

biases of the SVS. In short, they raised questions about its potential for bias against disabled or trauma patients, and women. They cautioned psychologists about these problems and limits to its inadmissibility. Similarly, Lees-Haley and colleagues (2002) cautioned that although the MMPI-2 certainly meets the Daubert standard of general acceptance within the scientific community, its reliability and validity may not be as high as many clinicians assumed. Gervais, Ben-Porath, Wygant, and Green (2007) developed the Response Bias Scale (RBS) for the MMPI-2. MMPI-2 items were particularly relevant in establishing the credibility of cognitive effort in civil forensic groups. Based on an archival sample of over 1,200 disability claimants without head injury, the resultant 28-item scale yielded excellent specificity (95% overall; 100% for chronic pain patients) but very poor sensitivity (25%). In short, the study employed a lax criterion when predicting effort failure, which was defined as failure on only one or more cognitive feigning tests. Cross-validation appears to support the utility of the RBS when differentiating mixed psychiatric, mixed neurological, TBI, TBI with poor effort, and epilepsy groups (Schroeder et al., 2012). This study evaluated the utility of the RBS in a criterion-group design involving multiple clinical groups, which likely increases the confidence of the findings. The FBS/SVS and RBS were among the changes made during the construction of the MMPI2-RF (Ben-Porath & Tellegen, 2008). One of the MMPI-2 criticisms was that the validity scales were not independent from the clinical scales, with substantial overlap between certain scales (e.g., SVS and Scales 1 and 3). The MMPI-2-RF addressed this problem by removing items that did not load orthogonally, meaning that elevations on the validity scales are no longer as prone to conflation with elevations on clinical scales. Recent studies (McBride, Crighton, Wygant, & Granacher, 2013; Sleep, Petty, & Wygant, 2015) indicate that the revised versions of SVS and RBS are both correlated with cognitive effort test failure and are resistant to the effects of genuine brain damage. Excellent specificity, but variable sensitivities, were found. PAI

The PAI (Morey, 1991) is newer to the forensic realm than the MMPI-2, but it continues to gain widespread acceptance in forensic and clinical

practice. The PAI appears to hold certain advantages over other multiscale inventories. These advantages include (1) a fourth-grade reading level, (2) a shorter length than some other inventories (e.g., the MMPI-2), (3) nonoverlapping scales that aid differential diagnosis, and (4) a larger range of response options (i.e., “false,” “slightly true,” “mainly true,” and “very true”). The PAI (see Boccaccini & Hart, Chapter 15, this volume) includes four validity scales for assessing defensiveness and malingering. The Malingering Index (Morey, 1996) and the Rogers Discriminant Function (RDF; Rogers, Sewell, Morey, & Ustad, 1996) were developed for feigned mental disorders. As this section reviews the PAI only as it pertains to TBI, the interested reader should refer to Boccaccini and Hart (Chapter 15, this volume) for a broader critical review of the PAI and response styles. A growing body of research establishes the utility of the PAI in mTBI cases. Demakis and colleagues (2007) found that genuine TBI was associated with elevations on the Somatic Complaints, Depression, Borderline Features, Paranoia, and Schizophrenia scales, which is generally in line with similar studies using the MMPI-2 (Warriner, Rourke, Velikonja, & Metham, 2003). The social isolation and confused thinking items associated with Cluster 8 profiles (Morey, 1991) were endorsed by almost 20% of the mTBI sample, suggesting that the PAI includes items that are at least somewhat sensitive to symptoms of TBI. Perhaps not surprisingly, the Negative Impression Management (NIM) scale had one of the highest elevations, with over 24% scoring 2 standard deviations or more above the standardization sample’s mean. Such marked elevations raise a critical concern about whether genuine TBI patients might be misclassified as feigners on the PAI NIM scale. The authors did not control for litigation status, leaving open questions about the NIM scale’s utility for classifying TBI in incentivized contexts. According to previous research, milder head injuries are associated with higher scale elevations on the MMPI-2, especially on the “neurotic triad” (Youngjohn, Davis, & Wolf, 1997). Kurtz, Shealy, and Putnam (2007) replicated this “paradoxical severity effect” on four MMPI-2 scales but noted a different result on the PAI. Only two PAI scales (Somatization and Depression) were elevated among patients with mTBI, but less so than on the MMPI-2. These and other data led the authors to conclude that the PAI is both valid and useful when assessing head injury.

7. Malingered Traumatic Brain Injury 137

The PAI has few published studies on its efficacy for identifying feigned TBI. In a study of claims for worker’s compensation, Sumanti, Boone, Savodnik, and Gorsuch (2006) found that correlations between the PAI validity scales and cognitive effort tests were modest (rs from .02 to .30). The results suggest that the PAI (a measure of psychopathology) lacks utility for detecting feigned cognitive dysfunction. But recent studies may be more promising. For example, Keiski, Shore, Hamilton, and Malec (2015) found that the overreporting scales of the PAI were sensitive both to simulators feigning symptoms globally and simulators feigning more specific symptoms of TBI (though less so in the latter group). As expected, the NIM scale performed better overall than the other scales with a cutoff score of 73, yielding a sensitivity of .67 and a specificity of .83. Though intriguing, replication is needed before the PAI can be used in forensic practice. Furthermore, though other PAI studies have been conducted in TBI populations, many use a differential prevalence design (e.g., Whiteside, Galbreath, Brown, & Turnbull, 2012), rendering the results largely uninterpretable with respect to classificatory utility.

FOUR TESTS DESIGNED TODETECT FEIGNED COGNITIVEDEFICITS No neuropsychological battery is complete without a measure designed expressly to detect feigned neurocognitive dysfunction. Dozens of tests are available, many of which containing multiple indicators. Only four tests are included here, based on their familiarity to most clinicians and their research support. Interested readers are encouraged to review Bender and Frederick (Chapter 3), Frederick (Chapter 17), and Garcia-Willingham, Bosch, Walls, and Berry (Chapter 18) in this volume, as well as other reviews (see Boone, 2007; Rogers & Bender, 2013). The four tests cover both verbal and nonverbal domains, thereby theoretically improving incremental validity when used in combination. The 21-Item Test is ostensibly a measure of verbal recall, but it has been shown to be insensitive to bona fide memory disorders, which makes it a good candidate for assessing effort (Iverson, Wilhelm, & Franzen, 1993). It also appears to be insensitive to cognitive decline in the elderly (Ryan et al., 2012), which suggests that it has potential as a measure of effort in this rapidly growing population. As a

strength, the 21-Item Test uses a detection strategy termed violations of learning principles. If the number of words in free recall exceeds the number in the recognition trial, then feigning is suspected. The Validity Indicator Profile (VIP; Frederick, 1997; Chapter 17, this volume) is also commonly used in forensic cases. It is unique in at least two ways: First, it employs multiple detection strategies, which makes it difficult for the examinee to know which strategy is being used for which items. Second, it assesses feigning in multiple cognitive domains (e.g., conceptualization, attention, and memory). The VIP Performance Curve strategy capitalizes on feigners that missing items too soon, which is indicated by a premature dip in performance. It is virtually impossible to explain improved performance in terms other than poor effort or feigning. The VSVT (Slick, Hopp, Strauss, & Thompson, 1997) requires the examinee to recognize the correct sequence of digits when provided with two choices. Performance curve can be examined by comparing performance on hard versus easy items, resulting in very good classification rates in mTBI litigants (Silk-Eglit, Lynch, & McCaffrey, 2016) and in military samples with mTBI (Jones, 2013). Finally, the TOMM, a forced-choice test of visual recognition memory, relies primarily on the floor effect. It is the most commonly used test of neurocognitive feigning, but unfortunately, suggested cutoff scores and sample reports have found their way to several websites, which raises questions about test security (Bauer & McCaffrey, 2006). Supplemental measures of effort for the TOMM have since been developed (Buddin et al., 2014; Gunner, Miele, Lynch, & McCaffrey, 2012). In short, these indices capture normatively unusual levels of inconsistency across the three trials of the TOMM. Both studies employ the atypical performance strategy to identify atypical learning and memory patterns, and have yielded slightly increased sensitivities over the traditional TOMM cutoff scores while maintaining high specificity. Vickery and colleagues (2001) conducted a meta-analysis on the aforementioned tests of feigned cognitive impairment (except the VSVT). The 21-Item Test showed an impressive average specificity of 100% but a low sensitivity of 22%. The meta-analysis also revealed large effect sizes for the 21-Item Test (mean d = 1.46). In contrast, the 15-Item Test (Rey, 1964) performed less impressively, especially with subtle feigning (Guilmette, Hart, Giuliano, & Leininger, 1994). It was also

138

I I .   D i a g n o s t ic I ss u e s

less successful with genuine neurocognitive disorders, including mental retardation (Goldberg & Miller, 1986; Schretlen, Brandt, Krafft, & van Gorp.1991). In an updated meta-analysis, Sollman and Berry (2011) replicated some of the prior findings. They included neurocognitive feigning tests only if at least three contrasts of criterion-supported honest patient groups and feigners were available. As a result of this rigorous standard, the measures differed somewhat from the 2001 study. In short, the VSVT, TOMM, WMT, Letter Memory Test, and Medical Symptom Validity Test all showed large effect sizes (mean d = 1.55). Mean specificity was 0.90, and mean sensitivity was 0.69. Notably, these measures, and the vast majority of all cognitive feigning measures, focus solely on recognition memory. It is reasonable to question the degree to which such tests measure effort in other cognitive domains. Indeed, their clinical utility rests solely on the untested premise that they do. Constantinou, Bauer, Ashendorf, Fisher, and McCaffery (2005) provide a preliminary answer to this question in the mTBI population. They found that though poor performance on the TOMM correlated significantly with multiple domain scores from tests of intellectual functioning and from the Halstead–Reitan Neuropsychological Test Battery, the amount of variance in neuropsychological performance accounted for by the TOMM was small (r2 = .19–.30). This supports the use of the TOMM as a test of effort as opposed to a measure of various cognitive abilities.

EVALUATION OFFEIGNEDTBI: ACASEEXAMPLE CaseFindings A 43-year-old right-handed man (FM) was exiting his pickup truck when another pickup struck his vehicle from behind at 5mph, causing the door frame to strike him in the back of the head. Witnesses noted that although he did not lose consciousness, he was confused. When paramedics arrived 10 minutes later, his GCS score was 15. His past medical history was noncontributory, and CT of the head was negative, but he had a headache. He was discharged from the emergency department the same day with instructions to rest, and was prescribed naproxen for pain. When he presented for neuropsychological assessment 12 months later as part of a workers’ compensation evaluation, he reported LOC of 30 minutes and

posttraumatic amnesia of 5–7 days. FM claimed disabling attention deficits, memory problems, anxiety, and insomnia. He complained to his primary care physician (PCP) that his cognitive symptoms were worsening and said that he had never experienced any symptoms before the accident. FM indicated that he “deserved some kind of restitution.” As part of an independent medical examination (IME) 2 years postaccident, FM was administered a comprehensive neuropsychological battery. He now reported that he had been knocked out “for at least an hour” and that he “could not remember anything anymore.” His scores on the TOMM yielded a PPP of 94.3% and negative predictive power (NPP) of 87.4% (assuming a malingering base rate of 30%). Contrary to well-known learning principles, FM’s performance curve on the VIP revealed a slight improvement in performance in the middle of the curve. Finally, the RBS scale from the MMPI-2-RF strongly suggested that he exaggerated cognitive and somatic symptoms. The degree to which his scores comport with those expected in light of injury severity was also assessed. Specifically, his scores were compared to data from Volbrecht and colleagues (2000) showing that injury severity is significantly correlated with neuropsychological test performance. For illustration, Table 7.3 compares FM’s transformed scores to the T-scores and standard deviations from selected tests taken by genuine patients with TBI. The severity of deficits increases in a dose– response curve as LOC increases. FM’s T-score of 22 on the Rey Auditory Verbal Learning Test (RAVLT) falls within the range of patients whose LOC was beyond 2 weeks, which is clearly incongruent with what was reported. Also, his perceptual reasoning score (Perceptual Reasoning Index [PRI] from the WAIS-IV) would suggest that he had been unconscious for at least 2 weeks.

CaseConceptualization FM stands to gain monetarily if he is found to be disabled. The neuropsychological evaluation included specific tests of feigning and analyzed patterns of performance to assess response bias. The medical history was carefully reviewed in order to identify incongruities between reported and observed symptoms. Finally, the claimant’s test scores were compared to those of known groups of patients with brain injury. FM reported that his symptoms (both cognitive and psychological) had worsened over time, which is inconsistent with the

7. Malingered Traumatic Brain Injury 139 TABLE 7.3.  Severity Indexing of Selected Scores from Neuropsychological Testing in the Case ofFM

T-scores from Volbrecht et al. (2000) Test RAVLT—Total

FM’s T-scores

LOC 29 days

22

40

34

26

19 33

Trails A (Time)

32

43

43

34

Trails B (Time)

40

46

48

32

37

WAIS-IV PRI

35

49

46

36

35

RCFT—Immediate

36

46

42

37

25

RCFT—Delayed

18

44

39

36

23

Note. Data are distilled from the case and Volbrecht et al. (2000). LOC, loss of consciousness; WAIS-IV, Wechsler Adult Intelligence Scale, 4th Edition; PRI, Perceptual Reasoning Index; RAVLT, Rey Auditory Verbal Learning Test; RCFT, Rey Complex Figure Test.

course of recovery from mTBI, and suggests that his responses do not reflect the effects of the mTBI itself. The clinical profile from the MMPI-2-RF was consistent with exaggerated cognitive and somatic problems. FM’s blanket dismissal of preaccident problems is consistent with research suggesting that patients with mTBI often overestimate their premorbid functioning (Greiffenstein, Baker, & JohnsonGreene, 2002), which leads to inflated and unrealistic expectations and subsequent dissatisfaction with cognitive performance following concussion. The patient was educated by his PCP about the expected symptoms of PCS. Questions remain about whether or not providing such information increases or decreases symptomatology. Mittenberg, DiGuilio, Perrin, and Bass (1992) demonstrated that psychoeducation about the course of mTBI can decrease recovery time and increase satisfaction with outcome. On the other hand, inadvertently contributing to a patient’s heightened illness perceptions (e.g., by suggesting that persisting symptoms are due to brain damage and that multiple referrals are needed) likely worsens outcome (Whittaker et al., 2007). Finally, FM voiced themes of injustice regarding his decision to litigate. Perceived injustice has been shown to be associated with poor outcome in various disability and workers’ compensation claims, and litigation involving whiplash, spinal cord, and low back pain (Yakobov, Scott, Thibault, & Sullivan, 2016; Tait & Chibnall, 2016; Trost, Mondon, Buelow, Boals, & Scott, 2016). Some investigators (e.g., Bender & Matusewicz, 2013; Silver, 2012) have theorized an equally potent role of these perceptions in protracted cases of PCS.

CLINICALAPPLICATIONS The critical role of malingering detection in neuropsychological assessment has been promulgated by both major academies in neuropsychology; moreover, the burden of proof lies with the clinician to justify why such testing would not be conducted (National Academy of Neurpsychology [NAN] Position Paper: Bush et al., 2005; AACN Consensus Statement: Heilbronner et al., 2009). Despite this requirement, assessment methods are not standardized. On the contrary, the methods of detection (e.g., observation, multiscale inventories, indices, intra- and intertest performance) vary markedly across practitioners and settings. There are dozens of published “malingering tests,” but in reality there is no such thing as a “malingering test.” These tests are largely tests of effort, and effort can be suboptimal for a number of reasons. It is the task of the forensic neuropsychologist to link poor effort with the intent to fake deficits. Relatedly, the role of bias in the examiner has not been studied nearly as extensively as it has in patients. Preliminary research (e.g., Murrie et al., 2009) strongly suggests that this area warrants further study within the broad context of forensic examination, but perhaps particularly so in malingering detection.

A Word ofCaution: UtilityEstimates andResearchDesigns In the assessment of feigned cognitive impairment, the most useful question is typically something akin to “Given this test score, how likely is it that the patient is feigning?” This question is

140

II .   D i a g n o s t ic I ss u e s

addressed by PPP, which is influenced by the base rate of malingering. Greve and Bianchini (2004) have stated that PPP and specificity should be emphasized above other utility estimates in order to reduce the number of false-positive errors. Valid and reliable research data can only be obtained from research employing sound designs. Rogers (Chapter 2, this volume) has cautioned clinicians not to confound litigation status with malingering (i.e., differential prevalence designs); a meta-analysis of the MMPI-2 and malingering (Rogers, Sewell, Martin, & Vitacco, 2003) illustrates this point nicely. Litigation status played a minor role in producing relatively small effect sizes that were overshadowed by diagnostic differences (e.g., genuine profiles of schizophrenia vs. depression). Ross, Putnam, and Adams (2006) made the same case with brain-injured patients; although both incomplete effort and psychological disturbance were good predictors of neuropsychological performance, compensation-seeking status was not. A superior design is the criterion-group design, in which multiple indicators (e.g., effort test failures and presence of external incentive) are used to establish specific group membership and validate tests of feigned cognitive impairment in those groups.

IncrementalValidity Studies have shown that using more than one test of feigning (Larrabee, 2012) and multiple detection strategies (Bender & Rogers, 2004) can improve detection. To the extent possible, each feigning-detection test in and of itself should measure largely independent constructs and should have good classification rates. For example, the TOMM (Tombaugh, 1997) is a test of visual recognition memory with very good classification rates that primarily uses the floor effect strategy. Ideally, the forensic neuropsychologist should complement such a feigning measure with an equally accurate feigning test of something other than recognition memory, such as the VIP (Frederick & Foster, 1991). When deciding which tests to combine, neuropsychologists should remember that using an additional test with a lower hit rate might decrease the accuracy of a single, more effective test. The practice of chaining likelihood ratios (LRs; also known as establishing posterior probabilities) may provide incremental validity. LRs have come into favor recently and appear to be a strongly positive step toward improving the transparency

of the statistics behind accuracy rates. In short, the LR considers both the number of findings for poor and good effort. For a review, see Bender and Frederick (Chapter 3, this volume).

Multistep Approach totheEvaluation ofMalingeredTBI Many of the decisions during an evaluation of mTBI must be rendered on a case-by-case basis. Nevertheless, all cases should involve a determination of the source or sources of the symptoms. At the outset, the clinician must try to answer the fundamental question: Was there a concussion? This determination may be difficult in itself, but probably the most difficult step is deciding whether the symptoms are actually attributable to the reported injury (Ruff & Richardson, 1999). Plus, the determination that the injury did not cause the symptoms is a multistep process that involves both qualitative and quantitative methods (see Table 7.4). The Slick and colleagues (1999) criteria for malingered neurocognitive dysfunction provided a useful heuristic for classifying individuals as malingerers but may have been both too conservative and too liberal in places (see Rogers & Bender, 2013; Rogers, Bender, & Johnson, 2011b). The revised criteria (Slick & Sherman, 2013) have yet to be validated but appear to be a step toward a fuller conceptual understanding of malingering and multifactorial methods for its detection. See Bender and Frederick (Chapter 3, this volume) for a description of how the new criteria differ from the original. Forensic neuropsychologists face major challenges with regard to mTBI, PCS, and malingering. Mild TBI is the most common diagnosis seen in forensic cases, and most of the features of PCS are nonspecific. mTBI is difficult to diagnose, partly because the most easily identified markers of injury (radiological evidence of contusion or hematoma) are exclusionary. Multiple conditions can masquerade as PCS, plus there are many reasons for a patient to put forth questionable effort during testing. Pain, depression, and symptoms of PTSD must be ruled out as potential causes for low scores on testing. Each of these conditions may also be malingered and may require an evaluation of their veracity. Moreover, as I discussed earlier in this chapter, multiple psychological and social factors may contribute to protracted recovery from PCS and be conflated with malingering if not adequately assessed.

7. Malingered Traumatic Brain Injury 141

TABLE 7.4.  Sequential Decision Model forMalingering Detection

1. Determine whether there is a brain injury. Evaluate the following: a. Were signs of brain injury present at the scene (e.g., LOC, confusion)? b. Are medical workups (e.g., imaging) suggestive? c. Does the postacute symptom constellation fit with TBI? d. Are the neuropsychological evaluation results suggestive of TBI? e. Are there viable alternative explanations for the symptoms? 2. If Step 1 indicates that a TBI occurred, whether or not the reported symptoms are attributable to the injury must be determined. Evaluate the following: a. History and course/time gradient of the symptom: i. Are patient and collateral reports consistent and congruent with TBI? ii. Does onset of symptoms coincide with the injury? Did their evolution follow the course seen in TBI? b. The impact of preexisting symptoms and comorbid mental disorders such as depression and somatoform disorders. c. Whether any symptoms are being malingered (i.e., exaggeration or prolongation of genuine symptoms) i. Results of validity tests. ii. Determine impact of psychosocial factors, such as litigation, attribution error, and good old days bias on effort and motivation. 3. If Step 1 does not indicate a TBI, then the clinician may conclude that symptom etiology is non-TBI related. a. Therefore, symptoms are associated with another medical condition, psychological disturbance, or malingering. b. However, duration and degree of debilitation are critical here: i. For instance, symptoms of PCS at 72 hours can be clinically debilitating and exist without frank neurological signs. ii. In contrast, PCS symptoms at 12 months posttreatment without evidence of trauma strongly suggest non-neurological mechanisms.

The application of conceptually based strategies to malingering detection (see Rogers, Harrell, & Liff, 1993; Rogers & Correa, 2008; Rogers, Chapter 2, this volume) appears to have improved detection accuracy, but there is still room for further improvement. Whether explicitly stated or not, fairly simple applications of the floor effect strategy still pervade the literature, largely because it is easy to understand and use. More sophisticated applications appear promising. For instance, Backhaus, Fichtenberg, and Hanks (2004) demonstrated the potential in their archival study that used a normative floor effect to detect suboptimal effort. In essence, standards expected in moderate to severe brain injury are applied to the performances of patients with mTBI. This approach (which is similar to severity indexing) resulted in excellent classification rates. Empirical studies have begun to show that hit rates improve when certain feigning detection scales are combined (e.g., Bender & Rogers, 2004; Nelson et al., 2003; Victor, Boone, Serpa, Buehler, & Ziegler, 2009). Larrabee (2014) has argued that the specificity of a finding of malingering increases when two or more tests of feigning are employed within a neuropsychological battery of tests. However, whether the approach should take into account not only the number of tests failed but also the number of tests administered has not been agreed upon. For dissenting views of specificity when using two or more effort tests, see Berthelson, Mulchan, Odland, Miller, and Mittenberg (2013), Bilder, Sugar, and Hellemann (2014), and Odland, Lammy, Martin, Grote, and Mittenberg (2015). Future research should include more sophisticated methods of distinguishing malingered mTBI from genuine neurological and psychiatric conditions, especially when multiple comorbidities are involved. To date, the vast majority of research has compared feigned mTBI with genuine mTBI, whereas less attention has been paid to the performances of patients with other genuine disorders. To achieve maximum clinical utility, future measures must be able to rule out these other conditions as well.

Summary Guidelines forForensicNeuropsychologists •• Be knowledgeable about the injury characteristics present at the time of the injury. This is important, because durations of LOC and PTA, GCS score, and neuroimaging in the acute period are reasonably good predictors of recovery from

142

II .   D i a g n o s t ic I ss u e s

mTBI (especially when used in combination). In contrast, postacute indicators of severity, including neuropsychological test scores, are less effective predictors in most cases involving mTBI. •• Be familiar with the natural recovery curves of the different severities of TBI and the expected outcomes at various points in time. Some degree of recovery is expected in almost all injuries to the brain, regardless of severity. But complete recovery is expected in almost all cases of single, uncomplicated mTBI. At no time would a worsening of symptoms be expected following mTBI, and when present almost certainly indicates the onset of symptoms with a psychological etiology (in the absence of an evolving bleed). •• Systematically evaluate the role of psychological and litigation factors, including those unrelated to (or only marginally so) external gain: jurisogenic factors, perceptions of injustice, biased recall about preexisting functioning, and attribution errors regarding the cause of their symptoms (Bender & Matusewicz, 2013; Silverberg et al., 2016). •• Be familiar with PCS and its masquerading syndromes, such as anxiety, anxiety sensitivity, PTSD, and depression. •• Use multiple tests of feigning (both embedded and freestanding tests) over the course of the evaluation. •• Only use effort tests that have been validated in the patient population in question. •• Do not rule out a priori the possibility of genuine mTBI and/or PCS. •• Be mindful of the apparent imbalance between the large number of tests available to assess effort and the less impressive knowledge base regarding what “effort” actually means.

REFERENCES Alexander, M. P. (1995). Mild traumatic brain injury: Pathophysiology, natural history, and clinical management. Neurology, 45(7), 1253–1260. Al-Ozairi, A., McCullagh, S., & Feinstein, A. (2015). Predicting posttraumatic stress symptoms following mild, moderate, and severe traumatic brain injury. Journal of Head Trauma Rehabilitation, 30(4), 283– 289. Alway, Y., Gould, K. R., Johnston, L., McKenzie, D., & Ponsford, J. (2016). A prospective examination of Axis I psychiatric disorders in the first 5 years fol-

lowing moderate to severe traumatic brain injury. Psychological Medicine, 46(6), 1331–1341. American Congress of Rehabilitation Medicine. (1993). Definition of mild traumatic brain injury. Journal of Head Trauma Rehabilitation, 8, 86–87. American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: Author. Andrikopoulos, J., & Greiffenstein, M. F. (2011). Something to talk about? In G. L. Larrabee (Ed.), Forensic neuropsychology: A scientific approach (pp.365–400). New York: Oxford University Press. Arbisi, P. A., & Butcher, J. (2004). Failure of the FBS to predict malingering of somatic symptoms: Response to critiques by Greve and Bianchini and Lees Haley and Fox. Archives of Clinical Neuropsychology, 19(3), 341–345. Ashendorf, L., O’Bryant, S. E., & McCaffrey, R. J. (2003). Specificity of malingering detection strategies in older adults using the CVLT and WCST. The Clinical Neuropsychologist, 17, 255–262. Axelrod, B. N., Fichtenberg, N. L., Millis, S. R., & Wertheimer, J. C. (2006). Detecting incomplete effort with digit span from the WAIS-III. The Clinical Neuropsychologist, 20, 513–523. Babikian, T., Boone, K. B., Lu, P., & Arnold, G. (2006). Sensitivity and specificity of various Digit Span scores in the detection of suspect effort. The Clinical Neuropsychologist, 20, 145–159. Backhaus, S. L., Fichtenberg, N. L., & Hanks, R. A. (2004). Detection of sub-optimal performance using a floor effect strategy in patients with traumatic brain injury. The Clinical Neuropsychologist, 18, 591–603. Barth, J. T., Macciocchi, S., Boll, T., Giordani, B., & Rimel, R. (1983). Neuropsychological sequelae of minor head injury. Neurosurgery, 13, 529–533. Bauer, L., & McCaffery, R. J. (2006). Coverage of the TOMM, VSVT, and WMT on the Internet: Is test security threatened? Archives of Clinical Neuropsychology, 21, 121–126. Bauer, L., Yantz, C. L., & Ryan, L. M. (2005). An examination of the California Verbal Learning Test II to detect incomplete effort in a traumatic brain-injury sample. Applied Neuropsychology, 12(4), 202–207. Bender, S. D., Barth, J. T., & Irby, J. (2004). Historical perspectives. In M. Lovell, R. Echemendia, J. T. Barth, & M. W. Collins (Eds.), Traumatic brain injury in sports: An international neuropsychological perspective (pp.1–21). Lisse, The Netherlands: Swets & Zeitlinger. Bender, S. D., & Matusewicz, M. (2013). PCS, iatrogenic symptoms, and malingering following concussion. Psychological Injury and Law, 6(2), 113–121. Bender, S. D., & Rogers, R. (2004). Detection of neurocognitive feigning: Development of a multi-strategy assessment. Archives of Clinical Neuropsychology, 19, 49–60. Ben-Porath, Y. S., & Tellegen, A. (2008). MMPI-2-RF:

7. Malingered Traumatic Brain Injury 143

Manual for administration, scoring and interpretation. Minneapolis: University of Minnesota Press. Berthelson, L., Mulchan, S. S., Odland, A. P., Miller, L. J., & Mittenberg, W. (2013). False positive diagnosis of malingering due to the use of multiple effort tests. Brain Injury, 27, 909–916. Bianchini, K. J., Aguerrevere, L. E., Guise, B. J., Ord, J. S., Etherton, J. L., Meyers, J. E., et al. (2014). Accuracy of the modified somatic perception questionnaire and pain disability index in the detection of malingered pain-related disability in chronic pain. The Clinical Neuropsychologist, 28(8), 1376–1394. Bianchini, K. J., Greve, K. W., & Glynn, G. (2005). On the diagnosis of malingered pain-related disability: Lessons from the cognitive malingering research. The Spine Journal, 5, 404–417. Bianchini, K. J., Greve, K. W., & Love, J. M. (2003). Definite malingered neurocognitive dysfunction in moderate/severe traumatic brain injury. The Clinical Neuropsychologist, 17(4), 574–580. Bigler, E. D. (2008). Neuropsychology and clinical neuroscience of persistent post-concussive syndrome. Journal of the International Neuropsychological Society, 14(1), 1–22. Bigler, E. D., & Maxwell, W. L. (2012). Neuropathology of mild traumatic brain injury: Relationship to neuroimaging findings. Brain Imaging and Behavior, 6(2), 108–136. Bilder, R. B., Sugar, C. A., & Hellemann, G. S. (2014). Cumulative false positive rates given multiple performance validity tests: Commentary on Davis & Millis (2014) and Larrabee (2014). The Clinical Neuropsychologist, 28(4), 1212–1223. Binder, L. M. (1997). A review of mild head trauma: Part II. Clinical applications. Journal of Clinical and Experimental Neuropsychology, 19, 432–457. Binder, L. M., Iverson, G. L., & Brooks, B. L. (2009). To err is human: “Abnormal” neuropsychological scores and variability are common in healthy adults. Archives of Clinical Neuropsychology, 24(1), 31–46. Binder, L. M., & Rohling, M. (1996). Money matters: A meta-analytic review of the effects of financial incentives on recovery after closed head injury. American Journal of Psychiatry, 153, 7–10. Binder, L. M., Rohling, M., & Larrabee, G. J. (1997). A review of mild head trauma: Part I. Meta-analytic review of neuropsychological studies. Journal of Clinical and Experimental Neuropsychology, 19, 421–431. Bolter, T., Picano, J. J., & Zych, K. (1985, October). Item error frequencies on the Halstead Category Test: An index of performance validity. Paper presented at the annual meeting of the National Academy of Neuropsychology, Philadelphia, PA. Boone, K. B. (2007). Assessment of feigned cognitive impairment. New York: Guilford Press. Boone, K. B., & Lu, P. (2003). Noncredible cognitive performance in the context of severe brain injury. The Clinical Neuropsychologist, 17(2), 244–254.

Bryant, R. A., Creamer, M., O’Donnell, M., Silove, D., Clark, C. R., & McFarlane, A. C. (2009). Post-traumatic amnesia and the nature of post-traumatic stress disorder after mild traumatic brain injury. Journal of the International Neuropsychological Society, 15(6), 862–867. Buddin, W. H., Jr., Schroeder, R. W., Hargrave, D. D., Von Dran, E. J., Campbell, E. B., Brockman, C. J., et al. (2014). An examination of the frequency of invalid forgetting on the Test of Memory Malingering. The Clinical Neuropsychologist, 28(3), 525–542. Busch, C. R., & Alpern, H. P. (1998). Depression after mild traumatic brain injury: A review of current research. Neuropsychology Review, 8(2), 95–108. Bush, S., Ruff, R., Troster, A., Barth, J., Koffler, S., Pliskin, N., et al. (2005). Symptom validity assessment: Practice issues and medical necessity: NAN Planning and Policy Committee. Archives of Clinical Neuropsychology, 20, 419–426. Butcher, J. N., Gass, C. S., Cumella, E., Kally, Z., & Williams, C. L. (2008). Potential for bias in MMPI-2 assessments using the Fake Bad Scale (FBS). Psychological Injury and Law, 1(3), 191–209. Carone, D. A., & Bush, S. S. (2013). Mild traumatic brain injury: Symptom validity assessment and malingering. New York: Springer. Chong, C. D., & Schwedt, T. J. (2015). White matter damage and brain network alterations in concussed patients: A review of recent diffusion tensor imaging and resting-state functional connectivity data. Current Pain and Headache Reports, 19(5), 1–10. Cisler, J. M., & Koster, E. H. (2010). Mechanisms of attentional biases towards threat in anxiety disorders: An integrative review. Clinical Psychology Review, 30(2), 203–216. Clarke, P., & MacLeod, C. (2013). The impact of anxiety on cognitive task performance. In P. A. Arnett (Ed.), Secondary influences on neuropsychological test performance (pp.94–116). New York: Oxford University Press. Constantinou, M., Bauer, L., Ashendorf, L., Fisher, J., & McCaffery, R. J. (2005). Is poor performance on recognition memory effort measures indicative of generalized poor performance on neuropsychological tests? Archives of Clinical Neuropsychology, 20, 191–198. Cornell, D. G., & Hawk, G. L. (1989). Clinical presentation of malingerers diagnosed by experienced forensic psychologists. Law and Human Behavior, 13(4), 375–383. Curtiss, G., & Vanderploeg, R. D. (2000). Prevalence rates for neuropsychological malingering indexes in traumatic brain injury. American Psychological Association Division 40 Newsletter, 18, 9–13. Davidoff, D. A., Kessler, H. R., Laibstain, D. F., & Mark, V. H. (1988). Neurobehavioral sequelae of minor head injury: A consideration of post-concussive syndrome versus post-traumatic stress. Cognitive Rehabilitation, 6, 8–13.

144

II .   D i a g n o s t ic I ss u e s

Davies, R., & McMillan, T. (2005). Opinion about postconcussion syndrome in health professionals. Brain Injury, 19(11), 941–947. Dearth, C. S., Berry, D. T., & Vickery, C. D. (2005). Detection of feigned head injury symptoms on the MMPI-2 in head injured patients and community controls. Archives of Clinical Neuropsychology, 20(1), 95–110. Deb, S., Lyons, I., Koutzoukis, C., Ali, I., & McCarthy, G. (1999). Rate of psychiatric illness 1 year after traumatic brain injury. American Journal of Psychiatry, 156, 374–378. Delis, D. C., Kramer, J. H., Kaplan, E., & Ober, B. A. (2000). CVLT-II: California Verbal Learning Test: Adult Version. San Antonio, TX: Psychological Corporation. Delis, D. C., Kramer, J. H., Kaplan, E., & Thompkins, B. A. O. (1987). CVLT: California Verbal Learning Test—Adult Version: Manual. San Antonio, TX: Psychological Corporation. Delis, D. C., & Wetter, S. R. (2007). Cogniform disorder and cogniform condition: Proposed diagnoses for excessive cognitive symptoms. Archives of Clinical Neuropsychology, 22(5), 589–604. Delouche, A., Attyé, A., Heck, O., Grand, S., Kastler, A., Lamalle, L., et al. (2016). Diffusion MRI: Pitfalls, literature review and future directions of research in mild traumatic brain injury. European Journal of Radiology, 85(1), 25–30. Demakis, G. J., Gervais, R. O., & Rohling, M. L. (2008). The effect of failure on cognitive and psychological symptom validity tests in litigants with symptoms of post-traumatic stress disorder. The Clinical Neuropsychologist, 22(5), 879–895. Demakis, G. J., Hammond, F., Knotts, A., Cooper, D. B., Clement, P., Kennedy, J., et al. (2007). The Personality Assessment Inventory in individuals with traumatic brain injury. Archives of Clinical Neuropsychology, 22(1), 123–130. Dikmen, S. S., Machamer, J. E., Fann, J., & Temkin, N. R. (2010). Rates of symptom reporting following traumatic brain injury. Journal of the International Neuropsychological Society, 16(3), 401–411. Dikmen, S. S., Machamer, J. E., Winn, H. R., & Temkin, N. R. (1995). Neuropsychological outcome at 1-year post head injury. Neuropsychology, 9(1), 80–90. Donnell, A. J., Kim, M. S., Silva, M. A., & Vanderploeg, R. D. (2012). Incidence of postconcussion symptoms in psychiatric diagnostic groups, mild traumatic brain injury, and comorbid conditions. The Clinical Neuropsychologist, 26(7), 1092–1101. Drob, S. L., Meehan, K. B., & Waxman, S. E. (2009). Clinical and conceptual problems in the attribution of malingering in forensic evaluations. Journal of the American Academy of Psychiatry and the Law, 37(1), 98–106. Etherton, J. L., Bianchini, K. J., Ciota, M. A., Heinly, M. T., & Greve, K. W. (2006). Pain, malingering,

and the WAIS-III Working Memory Index. The Spine Journal, 6, 61–71. Etherton, J. L., Bianchini, K. J., Greve, K. W., & Ciota, M. A. (2005). Test of memory malingering performance is unaffected by laboratory-induced pain: Implications for clinical use. Archives of Clinical Neuropsychology, 20(3), 375–384. Etherton, J. L., Bianchini, K. J., Greve, K. W., & Heinly, M. T. (2005). Sensitivity and specificity of reliable digit span in malingered pain-related disability. Assessment, 12(2), 130–136. Fakhran, S., Qu, C., & Alhilali, L. M. (2016). Effect of the suboccipital musculature on symptom severity and recovery after mild traumatic brain injury. American Journal of Neuroradiology, 37(8), 1556–1560. Faul, M., Xu, L., Wald, M. M., & Coronado, V. (2010). Traumatic brain injury in the United States: Emergency department visits, hospitalizations and deaths, 2002– 2006. Atlanta, GA: Centers for Disease Control and Prevention, National Center for Injury Prevention and Control. Forrest, T. J., Allen, D. N., & Goldstein, G. (2004). Malingering Indexes for the Halstead Category Test. The Clinical Neuropsychologist, 18(2), 334–347. Fox, D. D., Lees-Haley, P., Earnest, K., & Dolezal-Wood, S. (1995). Base rates of post-concussive symptoms in health maintenance organization patients and controls. Neuropsychology, 9(4), 606–611. Frederick, R. I. (1997). The Validity Indicator Profile. Minneapolis, MN: National Computer Systems. Frederick, R. I., & Foster, H. G. (1991). Multiple measures of malingering on a forced-choice test of cognitive ability. Psychological Assessment: A Journal of Consulting and Clinical Psychology, 3(4), 596. Galatzer-Levy, I. R., & Bryant, R. A. (2013). 636,120 ways to have posttraumatic stress disorder. Perspectives on Psychological Science, 8(6), 651–662. Gervais, R. O., Ben-Porath, Y. S., Wygant, D. B., & Green, P. (2007). Development and validation of a Response Bias Scale (RBS) for the MMPI-2. Assessment, 14(2), 196–208. Gierok, S. D., Dickson, A. L., & Cole, J. A. (2005). ­Performance of forensic and non-forensic adult psychiatric inpatients on the Test of Memory Malingering. Archives of Clinical Neuropsychology, 20(6), 755–760. Giza, C. C., & Hovda, D. A. (2004). The pathophysiology of traumatic brain injury. In M. R. Lovell, R. J. Echemendia, J. T. Barth, & M. W. Collins (Eds.), Traumatic brain injury in sports: An international neuropsychological perspective (pp.45–70). Lisse, The Netherlands: Swets & Zeitlinger. Goldberg, J. O., & Miller, H. R. (1986). Performance of psychiatric inpatients and intellectually deficient individuals on a task that assesses validity of memory complaints. Journal of Clinical Psychology, 42, 792–795. Gordon, S. N., Fitzpatrick, P. J., & Hilsabeck, R. C.

7. Malingered Traumatic Brain Injury 145

(2011). No effect of PTSD and other psychiatric disorders on cognitive functioning in veterans with mild TBI. The Clinical Neuropsychologist, 25(3), 337–347. Gouvier, W., Uddo-Crane, M., & Brown, L. M. (1988). Base rates of postconcussional symptoms. Archives of Clinical Neuropsychology, 3, 273–278. Green, P., Iverson, G. L., & Allen, L. (1999). Detecting malingering in head injury litigation with the Word Memory Test. Brain Injury, 13(10), 813–819. Green, P., Rohling, M. L., & Lees-Haley, P. (2001). Effort has a greater effect on test scores than severe brain injury in compensation claimants. Brain Injury, 15(12), 1045–1060. Greiffenstein, M., Baker, W., Axelrod, B., Peck, E., & Gervais, R. (2004). The Fake Bad Scale and MMPI2 F-family in detection of implausible psychological trauma claims. The Clinical Neuropsychologist 18(4), 573–590. Greiffenstein, M., Baker, W. J., & Gola, T. (1994). Validation of malingered amnesia measures with a large clinical sample. Psychological Assessment, 6, 218–224. Greiffenstein, M. F., Baker, W. J., & Johnson-Greene, D. (2002). Actual versus self-reported scholastic achievement of litigating postconcussion and severe closed head injury claimants. Psychological Assessment, 14(2), 202–208. Greve, K. W., & Bianchini, K. J. (2004). Setting empirical cut-offs on psychometric indicators of negative response bias: A methodological commentary with recommendations. Archives of Clinical Neuropsychology, 19, 533–541. Greve, K. W., Bianchini, K. J., & Roberson, T. (2007). The Booklet Category Test and malingering in traumatic brain injury: Classification accuracy in known groups. The Clinical Neuropsychologist, 21(2), 318–337. Greve, K. W., Ord, J. S., Bianchini, K. J., & Curtis, K. L. (2009). Prevalence of malingering in patients with chronic pain referred for psychologic evaluation in a medico-legal context. Archives of Physical Medicine and Rehabilitation, 90(7), 1117–1126. Greve, K. W., Ord, J., Curtis, K. L., Bianchini, K. J., & Brennan, A. (2008). Detecting malingering in traumatic brain injury and chronic pain: A comparison of three forced-choice symptom validity tests. The Clinical Neuropsychologist, 22(5), 896–918. Guilmette, T. J., Hart, K. J., Giuliano, A. J., & Leininger, B. E. (1994). Detecting simulated memory impairment: Comparison of the Rey Fifteen-Item Test and the Hisco*ck Forced Choice Procedure. The Clinical Neuropsychologist, 8, 283–294. Gunner, J. H., Miele, A. S., Lynch, J. K., & McCaffrey, R. J. (2012). The Albany Consistency Index for the Test of Memory Malingering. Archives of Clinical Neuropsychology, 27(1), 1–9. Gunstad, J., & Suhr, J. A. (2001). “Expectation as etiology” versus “the good old days”: Postconcussion syndrome symptom reporting in athletes, headache

sufferers, and depressed individuals. Journal of the International Neuropsychological Society, 7(3), 323–333. Hall, R. C. W., & Hall, R. C. W. (2012). Compensation neurosis: A too quickly forgotten concept? Journal of the American Academy of Psychiatry and the Law, 40(3), 390–398. Harvey, A. G., & Bryant, R. A. (1998). Predictors of acute stress following mild traumatic brain injury. Brain Injury, 12(2), 147–154. Heilbronner, R. L., Sweet, J. J., Morgan, J. E., Larrabee, G. L., Millis, S. R., & Conference Participants. (2009). American Academy of Clinical Neuropsychology consensus conference statement on the neuropsychological assessment of effort, response bias, and malingering, The Clinical Neuropsychologist, 23(7), 1093–1129. Hill, B. D., Rohling, M. L., Boettcher, A. C., & Meyers, J. E. (2013). Cognitive intra-individual variability has a positive association with traumatic brain injury severity and suboptimal effort. Archives of Clinical Neuropsychology, 28(7), 640–648. Hoffman, J. M., Dikmen, S., Temkin, N., & Bell, K. R. (2012). Development of posttraumatic stress disorder after mild traumatic brain injury. Archives of Physical Medicine and Rehabilitation, 93(2), 287–292. Hou, R., Moss-Morris, R., Peveler, R., Mogg, K., Bradley, B. P., & Belli, A. (2012). When a minor head injury results in enduring symptoms: A prospective investigation of risk factors for postconcussional syndrome after mild traumatic brain injury. Journal of Neurology, Neurosurgery and Psychiatry, 83(2), 217–223. Institute of Medicine. (2009). Gulf War and Health Committee: Vol. 7. Long-term consequences of traumatic brain injury. Washington, DC: National Academies Press. Iverson, G. L. (2003). Detecting malingering in civil forensic evaluations. In A. M. Horton, Jr., & L. C. Hartlage (Eds.), Handbook of forensic neuropsychology (pp.137–177). New York: Springer. Iverson, G. L. (2006). Ethical issues associated with the assessment of exaggeration, poor effort, and malingering. Applied Neuropsychology, 13(2), 77–90. Iverson, G. L., & Binder, L. M. (2000). Detecting exaggeration and malingering in neuropsychological assessment. Journal of Head Trauma Rehabilitation, 15, 829–858. Iverson, G. L., Brooks, B. L., Ashton, V. L., & Lange, R. (2010). Interview versus questionnaire symptom reporting in people with the postconcussion syndrome. Journal of Head Trauma Rehabilitation, 25(1), 23–30. Iverson, G. L., Brooks, B. L., & Holdnack, J. A. (2008). Misdiagnosis of cognitive impairment in forensic neuropsychology. In R. L. Heilbronner (Ed.), Neuropsychology in the courtroom: Expert analysis of reports and testimony (pp.243–266). New York: Guilford Press. Iverson, G. L., & McCracken, L. M. (1997). Postconcus-

146

II .   D i a g n o s t ic I ss u e s

sive symptoms in persons with chronic pain. Brain Injury, 11, 783–790. Iverson, G. L., & Tulsky, D. S. (2003). Detecting malingering on the WAIS-III: Unusual Digit Span performance patterns in the normal population and in clinical groups. Archives of Clinical Neuropsychology, 18, 1–9. Iverson, G. L., Wilhelm, K., & Franzen, M. D. (1993). Objective assessment of simulated memory deficits: Additional scoring criteria for the 21 Items Test. Archives of Clinical Neuropsychology, 8, 235–236. Jennett, B. (2002). The Glasgow Coma Scale: History and current practice. Trauma, 4(2), 91–103. Jennett, B., & Bond, M. (1975). Assessment of outcome after severe brain damage. Lancet, 1, 480–484. Jones, A. (2013). Victoria Symptom Validity Test: Cutoff scores for psychometrically defined malingering groups in a military sample. The Clinical Neuropsychologist, 27(8), 1373–1394. Ju, D., & Varney, N. (2000). Can head injury patients simulate malingering? Applied Neuropsychology, 7(4), 201–207. Kashluba, S., Casey, J., & Paniak, C. (2006). Evaluating the utility of ICD-10 diagnostic criteria for postconcussion syndrome following mild traumatic brain injury. Journal of the International Neuropsychological Society, 12, 111–118. Kay, T. (1999). Interpreting apparent neuropsychological deficits: What is really wrong? In J. Sweet (Ed.), Forensic neuropsychology: Fundamentals and practice (pp.145–183). Lisse, The Netherlands: Swets & Zeitlinger. Keiski, M. A., Shore, D. L., Hamilton, J. M., & Malec, J. F. (2015). Simulation of TBI symptoms on the PAI: An analogue study. Assessment, 22(2), 233–247. Kewman, D. G., Vaishampayan, N., Zald, D., & Han, B. (1991). Cognitive impairment in musculoskeletal pain patients. International Journal of Psychiatry in Medicine, 21, 253–262. Kirkwood, M. W. (2015). Validity testing in child and adolescent assessment: Evaluating exaggeration, feigning, and noncredible effort. New York: Guilford Press. Koren, D., Hemel, D., & Klein, E. (2006). Injury increases the risk for PTSD: An examination of potential neurobiological and psychological mediators. CNS Spectrums, 11(8), 616–624. Koren, D., Norman, D., Cohen, A., Berman, J., & Klein, E. M. (2005). Increased PTSD risk with combat-related injury: A matched comparison study of injured and uninjured soldiers experiencing the same combat events. American Journal of Psychiatry, 162, 276–278. Kurtz, J. E., Shealy, S. E., & Putnam, S. H. (2007). Another look at paradoxical severity effects in head injury with the Personality Assessment Inventory. Journal of Personality Assessment, 88, 66–73. Landre, N., Poppe, C. J., Davis, N., Schmaus, B., & Hobbs, S. E. (2006). Cognitive functioning and postconcussive symptoms in trauma patients with and

without mild TBI. Archives of Clinical Neuropsychology, 21, 255–273. Lange, R. T., Iverson, G. L., & Rose, A. (2010). Postconcussion symptom reporting and the “good-olddays” bias following mild traumatic brain injury. Archives of Clinical Neuropsychology, 256(5), 442–450. Lange, R., Iverson, G. L., & Rose, A. (2011). Depression strongly influences postconcussion symptom reporting following mild traumatic brain injury. Journal of Head Trauma Rehabilitation, 26(2), 127–137. Larrabee, G. J. (2003a). Detection of malingering using atypical performance patterns on standard neuropsychological tests. The Clinical Neuropsychologist, 17, 410–425. Larrabee, G. J. (2003b). Exaggerated pain report in litigants with malingered neurocognitive dysfunction. The Clinical Neuropsychologist, 17, 395–401. Larrabee, G. J. (2007). Assessment of malingered neuropsychological deficits. New York: Oxford University Press. Larrabee, G. J. (2012). Forensic neuropsychology: A scientific approach (2nd ed.). New York: Oxford University Press. Larrabee, G. J. (2014). False-positive rates associated with the use of multiple performance and symptom validity tests. Archives of Clinical Neuropsychology, 29(4), 364–373. Larrabee, G. J., Greiffenstein, M. F., Greve, K. W., & Bianchini, K. J. (2007). Refining diagnostic criteria for malingering. In G. L. Larrabee’s (Ed.), Assessment of malingered neuropsychological deficits. New York: Oxford University Press. Larrabee, G. J., & Rohling, M. L. (2013). Neuropsychological differential diagnosis of mild traumatic injury. Behavioral Sciences and the Law, 31(6), 686–701. Lees-Haley, P. R., & Brown, R. S. (1993). Neuropsychological complaint base rates of 170 personal injury claimants. Archives of Clinical Neuropsychology, 8, 203–209. Lees-Haley, P. R., English, L. T., & Glenn, W. J. (1991). A fake bad scale on the MMPI-2 for personal-injury claimants. Psychological Reports, 68, 203–210. Lees-Haley, P. R., Iverson, G. L., Lange, R. T., Fox, D. D., & Allen, L. M. (2002). Malingering in forensic neuropsychology: Daubert and the MMPI-2. Journal of Forensic Neuropsychology, 3(1–2), 167–203. Levin, H. S., Mattis, S., Ruff, R. M., Eisenberg, H. M., Marshall, L. F., Tabaddor, K., et al. (1987). Neurobehavioral outcome following minor head injury: A three-center study. Neurosurgery, 66(2), 234–243. Lezak, M. D., Howieson, D. B., Bigler, E. D., & Tranel, D. (2012). Neuropsychological Assessment (5th ed.). New York: Oxford University Press. Lipman, F. D. (1962). Malingering in personal injury cases. Temple Law Quarterly, 35, 141–162. Martin, P. K., Schroeder, R. W., & Odland, A. P. (2015). Neuropsychologists’ validity testing beliefs and practices: A survey of North American professionals. The Clinical Neuropsychologist, 29(6), 741–776.

7. Malingered Traumatic Brain Injury 147

McBride, W. F., Crighton, A. H., Wygant, D. B., & Granacher, R. P. (2013). It’s not all in your head (or at least your brain): Association of traumatic brain lesion presence and location with performance on measures of response bias in forensic evaluation. Behavioral Sciences and the Law, 31(6), 779–788. Meares, S., Shores, E. A., Taylor, A. J., Batchelor, J., Bryant, R. A., Baguley, I. J., et al. (2011). The prospective course of postconcussion syndrome: The role of mild traumatic brain injury. Neuropsychology, 25(4), 454–465. Meier, T. B., Bergamino, M., Bellgowan, P. S. F., Teague, T. K., Ling, J. M., Jeromin, A., et al. (2015). Longitudinal assessment of white matter abnormalities following sports-related concussion. Human Brain Mapping, 37, 833–845. Meyers, J. E., & Diep, A. (2000). Assessment of malingering in chronic pain patients using neuropsychological tests. Applied Neuropsychology, 7(3), 133–139. Miller, H. (1961). Accident neurosis. British Medical Journal, 1, 919–925. Millis, S. R., Putnam, S., Adams, K., & Ricker, J. (1995). The California Verbal Learning Test in the detection of incomplete effort in neuropsychological evaluation. Psychological Assessment, 7, 463–471. Millis, S. R., Rosenthal, M., Novack, T., Sherer, M., Nick, T. G., Kreutzer, J. S., et al. (2001). Long-term neuropsychological outcome after traumatic brain injury. Journal of Head Trauma Rehabilitation, 16, 343–355. Mittenberg, W., DiGuilio, D., Perrin, S., & Bass, A. (1992). Symptoms following mild head injury: Expectation as aetiology. Journal of Neurology, Neurosurgery, and Psychiatry, 55, 200–204. Mittenberg, W., Patton, C., Canyock, E. M., & Condit, D. C. (2002). Base rates of malingering and symptom exaggeration. Journal of Clinical and Experimental Neuropsychology, 24, 1094–1102. Mooney, G., Speed, J., & Sheppard, S. (2005). Factors related to recovery after mild traumatic brain injury. Brain Injury, 19(12), 975–987. Morey, L. C. (1991). Personality Assessment Inventory professional manual. Lutz, FL: Psychological Assessment Resources. Morey, L. C. (1996). An interpretive guide to the Personality Assessment Inventory. Lutz, FL: Psychological Assessment Resources. Murrie, D. C., Boccaccini, M. T., Turner, D., Meeks, M., Woods, C., & Tussey, C. (2009). Rater (dis) agreement on risk assessment measures in sexually violent predator proceedings: Evidence of adversarial allegiance in forensic evaluation? Psychology, Public Policy, and Law, 15(1), 19–53. Nampiaparampil, D. E. (2008). Prevalence of chronic pain after traumatic brain injury. Journal of the American Medical Association, 300(6), 711–719. Nelson, E. A., Lickel, J. J., Sy, J. T., Dixon, L. J., & Deacon, B. J. (2010). Probability and cost biases in social phobia: Nature, specificity, and relationship to treat-

ment outcome. Journal of Cognitive Psychotherapy, 24(3), 213–228. Nelson, N. W., Boone, K., Dueck, A., Wagener, L., Lu, P., & Grills, C. (2003). Relationships between eight measures of suspect effort. The Clinical Neuropsychologist, 17, 263–272. Nelson, N. W., Sweet, J. J., & Demakis, G. J. (2006). Meta-analysis of the MMPI-2 Fake Bad Scale: Utility in forensic practice. The Clinical Neuropsychologist, 20, 39–58. Nicholson, K., & Martelli, M. F. (2007). Malingering: Traumatic brain injury. In G. Young, A. W. Kane, & K. Nicholson (Eds.), Causality of psychological injury (pp.427–475). New York: Springer. Odland, A. P., Lammy, A. B., Martin, P. K., Grote, C. L., & Mittenberg, W. (2015). Advanced administration and interpretation of multiple validity tests. Psychological Injury and Law, 8(1), 46–63. Peck, C. P., Schroeder, R. W., Heinrichs, R. J., VonDran, E. J., Brockman, C. J., Webster, B. K., et al. (2013). Differences in MMPI-2 FBS and RBS scores in brain injury, probable malingering, and conversion disorder groups: A preliminary study. The Clinical Neuropsychologist, 27(4), 693–707. Ponsford, J., Cameron, P., Fitzgerald, M., Grant, M., Mikocka-Walus, A., & Schönberger, M. (2012). Predictors of postconcussive symptoms 3 months after mild traumatic brain injury. Neuropsychology, 26(3), 304. Rees, L. M., Tombaugh, T. N., Gansler, D. A., & Moczynski, N. P. (1998). Five validation experiments of the Test of Memory Malingering (TOMM). Psychological Assessment, 10, 10–20. Resnick, P. J., West, S., & Payne, J. W. (2008). Malingering of posttraumatic disorders. In R. Rogers (Ed.), Clinical assessment of malingering and deception (pp.109–127). New York: Guildford Press. Rey, A. (1964). The clinical examination in psychology. Paris: Presses Universitaires de France. Roebuck-Spencer, T., & Sherer, M. (2008). Moderate and severe traumatic brain injury. In J. E. Morgan & J. H. Ricker (Eds.), Textbook of clinical neuropsychology (pp.411–429). New York: Psychology Press. Rogers, R., & Bender, S. D. (2013). Evaluation of malingering and related response styles. In R. K. Otto & I. B. Weiner (Eds.), Handbook of psychology: Vol. 11. Forensic psychology (2nd ed., pp.517–541). Hoboken, NJ: Wiley. Rogers, R., Bender, S. D., & Johnson, S. (2011a). A critical analysis of the MND criteria for feigned cognitive impairment: Implications for forensic practice and research. Psychological Injury and Law, 4(2), 147–156. Rogers, R., Bender, S. D., & Johnson, S. (2011b). A commentary on the MND model and the Boone critique: “Saying it doesn’t make it so.” Psychological Injury and Law, 4(2), 163–167. Rogers, R., & Correa, A. A. (2008). Determinations of malingering: Evolution from case-based methods to

148

II .   D i a g n o s t ic I ss u e s

detection strategies. Psychiatry, Psychology, and Law, 15(2), 213–223. Rogers, R., Harrell, E. H., & Liff, C. D. (1993). Feigning neuropsychological impairment: A critical review of methodological and clinical considerations. Clinical Psychology Review, 13, 255–274. Rogers, R., Salekin, R. T., Sewell, K. W., Goldstein, A., & Leonard, K. (1998). A comparison of forensic and nonforensic malingerers: A prototypical analysis of explanatory models. Law and Human Behavior, 22, 353–367. Rogers, R., Sewell, K., & Goldstein, A. (1994). Explanatory models of malingering: A prototypical analysis. Law and Human Behavior, 18, 543–552. Rogers, R., Sewell, K. W., Martin, M. A., & Vitacco, M. J. (2003). Detection of feigned mental disorders: A meta-analysis of the MMPI-2 and malingering. Assessment, 10, 160–177. Rogers, R., Sewell, K. W., Morey, L. C., & Ustad, K. L. (1996). Detection of feigned mental disorders on the Personality Assessment Inventory: A discriminant analysis. Journal of Personality Assessment, 67, 629–640. Rohling, M. L., Binder, L. M., Demakis, G. J., Larrabee, G. J., Ploetz, D. M., & Langhinrichsen-Rohling, J. (2011). A meta-analysis of neuropsychological outcome after mild traumatic brain injury; Re-analyses and reconsiderations of Binder et al. (1997), Frencham et al. (2005), and Pertab et al. (2009). The Clinical Neuropsychologist, 25(4), 608–623. Rohling, M. L., Meyers, J. E., & Millis, S. R. (2003). Neuropsychological impairment following traumatic brain injury: A dose–response analysis. The Clinical Neuropsychologist, 17, 289–302. Root, J. C., Robbins, R. N., Chang, L., & van Gorp, W. G. (2006). Detection of inadequate effort on the California Verbal Learning Test–Second Edition: Forced choice recognition and critical item analysis. Journal of the International Neuropsychological Society, 12, 688–696. Rosen, G. M., & Taylor, S. (2007). Pseudo-PTSD. Journal of Anxiety Disorders, 21(2), 201–210. Ross, S. R., Millis, S. R., Krukowski, R. A., Putnam, S. H., & Adams, K. M. (2004). Detecting incomplete effort on the MMPI-2: An examinations of the FBS in mild head injury. Journal of Clinical and Experimental Neuropsychology, 26, 115–124. Ross, S. R., Putnam, S. H., & Adams, K. M. (2006). Psychological disturbance, incomplete effort, and compensation-seeking status as predictors of neuropsychological performance in head injury. Journal of Clinical and Experimental Neuropsychology, 28, 111–125. Ruff, R. M., Klopfer, J., & Blank, J. (2016). hom*ogenous base rates for malingering in neuropsychological examination of litigants. Psychological Injury and Law, 9(2), 143–153. Ruff, R. M., & Richardson, A. M. (1999). Mild traumat-

ic brain injury. In J. J. Sweet (Ed.), Forensic neuropsychology: Fundamentals and practice (pp.313–338). Lisse, The Netherlands: Swets & Zeitlinger. Ryan, J. J., Turpin, D. M., & Kreiner, D. S. (2012). Specificity of the 21-Item Test in two elderly samples. Aging, Neuropsychology, and Cognition, 19(6), 723– 740. Sanfey, A. G., Rilling, J. K., Aronson, J. A., Nystrom, L. E., & Cohen, J. D. (2003). The neural basis of economic decision-making in the Ultimatum Game. Science, 300(5626), 1755–1758. Schneiderman, A. I., Braver, E. R., & Kang, H. K. (2008). Understanding sequelae of injury mechanisms and mild traumatic brain injury incurred during the conflicts in Iraq and Afghanistan: Persistent postconcussive symptoms and posttraumatic stress disorder. American Journal of Epidemiology, 167(12), 1446–1452. Schretlen, D. J., Brandt, J., Krafft, L., & van Gorp, W. (1991). Some caveats in using the 15-Item Memory Test to detect malingered amnesia. Psychological Assessment, 3, 667–672. Schretlen, D. J., Munro, C. A., Anthony, J. C., & Pearlson, G. D. (2003). Examining the range of normal intra-individual variability in neuropsychological test performance. Journal of the International Neuropsychological Society, 9, 864–870. Schretlen, D. J., & Shapiro, A. M. (2003). A quantitative review of the effects of traumatic brain injury on cognitive functioning. International Review of Psychiatry, 15(4), 341–349. Schroeder, R. W., Baade, L. E., Peck, C. P., VonDran, E. J., Brockman, C. J., Webster, B. K., et al. (2012). Validation of MMPI-2-RF validity scales in criterion group neuropsychological samples. The Clinical Neuropsychologist, 26(1), 129–146. Scott, J. C., Matt, G. E., Wrocklage, K. M., Crnich, C., Jordan, J., Southwick, S. M., et al. (2015). A quantitative meta-analysis of neurocognitive functioning in posttraumatic stress disorder. Psychological Bulletin, 141(1), 105–140. Shively, S., Scher, A. I., Perl, D. P., & Diaz-Arrastia, R. (2012). Dementia resulting from traumatic brain injury: What is the pathology? Archives of Neurology, 69(10), 1245–1251. Silk-Eglit, G. M., Lynch, J. K., & McCaffrey, R. J. (2016). Validation of Victoria Symptom Validity Test cutoff scores among mild traumatic brain injury litigants using a known-groups design. Archives of Clinical Neuropsychology, 31, 231–245. Silver, J. M. (2012). Effort, exaggeration, and malingering after concussion. Journal of Neurology, Neurosurgery, and Psychiatry, 83(8), 836–841. Silver, J. M., McAllister, T. W., & Arciniegas, T. B. (2009). Depression and cognitive complaints following mild traumatic brain injury. American Journal of Psychiatry, 166(6), 153–161. Silverberg, N. D., & Iverson, G. L. (2011). Etiology of

7. Malingered Traumatic Brain Injury 149

the postconcussion syndrome: Physiogenesis and psychogenesis revisited. NeuroRehabilitation, 29(4), 317–329. Silverberg, N. D., Iverson, G. L., Brubacher, J. R., Holland, E., Hoshino, L. C., Aquino, A., et al. (2016). The nature and clinical significance of preinjury recall bias following mild traumatic brain injury. Journal of Head Trauma Rehabilitation, 31(6), 388–396. Sleep, C. E., Petty, J. A., & Wygant, D. B. (2015). Framing the results: Assessment of response bias through select self-report measures in psychological injury evaluations. Psychological Injury and Law, 8, 27–39. Slick, D. J., Hopp, G., Strauss, E., & Spellacy, F. J. (1996). Victoria Symptom Validity Test: Efficiency for detecting feigned memory impairment and relationship to neuropsychological tests and the MMPI-2 validity scales. Journal of Clinical and Experimental Neuropsychology, 18, 911–922. Slick, D., Hopp, G., Strauss, E., & Thompson, G. B. (1997). Victoria Symptom Validity Test. Odessa, FL: Psychological Assessment Resources. Slick, D. J., Iverson, G. L., & Green, P. (2000). California Verbal Learning Test indicators of suboptimal performance in a sample of head-injury litigants. Journal of Clinical and Experimental Neuropsychology, 22, 569–579. Slick, D. J., & Sherman, E. M. S. (2013). Differential diagnosis of malingering. In D. A. Carone & S. S. Bush (Eds.), Mild traumatic brain injury: Symptom validity assessment and malingering (pp.57–72). New York: Springer. Slick, D. J., Sherman, E. M. S., & Iverson, G. L. (1999). Diagnostic criteria for malingered neurocognitive dysfunction: Proposed standards for clinical practice and research, The Clinical Neuropsychologist, 13, 545–561. Snell, D. L., Siegert, R. J., Hay-Smith, E. J. C., & Surgenor, L. J. (2011). Associations between illness perceptions, coping styles and outcome after mild traumatic brain injury: Preliminary results from a cohort study. Brain Injury, 25(11), 1126–1138. Soble, J. R., Spanierman, L. B., & Fitzgerald-Smith, J. (2013). Neuropsychological functioning of combat veterans with posttraumatic stress disorder and mild traumatic brain injury. Journal Clinical and Experimental Neuropsychology, 35(5), 555–561. Sollman, M. J., & Berry, D. T. (2011). Detection of inadequate effort on neuropsychological testing: A metaanalytic update and extension. Archives of Clinical Neuropsychology, 26(8), 774–789. Spencer, R. J., Axelrod, B. N., Drag, L. L., WaldronPerrine, B., Pangilinan, P. H., & Bieliauskas, L. A. (2013). WAIS-IV reliable digit span is no more accurate than age corrected scaled score as an indicator of invalid performance in a veteran sample undergoing evaluation for mTBI. The Clinical Neuropsychologist, 27(8), 1362–1372. Suhr, J. A., & Gunstad, J. (2005). Further exploration of

the effect of “diagnosis threat” on cognitive performance in individuals with mild head injury. Journal of the International Neuropsychological Society, 11(1), 23–29. Suhr, J. A., & Wei, C. (2013). Response expectancies and their potential influence in neuropsychological evaluation. In P. A. Arnett (Ed.), Secondary influences on neuropsychological test performance (pp.182– 200). New York: Oxford University Press. Sullivan, K. A., & Edmed, S. L. (2012). The good-olddays bias and post-concussion syndrome symptom reporting in a non-clinical sample. Brain Injury, 26(9), 1098–1104. Sumanti, M., Boone, K. B., Savodnik, I., & Gorsuch, R. (2006). Noncredible psychiatric and cognitive symptoms in a workers’ compensation “stress” claim sample. The Clinical Neuropsychologist, 20, 754–765. Sweet, J. J., Goldman, D. J., & Guidotti-Breting, L. M. (2013). Traumatic brain injury: Guidance in a forensic context from outcome, dose–response, and response bias research. Behavioral Science and the Law, 31(6), 756–778. Sweet, J. J., King, J. H., Malina, A. C., Bergman, M. A., & Simmons, A. (2002). Documenting the prominence of forensic neuropsychology at national meetings and in relevant professional journals from 1990 to 2000. The Clinical Neuropsychologist, 16, 481–494. Tait, R. C., & Chibnall, J. T. (2016). Workers’ compensation claimants with low back pain: The role of dissatisfaction in the transition to disability. Psychological Injury and Law, 9(1), 16–22. Tenhula, W. N., & Sweet, J. J. (1996). Double crossvalidation of the Booklet Category Test in detecting malingered traumatic brain injury. The Clinical Neuropsychologist, 10, 104–116. Tombaugh, T. N. (1997). The Test of Memory Malingering (TOMM): Normative data from cognitively intact and cognitively impaired individuals. Psychological Assessment, 9(3), 260–268. Trafimow, D. (2014). More reasons why diffuse tensor imaging is problematic in medicolegal settings. AJOB Neuroscience, 5(2), 39–41. Trost, Z., Mondon, K. R., Buelow, M., Boals, A., & Scott, W. (2016). Perceived injustice predicts intention to litigate: Findings from a spinal cord injury sample. Psychological Injury and Law, 9(1), 31–40. Uomoto, J. M., & Esselman, P. C. (1993). Traumatic brain injury and chronic pain: Differential types and rates by head injury severity. Archives of Physical Medicine and Rehabilitation, 74, 61–64. Van Gorp, W. G., Humphrey, L. A., Kalechstein, A., Brumm, V. L., McMullen, W. J., & Stoddard, N. A. (1999). How well do standard clinical neuropsychological tests identify malingering?: A preliminary analysis. Journal of Clinical and Experimental Neuropsychology, 21, 245–250. Vanderploeg, R. D., Belanger, H. G., & Kaufmann, P. M. (2014). Nocebo effects and mild traumatic brain

150

II .   D i a g n o s t ic I ss u e s

injury: Legal implications. Psychological Injury and Law, 7(3), 245–254. Vickery, C. D., Berry, D. T. R., Inman, T. H., Harris, M. J., & Orey, S. A. (2001). Detection of inadequate effort on neuropsychological testing: A meta-analytic review of selected procedures. Archives of Clinical Neuropsychology, 16, 45–73. Victor, T. L., Boone, K. B., Serpa, J. G., Buehler, J., & Ziegler, E. A. (2009). Interpreting the meaning of multiple symptom validity test failure. The Clinical Neuropsychologist, 23(2), 297–313. Volbrecht, M. E., Meyers, J. E., & Kaster-Bundgaard, J. (2000). Neuropsychological outcome of head injury using a short battery. Archives of Clinical Neuropsychology, 15, 251–265. Walsh, K. W. (1985). Understanding brain damage: A primer of neuropsychological evaluation. Edinburgh, UK: Churchill Livingstone. Warriner, E. M., Rourke, B. P., Velikonja, D., & Metham, L. (2003). Subtypes of behavioral and emotional sequelae in patients with traumatic brain injury. Journal of Clinical and Experimental Neuropsychology, 25, 904–917. Weinborn, M., Orr, T., Woods, S. P., Conover, E., & Feix, J. (2003). A validation of the Test of Memory Malingering in a forensic psychiatric setting. Journal of Clinical and Experimental Neuropsychology, 25, 979–990. Weissman, H. B. (1990). Distortions and deceptions in self-presentation: Effects of protracted litigation on personal injury cases. Behavioral Sciences and the Law, 8, 67–74. Whiteside, D. M., Galbreath, J., Brown, M., & Turnbull, J. (2012). Differential response patterns on the PAI in compensation-seeking and non-compensationseeking mild TBI patients. Journal of Clinical and Experimental Neuropsychology, 43(2), 172–182. Whittaker, R., Kemp, S., & House, A. (2007). Illness perceptions and outcome in mild head injury: A lon-

gitudinal study. Journal of Neurology, Neurosurgery, and Psychiatry, 78(6), 644–646. Wolfe, P. L., Millis, S. R., Hanks, R., Fichtenberg, N., Larrabee, G. J., & Sweet, J. J. (2010) Effort indicators within the California Verbal Learning Test–II (CVLT-II). The Clinical Neuropsychologist, 24(1), 153–168. Wood, R. (2004). Understanding the “miserable minority”: A diathesis–stress paradigm for postconcussional syndrome. Brain Injury, 18(11), 1135–1153. World Health Organization. (1992). International statistical classification of diseases and related health problems (10th ed.). Geneva: Author. Wortzel, H. S., Tsiouris, A. J., & Filippi, C. G. (2014). The potential for medicolegal abuse: Diffusion tensor imaging in traumatic brain injury. AJOB Neuroscience, 5(2), 9–15. Yakobov, E., Scott, W., Tibault, P., & Sullivan, M. J. L. (2016). Treatment-related reductions in disability are associated with reductions in perceived injustice following treatment of whiplash injury. Psychological Injury and Law, 9(1), 41–47. Young, G. (2015). Malingering in forensic disability-related assessments: Prevalence 15 ± 15%. Psychological Injury and Law, 8(3), 188–199. Young, J. C., Sawyer, R. J., Roper, B. L., & Baughman, B. C. (2012). Expansion and re-examination of Digit Span effort indices on the WAIS-IV. The Clinical Neuropsychologist, 26(1), 147–159. Youngjohn, J. R., Davis, D., & Wolf, I. (1997). Head injury and the MMPI-2: Paradoxical severity effects and the influence of litigation. Psychological Assessment, 9, 177–184. Zakzanis, K. K., Leach, L., & Kaplan, E. (1998). Neuropsychological differential diagnosis. Visse, The Netherlands: Swets & Zeitlinger. Zakzanis, K. K., & Yeung, E. (2011). Base rates of postconcussive symptoms in a nonconcussed multicultural sample. Archives of Clinical Neuropsychology, 26(5), 461– 465.

CHAPTER 8

Denial andMisreporting ofSubstance Abuse LyndaA.R.Stein,PhD RichardRogers,PhD SarahHenry,PhD

Substance abuse is often associated with penalties including social sanctions and stigmatization. This association is illustrated by the adoption of penalties for driving under the influence (see Ying, Wu, & Chang, 2013), and by the widespread use of drug-testing policies in work settings (DiThomas, 2012). These policies resulted from substanceabusing employees having higher rates of absenteeism and poor performance (Stein, Smith, Guy, & Bentler, 1993). Research suggests dissimulation on substance use commonly occurs in a variety of settings even when guaranteed confidentiality. As an example, families in urban, low socioeconomic status areas significantly underreport cocaine use, with youth underreporting even more than their parents, presumably due to fear of repercussions (Delaney-Black et al., 2010). In contrast, elderly persons may underreport drug use unintentionally due to memory difficulties (Rockett, Putnam, Jia, & Smith, 2006). As a comparison, treatmentseeking adolescents are more likely to overreport marijuana use to enhance social status among their peers (Williams & Nowatzki, 2005). Moreover, national surveys find greater underreporting of more stigmatizing drugs, such as cocaine or heroin (Magura, 2010). Despite social sanctions and stigma, use of illicit substances is widespread (Substance Abuse and Mental Health Services Administration [SAM-

HSA], 2014). In 2013 alone, the prevalence rate of substance use disorder was 12.2% for adults and 5.2% for persons ages 12–17 years (SAMHSA, 2014). Individuals may withhold reports of their substance use in order to avoid legal consequences (Tourangeau & Yan, 2007) or to enhance social desirability (see Buchan, Dennis, Tims, & Diamond, 2002). Persons are unlikely to be forthcoming when faced with criminal or civil sanctions. Examples of the latter include parenting (e.g., fitness to parent) and employment (e.g., torts arising from drugrelated work injuries). In legal settings (e.g., detention centers), self-reports often underestimate substance use when compared to biological testing results (Mieczkowski, 2002; Knight, Hiller, Simpson, & Broome, 1998). Conversely, although largely underinvestigated, offenders arrested for serious crimes may be motivated to overreport or fabricate drug use to potentially serve as a mitigating factor in sentencing (Rogers & Mitchell, 1991), or to appear more sympathetic in some cases (Ortiz & Spohn, 2014).

OVERVIEW SpecificTerminology Specific terms are often used to describe response styles related to substance abuse:

151

152

I I .   D i a g n o s t ic I ss u e s

1. Disacknowledgment: Claiming a lack of knowledge (“I don’t remember”) of either drug usage or behavioral consequences. Disacknowledgment may or may not be sincere. 2. Misappraisal: These distortions are unintentional rather than deliberate attempts to provide false information regarding substance use. 3. Denial: Substance use or related consequences are purposely minimized for specific reasons. Examples of reasons include social desirability, desire to avoid consequences (e.g., unwanted treatment), or unwillingness to accept responsibility for behavior. 4. Exaggeration: Individuals intentionally magnify substance use and its consequences. For example, adolescents may overreport substance use to increase social desirability (Palmer, Dwyer, & Semmer, 1994).

Prevalence ofDissimulation amongSubstanceAbusers

of offenders referred for impaired driving, Lapham, C’de Baca, Chang, Hunt, and Berger (2001) found that nearly 30% of those denying substance abuse during initial screening actually met criteria for a substance use diagnosis when interviewed for a second time. However, the opposite pattern has also been seen in adult offenders. Some detainees are motivated to inflate their use if they believe it could result in mandated substance abuse treatment rather than serving a jail sentence (Hser, Maglione, & Boyle, 1999). ClinicalSettings

Among clinical populations, Winters, Stinchfield, Henly, and Schwartz (1991) reported low percentages of adolescents in treatment (5.1–6.3%) who exaggerated drug use. While low, these percentage are nearly double those of nonclinical adolescent populations (i.e., 2.8–3.8% (Meldrum & Piquero, 2015; Petzel, Johnson, & McKillip.1973).

ForensicSettings

Types ofDistortion

Youth involved with the juvenile justice system have higher rates of substance use than their counterparts not involved with the juvenile justice system. As an example, McClelland, Elkington, Teplin, and Abram (2004) found that nearly half of detainees from a juvenile justice sample met criteria for one or more substance use disorders— specifically, alcohol and marijuana use disorders, which are the most commonly seen disorders in juvenile offenders. However, survey data suggest that respondents are likely to minimize their substance abuse even with an assurance of anonymity (Colon, Robles, & Sahai, 2002). On this point, Williams and Nowatzki (2005) studied 367 adolescents referred for substance use assessment; 26% incorrectly denied substance use, as determined by biochemical testing. Additionally, Stein (2016) asked incarcerated adolescents (N = 164), about whether they had under- or overreported amount, frequency, or problems associated with alcohol and marijuana. Inaccurate reports ranged from 13.4% for alcohol problems to 22.6% for marijuana. Despite assurances of confidentiality, 45.1% indicated some concern that the information might not be private. Denial of substance use in adult offender samples is also common, particularly for “harder” drugs such as powder and crack cocaine (Bureau of Justice Statistics, 2010), especially if these drugs could result in additional legal sanctions. In a study

Distortions regarding substance abuse may occur on three dimensions: (1) amount and type of substance abuse, (2) immediate behavioral and psychological effects of substance abuse, and (3) consequent impairment and psychological sequelae from cumulative substance abuse. We focus on the latter two dimensions. Distortions about the immediate behavioral and psychological effects of substance abuse have not been systematically investigated. No standardized measures are available to assess the immediate behavioral consequences of substance use. As a proxy, general guidelines for some substances have been promulgated, such as the relationship of estimated blood alcohol content (BAC) to general level of impairment (National Institute on Alcohol Abuse and Alcoholism [NIAAA], 2015). A respondent may be forthcoming about the amount of drug abuse but inaccurate about impaired judgment and erratic behavior, which often are more relevant to a specific incident than general usage. The Biphasic Alcohol Effects Scale (Martin, Earleywine, Musty, Perrine, & Swift, 1993) was developed to measure intoxicating effects of alcohol according to the client’s report. However, the relationship of BAC to intoxicated behavior continues to be highly variable. The same individual may respond differently with identical BACs depending on mood state and other situational factors (de Wit, Uhlenhuth, Pierri, & Johanson, 1987). Like-

8.  Denial and Misreporting of Substance Abuse 153

wise, the relationship between drug consumption and consequences (i.e., intoxicated behavior and psychological effects) is also variable, and may depend, for example, on tolerance and environmental cues (Siegel, 2005). In contrast, psychometric methods tend to focus on long-term effects and sequelae of substance abuse, including disorders as a result of chronic use (e.g., dependence), cognitive deficits (e.g., memory impairment), and clinical correlates (e.g., personality characteristics). For instance, versions of the Substance Abuse Subtle Screening Inventory (see below) have examined the effects of both reported and unacknowledged substance abuse.

CLINICALMETHODS In this section, brief screens offer rapid and relatively effective means of evaluating self-acknowledged substance use. Given the simplicity of these screens and typical face-valid nature, more elaborate procedures with greater discriminant validity may be particularly useful when misreporting is suspected. Examples of these include structured interviews and multiscale inventories that employ a broad-spectrum approach toward psychological impairment and incorporate syndromes and scales for substance abuse. Finally, specialized methods have been developed specifically for assessment of substance abuse. The validation of substance abuse measures is predicated on the accurate measurement of the external criterion, namely the use/abuse of alcohol and drugs (see Table 8.1). Unfortunately, many studies attempt to satisfy this criterion by simply using uncorroborated self-reporting. Practitioners can quickly recognize the non-independence and paradoxical nature of this simplistic approach. Other research has relied on informants and convergent indicators from other measures to address this concern. While a modest improvement is achieved over “self-report by self-report” criterion, such studies are still largely dependent on secondary sources of self-reporting. In contrast to these approaches, two types of external validity are generally effective: (1) treatment history for establishing drug-related disorders and (2) biochemical methods for confirming the use or nonuse of alcohol and drugs. However, even these methods are limited, because they do not provide a good indication of amounts consumed over a specific time period, or behavioral effects, if any. Research can easily demonstrate the vulnerability of substance

TABLE 8.1.  Types of Test Validation forSubstance Abuse Measures and Their Relevance to Honesty and Dissimulation

1. Self-report by self-report. The criterion is acknowledged use/nonuse. Paradoxically, honesty is assumed for the self-described use/nonuse to assess forthrightness on the substance measure. 2. Self-report by informant report. The approach assumes a collateral source, typically a family member, has firsthand knowledge of the respondent’s substancerelated behavior. It is recommended that informants provide (a) a rating of how confident they are in their reports, and (b) a specific description of what they observed. 3. Self-report by treatment history. Treatment history provides an excellent criterion for a longitudinal perspective of substance abuse. Substance abusers are likely to know denial will be detected; therefore, they are likely to be more forthcoming than the target group (i.e., individuals for whom substance abuse has not been established). 4. Self-report by biochemical methods. Biochemical methods can accurately establish current (urine analysis) and long-term (hair analysis) drug use. This is a very effective approach to verify use, although not behavioral effects, of drugs. 5. Self-report by convergent indicators. Degree of association between two or more substance abuse measures is evaluated. To the extent that external validity of measures is based on either item 1 or 2 (above), the convergence of measures also remains vulnerable to response styles. 6. Self-report by simulation design. Users and nonusers are asked simulate on substance abuse measures. Although rarely used, the simulation design provides direct information on the vulnerability of substance abuse measures to distortion. Consideration must be given to type of dissimulation and whether it is general (e.g., social desirability) or substance-specific (e.g., denied drug use).

abuse measures to denial and exaggeration. Therefore, studies should ideally combine treatment history, laboratory methods, and corroborated self-reports (e.g., via significant other) as external criteria with simulation studies of denial and exaggeration. The measures listed below are not intended to be exhaustive. Some measures were included because they contain indicators of response bias, despite relatively little use in the literature. Cutoff scores and classification rates vary depending on factors such as setting (e.g., legal vs. primary care),

154

II .   D i a g n o s t i c I s s u e s TABLE 8.2.  Brief Summary of Substance Use Screens

Screen

Items

Coverage

Populations

Face valid

Validity scales

AUDIT

10

Alcohol

Adults/adolescents

Yes

No

MAST

25

Alcohol

Adults

Yes

No

RAPS4

4

Alcohol

Adults

Yes

No

RAPI

23

Alcohol

Adolescents

Yes

No

ASSIST

7–8

Alcohol/drugs

Adults/adolescents

Yes

No

CRAFFT

6

Alcohol/drugs

Adolescents

Yes

No

DUSI

152

Alcohol/drugs

Adults/adolescents

Yes

No

GAIN-SS

20–23

Alcohol/drugs

Adults/adolescents

Yes

No

MAYSI-2

52

Alcohol/drugs

Adolescents

Yes

No

PESQ

40

Alcohol/drugs

Adolescents

Yes

Yes

SIP

10–17

Alcohol/drugs

Adults

Yes

No

TCUDS

15–17

Alcohol/drugs

Adolescents

Yes

No

DAST

10–28

Drugs

Adults/adolescents

Yes

No

Note. See text for full names. Item length varies by version.

variant of the instrument (e.g., Alcohol Use Disorders Identification Test vs. Alcohol Use Disorders Identification Test–Consumption, see below), and criterion (e.g., use, extensive use, or substance diagnosis). Given this variability, an extensive review of cutoff scores and classification rates is not provided (see Table 8.2).

Screens forSubstanceAbuse Adult AlcoholScreeners MICHIGAN ALCOHOLISM SCREENINGTEST

The Michigan Alcoholism Screening Test (MAST; Selzer, 1971) consists of 25 items to determine alcohol abuse via interview or self-administration. Short-MAST (SMAST, 13 items; Selzer, Vinokur, & van Rooijen, 1975) and brief MAST (bMAST, 10 items; Pokorny, Miller, & Kaplan, 1972) are also available. The bMAST has been shown to be effective in assessing self-disclosed problems with drinking severity in some samples (see Connor, Grier, Feeney, & Young, 2007). The MAST is effective in classifying alcohol use disorders among inpatients seeking treatment for self-acknowledged problems (Moore, 1972; Ross, Gavin, & Skinner, 1990). However, because of its high face validity, respondents can easily fake on the MAST by denying alcohol abuse (Otto & Hall, 1988; Nochajski & Wieczorek, 1998).

FOUR-ITEM RAPID ALCOHOL PROBLEMSSCREEN

Items from several screens including AUDIT (see below) and bMAST were evaluated to develop the Four-Item Rapid Alcohol Problems Screen (RAPS4; Cherpitel, 2000). Two quantity–frequency items were added to increase sensitivity for alcohol abuse and harmful drinking (RAPS4QF; Cherpitel, 2002). These screens have been used internationally in emergency departments and may be administered orally or in writing, although some decrement was found for their internal consistency as compared to the use in the United States (Cherpitel, Ye, Moskalewicz, & Swiatkiewicz, 2005; Cremonte, Ledesma, Cherpitel, & Borges, 2010; Geneste et al., 2012). Its use of face-valid items leaves this measure vulnerable to misreporting. Adolescent AlcoholScreeners RUTGERS ALCOHOL PROBLEMINDEX

White and Labouvie (1989) developed the Rutgers Alcohol Problem Index (RAPI) questionnaire for quick screening of alcoholism among adolescents. The RAPI’s 23 face-valid items address potential negative effects of problematic alcohol use, as well as consequences unique to adolescents. The authors have experimented with different time frames (e.g., the last 3 years or “ever”) and different

8.  Denial and Misreporting of Substance Abuse 155

response categories. Frequently applied to college students, the RAPI data indicate three factors: Abuse/Dependence Symptoms, Personal Consequences, and Social Consequences, each with adequate internal consistency and convergent validity (Martens, Neighbors, Dams-O’Connor, Lee, & Larimer, 2007). With face-valid measures, White and Labouvie (1989) have acknowledged the potential for denial or exaggeration on the RAPI, neither of which has been researched. Adult Alcohol andDrugScreeners TEXAS CHRISTIAN UNIVERSITY DRUGSCREEN

The 15-item Texas Christian University Drug Screen–II (TCUDS; Institute of Behavioral Research [IBR], 2007) provides an interview-based, self-report measure of frequency of substance use, treatment history, substance-use diagnoses, and treatment motivation. Its 17-item revision (TCUDS-II) has recently been updated to reflect DSM-5 diagnostic criteria (IBR, 2014). In a study detecting substance use disorders among male inmates (N = 400), sensitivity (.85) and specificity (.78) were good, but the “external” criterion simply involved an earlier version of the instrument (Peters et al., 2000), which may be viewed as criterion contamination. Frequently used in correctional systems (Peters, LeVasseur, & Chandler, 2004), it appears to be susceptible to dissimulation (Richards & Pai, 2003). SHORT INDEX OFPROBLEMS

The Short Index of Problems (SIP; Miller, Tonigan, & Longabaugh, 1995) is a 15-item questionnaire for assessing lifetime and past 3-month alcohol consequences. As a strength, the SIP also has a collateral version for evaluating report consistency. In addition to a total score, five subscales assess alcohol problems (Physical, Social Responsibility, Intrapersonal, Impulse Control, and Interpersonal). Its 2-day test–retest reliability ranges from good to excellent (rs from 71 to .95; Miller et al., 1995). However, the SIP total score only modestly correlated with alcohol dependence criteria (r = .36; Feinn, Tennen, & Kranzler, 2003). Many versions of the SIP share the majority of items, such as SIP modified for Drug Use (SIP-DU; Allensworth-Davles, Cheng, Smith, Samet, & Saitz, 2012) and SIP–Drugs (SIP-D; Alterman, Cacciola, Ivey, Habing, & Lynch, 2009). Given their face validity, all SIP measures appear to be susceptible to

dissimulation, so the collateral version should also be routinely administered. Adolescent Alcohol andDrugScreeners PERSONAL EXPERIENCE SCREENINGQUESTIONNAIRE

The 40-item Personal Experience Screening Questionnaire (PESQ; Winters, 1992) has a Likert-type response format addressing drug problem severity, problems often associated with drug use (e.g., physical abuse), and drug use history. Of particular importance, it contains two scales for assessing Defensiveness and Infrequent Responding. The data suggest that only about 15% of protocols were invalid due to compromised self-reporting. However, no formal studies were conducted to determine efficacy of scales to detect response distortion. The PESQ appears to be effective at identifying those with clinical needs for further drug-abuse evaluation (Winters & Kaminer, 2008). CRAFFT

Via interview or written format, yes–no inquiries cover six content items: riding in a Car driven by someone using alcohol/drugs (AD), using substances to Relax or when Alone, dependence problems identified by Friends/family, Forgetting/ regretting actions when using, and getting into Trouble while using. Excellent concurrent validity has been demonstrated in psychiatric (Oesterle, Hitschfeld, Lineberry, & Schneekloth, 2015) and primary care settings (Knight, Sherritt, Shrier, Harris, & Chang, 2002). Obviously, the CRAFFT is vulnerable to drug denial. MASSACHUSETTS YOUTH SCREENINGINVENTORY–2

The 52-item Massachusetts Youth Screening Inventory–2 (MAYSI-2; Grisso & Barnum, 2006) has a yes–no response format with seven scales, including Alcohol/Drug Use. As evidence of reliability, adjudicated youth (N = 248) were administered the MAYSI-2 prior to placement decisions, producing a very high correlation (r = .92) between paper and interview-based formats on the Alcohol/Drug Use scale (Hayes, McReynolds, & Wasserman, 2005). When using substance diagnoses, a British study found poor agreement with the Alcohol/Drug Use scale (Lennox, O’Malley, Bell, Shaw, & Dolan, 2015). Important questions have

156

I I .   D i a g n o s t ic I ss u e s

been raised about the MAYSI-2 regarding external validity, response styles, and possible cultural influences. Adult andAdolescent Substance AbuseScreeners ALCOHOL USE DISORDERS IDENTIFICATIONTEST

The Alcohol Use Disorders Identification Test (AUDIT; Saunders, Aasland, Babor, De La Fuente, & Grant, 1993) consists of 10 questions about alcohol-related symptoms and is administered via interview or written format. Since its inception, various studies (Shields & Caruso, 2003; Reinert & Allen, 2002; Donovan, Kivlahan, Doyle, Longabaugh, & Greenfield, 2006) in hospitals and other settings have demonstrated strong psychometric properties. It has also been well validated for use with youth (Knight, Sherritt, Kim, Gates, & Chang, 2003; Cook, Chung, Kelly, & Clark, 2005). Promoted by the World Health Organization (WHO; Babor, Biddle, Saunders, & Monteiro, 2001), the AUDIT has been studied internationally in nine countries and has demonstrated fairly good results with clinical samples (see Cassidy, Schmitz, & Malla, 2007; Lundin, Hallgren, Balliu, & Forsell, 2015; Pradhan et al., 2012). Higher scores on Impression Management are associated with low AUDIT score, indicating its vulnerability to underreporting (Zaldivar, Molina, Rios, & Montes, 2009). Face-valid items leave the AUDIT vulnerable to dissimulation, which has not been well-studied.

ever, among youth admitted for inpatient evaluations, DAST-A was unrelated to social desirability (Martino et al., 2000). However, it should be noted these correlations indicate “faking good” and not specifically the denial of substance use. Finally, a simulation study found that inpatients could significantly suppress their DAST scores (Wooley, Rogers, Fiduccia, & Kelsey, 2012). Adult andAdolescent Alcohol andDrugScreeners DRUG USE SCREENINGINVENTORY

The Drug Use Screening Inventory (DUSI) was developed in the United States for use with both adolescent and adult substance abusers (Tarter & Hegedus, 1991), although most studies focus on adolescents. As the longest screen (152 items covering 10 domains), DUSI items extend beyond substance abuse to address general functioning (school/work, social, mental/physical functioning). International studies have also found that the DUSI can distinguish youth with and without substance problems (e.g., De Micheli & Formigoni, 2002). The revised version (DUSI-R; Kirisci, Hsu, & Tarter, 1994) includes a general Lie scale. Among Brazilian youth from a school setting, the Lie scale classified over half as underreporting, which raises questions about its clinical usefulness (Dalla-Déa, De Micheli, & Formigoni, 2003). A recent simulation study also found that the DUSIR is highly susceptible to complete and partial denial of substance use or related problems in adults (Wooley et al., 2012).

DRUG ABUSE SCREENINGTEST

The Drug Abuse Screening Test (DAST; Skinner, 1982) comprises 28 items (yes–no responses) and was developed from the MAST (discussed earlier). The DAST’s content reflects frequency of drug use, and interpersonal, legal, and medical problems associated with use. It has been extensively tested in psychiatric, work, substance treatment, and justice settings (Yudko, Lozhkina, & Fouts, 2007). Shorter versions, as well as a 27-item youth version (DAST-A; Martino, Grilo, & Fehon, 2000) are available. A review of these instruments indicates good reliability (Yudko et al., 2007). Importantly, DAST studies have evaluated the effects of underreporting. Skinner (1982) found DAST with modest negative correlations to denial (r = –.28) and social desirability (r = –.31 to –.38) among treatment-seeking substance users. How-

ALCOHOL, SMOKING ANDSUBSTANCE INVOLVEMENT SCREENINGTEST

The Alcohol, Smoking and Substance Involvement Screening Test (ASSIST) classifies low, moderate, or high risk for a range of drugs (Humeniuk, Henry-Edwards, Ali, Poznyak, & Monteiro, 2010). The ASSIST was developed by recruiting subjects from medical, addiction treatment, and psychiatric settings across several countries. Although initially validated on adults, ASSIST scores for tobacco, alcohol, and marijuana in an adolescent sample have demonstrated good internal consistency and concurrent validity (Gryczynski et al., 2014). Further work produced a second edition (ASSIST-2), establishing concurrent and construct validity. Its other strengths include discriminant validity (e.g., distinguishing between use, abuse, and de-

8.  Denial and Misreporting of Substance Abuse 157

pendence), and 3-month stability follow-up (Newcombe, Humeniuk, & Ali, 2005). The ASSIST-3 added weighted scoring procedures with validation on samples from primary care and substance abuse treatment settings (Humeniuk et al., 2008). It comprises eight questions. As with all face-valid measures, it is susceptible to misreporting. To partially address this issue, fictitious drug names were included to identify overreporting; no data are available for underreporting or drug denial. THE GLOBAL APPRAISAL OFINDIVIDUAL NEEDS—SHORTSCREENER

The Global Appraisal of Individual Needs—Short Screener (GAIN-SS; Dennis, Chan, & Funk, 2006) originally comprised 20 items addressing lifetime and past-year substance use, crime involvement, and internalizing–externalizing problems for adults and youth. More recent versions include the 23-item GAIN-SS-3 (Dennis, Feeney, & Titus, 2013), which is available in several languages. However, most research pertains to the GAIN-SS version; Dennis et al. (2006) studied almost 8,000 youth and adults from a wide array of treatment settings in the United States and found good to very good alphas and agreement with the full GAIN General Individual Severity scale. As an important note, an independent study of youth recruited from clinical settings found the GAINSS to be highly effective in detecting substance abuse (McDonell, Comtois, Voss, Morgan, & Reiss, 2009). However, another large study of adolescents from outpatient settings raised concern that GAIN-SS may misclassify persons with substance or behavioral disorders (Stucky, Orlando, & Ramchad, 2014). StructuredInterviews

An advantage of structured interviews is the standardization of clinical inquiries and responses so that direct comparisons can be made between the respondent and significant others (Rogers, 2001). However, respondent–informant agreement tends to be relatively modest, even when no evidence of dissimulation is found. Therefore, a lack of agreement does not signify deception. In this section, we selectively review (1) diagnostic interviews that contain substance abuse components and (2) targeted substance abuse interviews. Diagnostic interviews for adults and youth include sections on substance disorders covering life-

time or past-year diagnoses. Selected examples for interviews include the following: 1. The Structured Clinical Interview for DSM-5 Disorders (SCID-5; First, Williams, Karg, & Spitzer, 2016) 2. The Diagnostic Interview Schedule for Children–IV (DISC-IV; Shaffer, Fisher, Lucas, Dulcan, & Schwab-Stone, 2000) 3. The Kiddie Schedule for Affective Disorders and Schizophrenia—Present/Lifetime versions (K-SADS-PL; Axelson, Birmaher, Zelazny, Kaufman, & Gill, 2009) 4. The Composite International Diagnostic Interview–3.0 (CIDI-3), validated for youth and adults (Kessler et al., 2009; Haro et al., 2006), is descended from Diagnostic Interview Schedule (Robins, Heizer, Croughan, & Ratcliff, 1981; Kessler & Ustun, 2004). At the time this chapter was completed (January 2017), most validation of structured interviews focused on psychometric characteristics with respect to DSM-IV (American Psychiatric Association, 2000) diagnoses or diagnoses based on earlier versions of the DSM. For DSM-IV, SCID-IV substance use disorders have strong concurrent and discriminant validity among adults in substance treatment (DeMarce, Lash, Parker, Burke, & Grambow, 2013). In terms of changes, DSM-5 (American Psychiatric Association, 2013) combined abuse and dependence criteria into a single disorder; it also adds craving to diagnoses, and increases the diagnostic threshold for substance disorder. Diagnoses specify different levels of severity (i.e., mild, moderate, and severe) classifications. The most recent is the World Mental Health CIDI (WMH-CIDI; Kessler & Ustun, 2004) but sections of the CIDI that continue to be updated and include Medications and Illegal Substance Use (see www.hcp.med.harvard.edu/wmhcidi/instruments_capi.php). As a WHO instrument, the CIDI has been used worldwide and recently in Nepal (Ghimire, Chardoul, Kessler, Axinn, & Adhikari, 2013). In classifying lifetime substance disorder for community youth and adults, utility estimates are generally excellent (Haro et al., 2006; Kessler et al., 2009). Few studies have examined the vulnerability of structured interviews to the denial or exaggeration of substance abuse. Cottler, Robins, and Helzer (1989) conducted a 1-week follow-up on clients, almost one-third of whom were on parole. They

158

I I .   D i a g n o s t ic I ss u e s

found that 36% of discrepancies occurred due to forgetfulness and 14% due to misunderstanding. Apparently, none were due to underreporting (e.g., to shorten the interview or avoid interviewer disapproval). For interviews, items that address substance use are easily identified, and respondents should have no difficulty in modifying reports, either through exaggeration or denial. Targeted interviews have been developed for substance use. The following section covers three such interviews. ADDICTION SEVERITYINDEX

The Addiction Severity Index (ASI; McLellan, Luborsky, Woody, & O’Brien, 1980) covers medical, work, alcohol, drugs, family/social, legal and psychiatric domains. Chiefly validated with male Veterans Administration (VA) patients, it emphasizes treatment needs. The ASI has been translated into over 20 languages (McLellan, Cacciola, Alterman, Rikoon, & Carise, 2006). For each ASI section, interviewers rate confidence in the respondent’s answers, although no formal definition of invalidity is provided (University of Pennsylvania VA Center for Studies of Addiction [UPVACSA], 1990). McLellan et al. (1980) reported that only 11 of 750 (1.5%) interviews produced invalid information; however, the manual notes that formal work is needed in response distortion (UPVACSA, 1990). Adults in drug treatment underreported 30-day cocaine and opiate use on the ASI as compared to biological testing (Chermack et al., 2000). Interestingly, however, self-administered ASIs (via the Web or telephone voice-response technology [VRT]) have produced significantly higher drug scores than clinician-administered ASIs, despite respondents reporting greater likelihood of being honest with interviewers (Brodey, Rosen, Brodey, Sheetz, Steinfeld, & Gastfriend, 2004). These findings suggest that some substance abusers may minimize drug use when asked via interviews. The Teen ASI (T-ASI; Kaminer, Bukstein, & Tarter, 1991), modeled after the ASI, has been tested on small samples of adolescent inpatients (see Kaminer, Wagner, Plummer, & Seifer, 1993) and has been translated into nine languages (Kaminer, 2008). In contrast to adults, adolescents did not differ significantly in reported substance use between interviewer or self-administered (via the Web or telephone VRT) methods, although they endorsed being the most honest on the Web-based format (Brodey et al., 2005). The T-ASI-2 (Brodey

et al., 2008) expands coverage of psychological functioning; the authors found good internal consistency for alcohol and drug use in youth recruited from substance use clinics. GLOBAL APPRAISAL OFINDIVIDUALNEEDS

The Global Appraisal of Individual Needs—5th edition (GAIN-5; Dennis, White, Titus, & Unsicker, 2008) covers a full biopsychosocial assessment, including substance use, and physical and mental health, including scales for DSM-IV disorders. Numerous variations of the instrument are widely used in SAMHSA programming (Dennis et al., 2008). The GAIN-6 is being developed to address DSM-5. For adults and youth in substance treatment, reliability and construct validity of the substance scale are very robust (Dennis et al., 2008). Its unidimensionality and applicability across gender and setting was generally confirmed on adults in substance treatment (Kenaszchuk, Wild, Rush, & Urbanoski, 2013). GAIN-5 has keyed items alerting interviewers to inconsistencies that can be reconciled with respondents. Interviewers rate the quality of responses per section according to no problems (0), respondent appeared to guess (1), misunderstand (2), deny (3), or misrepresent (4). According to Dennis et al. (2008), variations in scores across sections may indicate biased reporting. However, no published data were found to substantiate this scoring.

Focus ontheMinnesota Multiphasic Personality Inventory andtheMillon Clinical MultiaxialInventory Many multiscale inventories include scales for the assessment of substance abuse and behavioral correlates. This chapter focuses on two inventories that have considerable substance abuse research: the Minnesota Multiphasic Personality Inventory–2 and the Millon Clinical Multiaxial Inventory. MMPI-2

The MMPI-2 (Butcher, Dahlstrom, Graham, Tellegen, & Kaemmer, 1989) and MMPI-2-RF (Restructured Form; Tellegen & Ben-Porath, 2008) are both used widely. Additional formats have included computerized adaptive versions (see Forbey, Ben-Porath, & Arbisi, 2012). Adolescents were administered the MMPI for Adolescent (MMPI-A; Butcher et al., 1992) until the recent release of the

8.  Denial and Misreporting of Substance Abuse 159

MMPI-A-RF (Archer, Handel, Ben-Porath, & Tellegen, 2016). MMPI-2 substance-abuse indicators have included the MacAndrew Alcoholism scale (MAC; MacAndrew, 1965); Substance Abuse Proclivity scale (SAP; MacAndrew, 1986); the revised MAC (MAC-R; Butcher et al., 1989, 1992); Addiction Acknowledgment and Addiction Potential scales (AAS, APS; Weed, Butcher, McKenna, & BenPorath, 1992); the Substance Abuse scale (SUB; Tellegen & Ben-Porath, 2008); and, for adolescents specifically, the Alcohol/Drug Problem Acknowledgment and Proneness scales (ACK, PRO; Weed, Butcher, & Williams, 1994). Despite initially positive results (MacAndrew, 1981), subsequent MAC research found poor specificity, limiting its clinical usefulness (Gripshover & Dacey, 1994). In addition, Otto, Lang, Megargee, and Rosenblatt (1988) demonstrated that the MAC is susceptible to denied substance use. Adolescent studies have generally failed to demonstrate the effectiveness of the MAC (Gantner, Graham, & Archer, 1992; Stein & Graham, 2001). Mixed results have been found for the effectiveness of the MAC-R in accurately detecting adult substance abuse (Clements & Heintz, 2002; Greene, Weed, Butcher, Arredondo, & Davis, 1992; Stein, Graham, Ben-Porath, & McNulty, 1999). The SAP has produced very similar results to MAC, therefore raising doubt regarding SAP’s incremental validity (Greene, 1991) and subsequently its discriminant validity (Greene et al., 1992). The methodological advantage of the APS involves items of varied content unrelated to substance misuse. Data regarding its discriminant validity have yielded variable results in English (e.g., Weed et al., 1992, 1994) and Spanish (Fantoni-Salvador & Rogers, 1997) versions. Despite its nontransparent content, unfortunately, it has not been tested on samples whose members deny or minimize their substance abuse. MMPI-2-RF no longer contains AAS or APS, but introduces the SUB scale (Tellegen & BenPorath, 2008). It tends to have modest predictive validity of substance-related disorders when combined with RC3 (i.e., Cynicism scale; Haber & Baum, 2014). Of particular relevance to this chapter, the SUB appears markedly susceptible to under- and overreporting in both naturalistic and simulation studies (Burchett & Ben-Porath, 2010; Forbey, Lee, Ben-Porath, Arbisi, & Garland, 2013). Clearly, much more work is needed in validating SUB, and for determinations of SUB-specific denial.

As an important caution, practitioners should definitely avoid using measures of general defensiveness as proxy indicators of denied substance abuse. Some substance abusers’ elevations on these scales would represent an impermissible extrapolation to assume that general defensiveness was masking substance abuse. In doing so, examinees without substance abuse issues would be wrongly categorized as substance abusers and their putative “denial” would possibly being viewed as a barrier to treatment interventions. MCMI

The MCMI-IV, the most recent iteration of this instrument, aligns with DSM-5 (Choca & Grossman, 2015). Understandably most studies focus on previous versions with limited generalizability to the MCMI-IV, including Alcohol Use and Drug Use scales (B and T, respectively). The MCMIIV manual (Millon, Grossman, & Millon, 2015) reported moderate to moderately high utility estimates for these scales based on clinician ratings; however, it does not appear that any standardized interviews or methods were used to establish these diagnoses. In addition, no data are provided on how underreporting or denial would affect the psychometric value of these scales. Clearly, crossvalidation is needed. Focusing on the MCMI-III, a large review by Tiet, Finney, and Moos (2008) of the T scale indicated poor sensitivity. In contrast, a study of substance treatment in Denmark (Hesse, Guldager, & Linneberg, 2012) found that B and T scales correlated strongly with alcohol and drug dependence symptom counts (rs > .60), but provided poor agreement with dependence diagnosis. The Millon Adolescent Clinical Inventory (MACI; Millon, 1993) aligns with DSM-IV and includes a Substance Abuse Proneness (BB) scale that distinguishes between clinical youth with and without a substance abuse diagnosis (Grilo, Fehon, Walker, & Martino, 1996; Pinto & Grilo, 2004) and juvenile offenders (Branson & Cornell, 2008). However, BB may have less clinical utility with African American juvenile offenders (Woodland et al., 2014). The MCMI appears to be vulnerable to faking. Respondents instructed to appear psychologically disturbed produced marked elevations on B and T (Millon, 1983) for the original MCMI. In addition, substance abusers appear to be able to conceal substance disorders if so motivated (Fals-Stewart, 1995). In a study to detect denied drug use, 52%

160

II .   D i a g n o s t i c I s s u e s

of drug abusers were able to successfully deny drug or alcohol abuse, although those unable to elude detection tended evidence a more severe course in their disorder (Craig, Kuncel, & Olson, 1994). Of central importance to this chapter, the MCMI-IV and the MACI have yet to publish research on the vulnerability of substance use scales to detect denied or exaggerated substance abuse.

SpecializedMeasures The following specialized measures (see Table 8.3) are included. Specialized Alcohol Measure forAdults DRINKER INVENTORY OFCONSEQUENCES

The Drinker Inventory of Consequences (DrInC; Miller et al., 1995), with 50 items (Likert-type or Yes/No, depending on format), was designed to measure drinking consequences, while avoiding the direct measurement of dependence, pathologic drinking, and help seeking. Normative data and acceptable validity information exist on a large sample of alcohol-dependent treatment-seeking adults. Five subscales (see the SIP, discussed earlier) and a total score are derived for not only the individual’s lifetime but also the past 3 months’ problems related to alcohol misuse. Forms for collateral reports aid in detecting potential differences in reporting. DrInC was validated for heavy-drinking, intravenous drug users, with data suggesting a single factor (Anderson, Gogineni, Charuvastra, Longabaugh, & Stein, 2001). Notably, Miller et al. (1995) cautioned that cultural difference may affect scores and their interpretation. As a means to detect possible dissimulation, carelessness or naysaying is assessed with five items; however, no

published studies have examined utility, or careless or defensive responding. Specialized Alcohol andDrug Measure forAdults THE INVENTORY OFDRUG USECONSEQUENCES

The Inventory of Drug Use Consequences (InDUC; Tonigan & Miller, 2002) was derived from the DrInC using treatment-seeking substance users. It assesses both alcohol and drug consequences for the past 3 months and lifetime. Among substance-involved clients, test–retest reliability was excellent (rs from .89 to .97). Investigators have suggested that the InDUC assesses a single factor, as compared to five (Blanchard, Morgenstern, Morgan, Labouvie, & Bux, 2003; Gillaspy & Campbell, 2006). Like the DrInC, carelessness or naysaying is assessed with five items with unknown validity. Specialized Alcohol andDrug Measures forAdults andAdolescents PERSONAL EXPERIENCEINVENTORY

The Personal Experience Inventory (PEI; Winters & Henly, 1989) includes 33 scales and 300 items that address the severity of drug use and associated psychosocial problems. A PEI—Parent Version (PEI-PV; Winters, Anderson, Bengston, Stinchfield, & Latimer, 2000) demonstrates significant convergence between mother–child reports. The PEI demonstrates good reliability (see Winters & Henly, 1989; Winters, Stinchfield, & Latimer, 2004), and consistent evidence of diagnostic validity (e.g., Winters, Stinchfield, & Henly, 1993; Winters et al., 2004). Similar results are produced across ethnic groups (Winters et al., 2004). The

TABLE 8.3.  Brief Summary of Specialized Measures

Measure

Items

Coverage

Populations

Face valid

Validity scales

DrInC

 50

Alcohol

Adults

Yes

Yes

InDUC

 50

Alcohol/drugs

Adults

Yes

Yes

PEI

300

Alcohol/drugs

Adolescents

Yes

Yes

SASSI

 90a

Alcohol/drugs

Adults/adolescents

Partiallyb

Yes

TLFB

Var.

Alcohol/drugs

Adults/adolescents

Yes

No

Note. See text for full names. Var., variable. a Approximately 90 items, depending on which version is used. b Scales are included that are face valid, whereas other scales are specifically designed to be less face valid.

8.  Denial and Misreporting of Substance Abuse 161

PEI has a Defensiveness scale (11 items), not specific to substance use, and an Infrequency scale (seven items), intended to measure faked drug use by the endorsem*nt of fictitious drugs and improbable statements related to drug use and procurement. Winters (1991) determined cutoff scores for faking bad and found that 5.7% of delinquent youth in drug treatment qualified as faking bad in contrast to 2.6% of high school controls. However, much larger percentages qualified as moderately defensive: 44.5 and 61.6% of drug users and controls, respectively. Subsequently, Winters et al. (1993) adopted a different standard to determine protocol invalidity. Participants who scored in the 90th percentile on either the Infrequency or Defensiveness scales of the normative sample for drug users were classified as invalid. Using this standard, 18.9% were predictably declared invalid for falling beyond 90th percentile. Several later studies, mostly on adolescents in drug treatment, indicated general response distortion rates of about 5–7% (Stinchfield & Winters, 2003; Winters et al., 2004; Winters, Latimer, Stinchfield, & Henly, 1999). No PEI simulation studies have been conducted to determine discriminant validity for response styles. Presently, marked elevations on the Infrequency scale should trigger a more complete evaluation of exaggeration or careless responding. Marked elevations on the Defensiveness scale should not be interpreted as denial or minimization of drug use, since these elevations represent general defensiveness and are more common among nonusers than users. SUBSTANCE ABUSE SUBTLE SCREENINGINVENTORY

The adult versions of the Substance Abuse Subtle Screening Inventory (SASSI) include the original SASSI (Miller, 1985), SASSI-2 (Miller, 1994), SASSI-3 (Miller & Lazowski, 1999), and the recently released SASSI-4 (Lazowski & Geary, 2016). Adolescent versions include the SASSI-A (Miller, 1990) and the SASSI-A2 (Miller & Lazowski, 2001). Understandably, most studies focus on the earlier versions. However, the SASSI-3 is no longer available for purchase. The SASSI measures were intended to be effective with both acknowledged and unacknowledged substance abuse. They contain face-valid (FV) and subtle scales for detecting substance problems and validity indicators, and other scales to inform counselors working with respondents.

Feldstein and Miller (2007) have provided a comprehensive critique of the SASSI measures. Internal consistency ranged remarkably from poor to excellent (alphas ranging from 0.27 to 0.95), with the highest values found for the FV scales. Test–retest reliability also evidenced marked variability. Another critique (Miller, Woodson, Howell, & Shields, 2009) noted the poor alphas for subtle scales, and cautioned about the use of SASSI. The effectiveness of the SASSI versions at detecting substance abuse constitutes a matter of some debate. The SASSI-3 manual (Miller & Lazowski, 1999) has claimed outstanding sensitivity (94%), which has not been confirmed by other investigators. The same appears to be true about the SASSI-4, with a claimed sensitivity of 93% when used with a clinical sample (Lazowski & Geary, 2016). However, Feldstein and Miller (2007) reported sensitivity for the SASSI-3 ranging from poor (33%) in a college sample) to excellent (87%) in an offender sample, with overall average equaling 69.8%. The recently released SASSI-4 will need to be investigated by independent researchers. The SASSI may be the preferred choice over other screeners when examining substance-using offenders. Laux, Piazza, Salyer, and Roseman (2011) found the SASSI-3 subtle scales improved overall sensitivity by 10.5–36.4%, above and beyond just the FV scales alone. The SASSI has been studied with response styles. While not usually a clinical concern, participants can overreport substance abuse (Myerholtz & Rosenberg, 1998). With respect with faking good, these same investigators found an alarming 71% of offenders with substance dependence avoided detection on the SASSI-2. College students, attempting to simply fake good rather than specifically deny substance abuse, suppress most SASSI-3 scales and show an elevation the Defensiveness (DEF) scale (Burck, Laux, Harper, & Ritchie, 2010). Refinement of SASSI detection strategies appears to be warranted. Using inpatients, Wooley et al. (2012) studied the SASSI-3 under standard instructions, partial denial, and complete denial of drug use. Only one of the subtle scales, the Subtle Attributes scale (SAT), was modestly helpful in detecting partial but not complete denial. However, the DEF scale was significantly elevated for both denial groups. To differentiate between partial and total denial, these researchers constructed a SAT – DEF index (i.e., subtracting DEF from SAT), which produced moderately robust utility estimates.

162

II .   D i a g n o s t ic I ss u e s

Lazowski and Geary (2016) studied the SASSI4 and reported a sensitivity rate of .79 among substance abusers instructed to “fake good” and further claimed that only four out of 120 successfully denied drug use. However, this claim could be questioned, because 23 of 115 participants in the fake good condition were excluded from the classifications because of inconsistent responding. While not yielding valid test data, inconsistent responding cannot be assumed to represent the denial of substance abuse. TIMELINEFOLLOWBACK

The Timeline Followback (TLFB; Fals-Stewart, O’Farrell, Freitas, McFarlin, & Rutigliano, 2000) uses a calendar method to evaluate daily patterns of alcohol and drug use. It has been used with adults and adolescents (Hjorthoj, Hjorthoj, & Nordengoft, 2012). Good reliability and validity data exist and a collateral form can be used (see Fals-Stewart et al., 2000). Validity has been established with biological testing, record review, and collateral reports (Fals-Stewart et al., 2000; O’Farrell & Langenbucher, 1988). Meta-analytic work demonstrates that the TLFB generally agrees with biological measures, although higher agreement may be found among persons without psychiatric comorbidity (Hjorthoj et al., 2012). It can be used to collect psychometrically sound substance data for up to 12 months (Robinson, Sobell, Sobell, & Leo, 2014). The face valid nature of its interview questions is vulnerable to faking or denial; however, the detailed nature of TLFB may make it hard for respondents to reliably fabricate a convincing pattern of substance use over multiple interviews covering a specified time period. In contrast, denial would be relatively straightforward. No relationship has been observed between TLFB substance use indices and a general measure of social desirability in adults seeking drug treatment (Fals-Stewart et al., 2000). However, as previously underscored, the lack of a relationship does not address the specific denial of substance abuse.

BIOCHEMICALMETHODS Three common methods of detecting substance use are reviewed in this chapter, including breathalyzer, urinalysis, and hair analysis. Eye-scanning, more aptly described as physiological than biochemical, is also noted. Other biochemical methods are not

addressed: blood and nail sampling, skin sensors to detect metabolic processes via sweat, and saliva sampling (see American Society of Addiction Medicine [ASAM], 2013). Biological detection of drug use usually involves screening via immunoassay (IA), which, if positive, is followed by confirmatory testing via gas chromatography/mass spectrometry, liquid chromatography/mass spectrometry, or tandem mass spectrometry (GC-MS, LC-MS, or LC-MS/MS). Alternatively, LC-MS/MS may be used in a single step to identify a far wider range of drugs than is possible with IA (ASAM, 2013). Box 8.1 provides a summary of the methods. IA uses antigen–antibody interactions to compare the specimen with a calibrated quantity of the drug being tested (SAMHSA, 2012). Drug concentrations in various specimens are highly variable and depend on many factors, such as amount of drug consumed, time since use, metabolism rate, body fat, and consumption of liquids (see Jaffee, Trucco, Levy, & Weiss, 2007). The general detection window varies by the sample types. As a general benchmark, the following ranges are presented: 1–36 hours for saliva, 1–3 days for urinalysis, 7–100+ days for hair-analysis, and 1–14 days for sweat via continuous monitoring (Dolan, Rouen, & Kimber, 2004). Detection times for blood and breath are very brief, usually on the order of several hours (ASAM, 2013; SAMHSA, 2012). BOX 8.1.  Laboratory Methods for Detecting Substance Use Screen • Immunoassay (IA). IA is a screener that uses antibodies to detect certain substances in urine. It does not measure amount of sub‑ stance in urine.

Definitive measures • Gas chromatography/mass spectrometry (GCMS). GC-MS provides identification of specific metabolites that are present in different types of drugs. • Liquid chromatography/mass spectrometry (LC-MS). LC-MS scans the breath of subjects to detect specific formations of aerosol pat‑ terns, which, if contaminated by drugs, will be different depending on type of drug ingested. • Tandem mass spectrometry (TMS). TMS allows for identification of specific compounds and metabolites that can differentiate, for example, different medications within a class of opiates.

8.  Denial and Misreporting of Substance Abuse 163

Breathalyzer Although under development for other substances, breath analysis is used primarily for rapid alcohol detection in the field (ASAM, 2013). The provided estimate of BAC is based on metabolism, which may be affected by gender, age, physical condition (especially the liver), and weight (SAMHSA, 2012). Because BAC is employed to define intoxication, clinicians must be careful to distinguish between legal intoxication and clinical intoxication, for which BACs are highly variable. Legal intoxication is generally set at BAC ≥ 80 mg/dL for adults (NIAAA, 2016b). For clinical intoxication, behavioral observations likely include one or more of the following: slurred speech, poor coordination, unsteady gait, nystagmus, attention and memory problems, and stupor/coma (American Psychiatric Association, 2013). At least some degree of tolerance is presumed if BAC ≥ 150 mg/ dL with no signs of intoxication. To put this in perspective, levels at 200 mg/dL render most nontolerant persons extremely intoxicated (American Psychiatric Association, 2013). Body temperature and breathing patterns can affect breath alcohol test results (SAMHSA, 2012). BAC obtained from breath may underestimate actual BAC (Kapur, 1994) by approximately 8.5% (Garriott, 2008), although a substantial minority of results (19–23%) may represent a marked overestimate of BAC (Simpson, 1987). Because of residual alcohol vapor in the mouth, artificially high readings are possible if readings are taken within 20 minutes of alcohol consumption. According to Watson and colleagues (2006), with properly sampled breath, BAC should be highly accurate. Few technical problems occur in breathalyzer administration, particularly with the widely adopted computer-based models. Although concerns have been raised regarding the effective maintenance of breathalyzers (Trichter, McKinney, & Pena, 1995), these concerns can easily be addressed through documentation and service records. False-positive readings may be created by exposure to alcohol-related products. For example, the gas additive, methyl tert-butyl ether (MBTE), may result in false positives on commercial breathalyzers (Buckley, Pleil, Bowyer, & Davis, 2001). This finding may be particularly germane for persons working with gasoline and other fuels oxygenated with MBTE (e.g., auto mechanics, gas station attendants). However, breathalyzer technology employing infrared and electrochemical detec-

tors may mitigate this concern. Alarmingly, use of common alcohol-based hand sanitizers and mouthwashes by persons operating a breathalyzer may cause false-positive readings even when sanitizer is used correctly (Ali et al., 2013).

Urinalysis Urinalysis, the most widely used method (Dolan et al., 2004; SAMHSA, 2012), can detect drugs including amphetamines, barbiturates, benzodiazepines, cocaine metabolites, methadone, phencyclidine (PCP), morphine, ethanol and cannabinoids (Morgan, 1984). It is generally effective, with a detection window of about 1–3 days, except for cannabis, which may be detected for 30 days or more (Musshoff & Madea, 2006; SAMHSA, 2012). The detection window depends on a variety of factors, including drug type, users’ characteristics (e.g., body mass), short or long-term drug use, and urine pH (Moeller, Lee, & Kissack, 2008; ASAM, 2013; SAMHSA, 2012). Of course, sample integrity should be maintained by using proper storage (i.e., refrigeration), if there is lag between collection and testing (ASAM, 2013; SAMHSA, 2012). Because ethanol is rapidly metabolized and eliminated from the body, methods that depend on sampling breath, blood, urine, and saliva are somewhat limited to detection of alcohol consumption, even within the past few hours (Jones, 1993). Therefore, urinalysis is not generally utilized to detect alcohol consumption (Moeller et al., 2008). However, some methods do allow for detection of even moderate amounts of alcohol (roughly 50 g, or a little less than about five bottles of beer) in urine within approximately 24 hours (Helander et al.,1999), and other methods may extend the window to 72 hours (ASAM, 2013). Accuracy of urinalysis may be affected by at least three factors: First, medication (e.g., nasal spray or diet agents) may lead to positive test results for amphetamine (SAMHSA, 2012). Second, foods (e.g., poppy seeds) may produce positive opiate results (Tenore, 2010). Third, certain teas (e.g., teas made from coca leaves) may yield positive cocaine results (SAMHSA, 2012). The accuracy of urinalysis may be substantially affected by methods to avoid detection. The simplest involves dilution by drinking copious amounts of water to produce false negatives. Adulterants can be ingested before urination (e.g., diuretic agents) or posturination (e.g., bleach) to decrease drug detection (Jaffee et al., 2007). Covert urine substitution (“clean” urine substituted

164

II .   D i a g n o s t ic I ss u e s

for a “dirty” specimen) can be very effective if not observed (Jaffee et al., 2007). Certain steps can help to ensure an accurate appraisal of substance abuse results. For example, a detailed history of medications can assist in ferreting out unexpected positive results, as can sophisticated confirmatory tests (Moeller et al., 2008; SAMHSA, 2012). In addition, direct observation of urine collection, while intrusive, can minimize risk of urine substitution and addition of adulterants posturination (Jaffee et al., 2007). Whether ingested or added after voiding, adulterants can be detected via several methods. Specimen validity tests compare characteristics of the specimen with acceptable density and composition ranges for human urine, and test for adulterants (SAMHSA, 2012). During collection, specimen should be checked for color and temperature (SAMHSA, 2012), although color may be altered by food, medication, and other factors (ASAM, 2013).

HairAnalysis Hair follicles absorb drugs and metabolites from blood and sweat glands (Musshoff & Madea, 2006). As hair grows (at about 1 cm per month), it chronicles substance use, so that laboratories can estimate extent of use (Dolan et al., 2004). Hair analysis for illicit drugs began in 1979 with radioimmunoassay (RIA); methodologies expanded through the 1980s to include LC and MS; today, GC-MS, LC-MS and LC-MS/MS are “gold standards,” although LC-MS/MS has demonstrated superiority (Liu, Liu, & Lin, 2015). Steps in hair analysis generally involve collection, decontamination, grinding or cutting the sample, dissolving the hair, extraction, then analysis (Kintz, Villain, & Cirimele, 2006). Most laboratories analyze 3 months of hair growth, but it may be possible to detect substances over a year postuse (SAMHSA, 2012). Hair analysis appears to be the most reliable method for detecting frequent and heavy use of cocaine, opioids, amphetamine, PCP, and ecstasy, but it is understandably less suited for detection of occasional drug use or binge use (SAMHSA, 2012). For occasional use of these drugs, a low cutoff score can improve classification (Gryczynski, Schwartz, Mitchell, O’Grady, & Ondersma, 2014). Hair analysis may also be useful with marijuana (Han, Chung, & Song, 2012; Musshoff & Madea, 2006) but not alcohol use (ASAM, 2013; Cooper, Kronstrand, & Kintz, 2012). Hair analysis for the detection of denied substances has its limitations. Drugs can move down

the hair shaft via sweat, confounding use of hair to formulate chronology of drug use. Additionally, drugs or metabolites may be found in hair from simple environmental exposure (Moosmann, Roth, & Auwarter, 2015; SAMHSA, 2012). Moreover, elements in the environment (e.g., sunlight) may degrade hair, producing false negatives (Pragst & Balikova, 2006; Suwannachom, Thananchai, Junkuy, O’Biren, & Sribanditmongkol, 2015). However, the hair sample undergoes extensive chemical washing that can mitigate contamination or environmental effects. Additional challenges to accurate hair analysis include variations in hair structure, growth rate, melanin, hygiene, and cosmetic hair treatment (e.g., bleach; Dasgupta, 2008). For example, dark pigmented hair tends to bind greater amounts of drug than less pigmented hair (i.e., blonde). A strength of hair analysis involves its ability to detect positive results, even if the individual has abstained for weeks (SAMHSA, 2012). In fact, chronic methamphetamine use appears to be detectable for up to 5 months after abstinence (Suwannchom et al., 2015). In summary, hair analysis is well suited to detection of frequent and heavy illicit drug use. Nonetheless, it can sometimes be used to detect low or single doses, as in drug-facilitated crimes (Cooper et al., 2012; Xiang, Shen, & Drummer, 2015). Newer techniques are expanding the spectrum of drugs that can be tested in hair, even at single or low doses (Montesano, Johansen, & Nielsen, 2014). For measuring treatment effectiveness, it can also be used to indicate periods of abstinence (SAMHSA, 2012). As a laboratory measure, hair analysis cannot be expected to provide information about behavioral responses or psychological sequelae to substance abuse.

EyeScanning Eye scanning is a relatively new method to assist in drug screens for denied substance use. Equipment can range from binocular size to the size of an automatic teller machine (Tahir & Bakri, 2006). Some require baseline readings when participants are drug-free, whereas others do not (Fazari, 2011; Richman & Noriega, 2002). The primary method used in a pupilometer drug screening is based on reactions to flashing lights: specifically, comparing the individual’s baseline eye reaction to controlled amounts of light to the current reaction. In addition, examination of retinal movement tracking may also be employed (Fazari, 2011). This type of screening can also

8.  Denial and Misreporting of Substance Abuse 165

identify the type of drug used as long as use occurred within the last 48 hours (Fazari, 2011; Tahir & Bakri, 2006). The advantages of pupilometer technology over other laboratory methods include relative ease and brevity. This technique requires only 30 seconds to complete and is a fully automated system. It is noninvasive, and it is presumably difficult to fake or alter results. For probationers, who require frequent substance use monitoring, this method is particularly useful. A further advantage involves cost savings as compared to other methods, which is particularly important for frequent monitoring (Fazari, 2011; Richman & Noriega, 2002). Challenges to accurate assessments via eye scanning may include situational factors (e.g., fatigue or diet), medication, and medical conditions (Fazari, 2011; Hoskins, 2005; Tahir & Bakri, 2006). More work is needed to better understand possible racial, age, and gender differences in detection rates, as this technique is still relatively new and underutilized (Fazari, 2011).

CLINICALAPPLICATIONS Many, if not most, diagnoses of substance use disorders are missed simply because clients are not asked about these disorders (see NIAAA, 2015). In a classic study of 705 patients with alcoholism, Sobell, Sobell, and Vanderspek (1979) found that clinicians were not particularly effective at detecting patients who were under the influence of alcohol. Of those patients denying intoxication, clinicians correctly identified only between 50 and 67%, as subsequently verified by breath analysis. Their false positives ranged across samples from 0 to 17%. Clearly, clinical observation alone is insufficient for the detection of substance abuse. Screens and many specialized substance use measures trustingly assume forthrightness and complete self-disclosure about not only about the type and frequency of substance abuse but also its short- and long-term sequelae. Systematic screening is likely to be helpful in discovering undiagnosed cases of substance abuse. Toward this objective, systematic use of screens, such as AUDIT, DAST, or GAIN-SS, is recommended. Minimized substance use may be evidenced in marked discrepancies among a variety of sources. These include (1) acknowledgment of use with peers but denial with authority figures; (2) minimized history of use in contradiction to treatment/ arrest records; and (3) denial of use in contradic-

tion to performance on face-valid measures of substance use. Because inconsistencies in reported drug use may have many determinants (e.g., confusion secondary to drug use, unawareness of consumed drugs), discrepancies alone should not be considered evidence of dissimulation. Similarly, exaggerated substance use may be evidenced in marked discrepancies across several sources of information, including (1) denial of substance use with peers but exaggeration with authorities; (2) gross exaggeration of past substance use in contradiction of past treatment/arrest records; (3) endorsem*nt of fictitious drugs; and (4) ascribing very atypical behavioral effects to known drug use (e.g., prolonged hallucinatory experiences from the use of marijuana). Careful inquiry into atypical effects may be helpful in distinguishing purely self-serving explanations from more believable descriptions. For instance, use of hallucinogens is unlikely to produce command hallucinations focused solely on execution of a well-planned bank robbery. In particular, caution must be applied in drawing conclusions about the endorsem*nt of fictitious drugs and deliberate exaggeration. Drug users may be motivated to appear knowledgeable about unfamiliar drugs. Alternatively, they may become confused by the myriad street and scientific terms. To identify deliberate exaggerations, clinicians can easily test the limits of credibility via a rare symptom approach. For example, queries could include very uncommon effects (e.g., accelerated hair growth) to different fictitious drugs. Clinical decision models assist in determining whether persons reporting substance use are engaged in dissimulation. For denied or minimized drug use, several sources provide useful information: • Independent witnesses who disconfirm the respondent’s denial or minimization of substance use for a specific event. • Positive alcohol breathalyzer results covering a very limited time period (2–12 hours), which depend on level of intoxication and metabolism rates. • Positive urinalysis results covering a circ*mscribed period of time (36–72 hours), although marijuana may be detected for more than 30 days. • Positive hair analysis covering an extended period of time (typically from 7 days to 3+ months). • Positive eye scanning for drug use (48 hours). • Observed data (biochemical concentrations or behavioral observations) that disconfirm the

166

II .   D i a g n o s t ic I ss u e s

respondent’s minimized substance use for a specific event or period of time. These are data sources within the clinical decision model. More specifically, many defensiveness scales are not specific to substance use and lack discriminant validity with respect to denied drug use (e.g., PEI DEF scale). Similarly, SASSI scales appear to have limited discriminant validity, although this needs to be further tested with the SASSI-4. Future work might establish utility estimates for promising instruments using a range of cutoff scores for a given setting. For exaggerated drug use, several sources provide useful information in clinical decision making, including the following: • Independent witnesses who disconfirm respondent-reported substance use for a specific event. • Negative alcohol breathalyzer results, covering a very limited time period (2–12 hours), that depend on level of intoxication and metabolism rates. • Negative urinalysis results covering a circ*mscribed period of time (36–72 hours), although marijuana may be detected for more than 30 days. • Negative hair analysis covering an extended period of time (from 7 days to 3+ months). • Observed data (biochemical concentrations or behavioral observations) that disconfirm the respondent’s exaggerated substance use for a specific event or period of time.

SUMMARY This chapter represents the breadth and diversity of methods that have been applied to the assessment of substance abuse. Sadly, very few advances have been observed in the last decade for psychometric assessment of illicit drugs and nonprescribed use of medications. Most measures appear to be based on completely unwarranted and professionally naive assumption that examinees will be entirely forthcoming about their substance abuse. According to Delaney-Black et al. (2010), as a stark example, parents in a high-risk sample had 650% more use of cocaine than they reported, which was virtually eclipsed by their teenagers (5,200% more use). However, more sophisticated detection strategies hold promise. As illustrated by Wooley et al. (2012) unlikely patterns between substance abuse scales may diminish face validity and improve the detection of denied substance abuse.

Laboratory-based methods of assessing substance use continue to develop and have become more sophisticated. For day-to-day use, their methods are often challenging to administer, vulnerable to countermeasures, and very expensive. The use of eye-scanning technology for drug use—despite being available for more than a decade—has not garnered the systematic research that it clearly deserves. This technology stands out because of its simplicity, low cost, and resistance to countermeasures.

REFERENCES Ali, S. S., Wilson, M. P., Castillo, E. M., Witucki, P., Simmons, T. T., & Vilke, G. M. (2013). Common hand sanitizer may distort readings of breathalyzer tests in the absence of acute intoxication. Academy of Emergency Medicine, 20(2), 212–215. Allensworth-Davles, D., Cheng, D. M., Smith, P. C., Samet, J. H., & Saitz, R. (2012). The Short Inventory of Problems—modified for drug use (SIP-DU): Validity in a primary care sample. American Journal on Addictions, 21, 257–262. Alterman, A. I., Cacciola, J. S., Ivey, M. A., Habing, B., & Lynch, K. G. (2009). Reliability and validity of the Alcohol Short Index of Problems and a newly constructed drug short index of problems. Journal of Studies Alcohol and Drugs, 70(2), 304–307. American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: Author. American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: Author. American Society of Addiction Medicine. (2013). Drug testing: A white paper of the American Society of Addiction Medicine. Chevy Chase, MD: Author. Anderson, B. J., Gogineni, A., Charuvastra, A., Longabaugh, R., & Stein, M. D. (2001). Adverse drinking consequences among alcohol abusing intravenous drug users. Alcoholism: Clinical and Experimental Research, 25(1), 41–45. Archer, P. A., Handel, R. W., Ben-Porath, Y. S., & Tellegen, A. (2016). Minnesota Multiphasic Personality Inventory—Adolescent Restructured Form professional manual. Minneapolis: University of Minnesota Press. Axelson, D., Birmaher, B., Zelazny, J., Kaufman, J., & Gill, M. K. (2009). K-SADS-PL: 2009 working draft. Retrieved April 21, 2016, from www.psychiatry.pitt. edu/sites/default/files/documents/assessments/ksadspl_2009_working_draft_full.pdf. Babor, T. F., Biddle, J. C., Saunders, J. B., & Monteiro, M. G. (2001). AUDIT: The Alcohol Use Disorders Identification Test, guidelines for use in primary care (2nd ed.). Geneva: World Health Organization. Blanchard, K. A., Morgenstern, J., Morgan, T. J., Labouvie, E. W., & Bux, D. A. (2003). Assessing conse-

8.  Denial and Misreporting of Substance Abuse 167

quences of substance use: Psychometric properties of the Inventory of Drug Use Consequences. Psychology Addictive Behavior, 17, 328–331. Branson, C. E., & Cornell, D. G. (2008). Correspondence of the Millon Adolescent Clinical Inventory (MACI) and the Personality Inventory for Youth (PIY) in juvenile offenders. Journal of Forensic Psychology Practice, 7(4), 65–100. Brodey, B. B., McMullin, D., Kaminer, Y., Winters, K. C., Mosshart, E., Rosen, C. S., et al. (2008). Psychometric characteristics of the Teen Addiction Severity Index–2 (T-ASI-2). Substance Abuse, 29(2), 19–32. Brodey, B. B., Rosen, C. S., Brodey, I. S., Sheetz, B. M., Steinfeld, R. R., & Gastfriend, D. R. (2004). Validation of the Addiction Severity Index (ASI) for internet and automated telephone self-report administration. Journal of Substance Abuse Treatment, 26, 253–259. Brodey, B. B., Rosen, C. S., Winters, K. C., Brodey, I. S., Sheetz, B. M., Steinfeld, R. R., et al. (2005). Conversion and validation of the Teen-Addiction Severity Index (T-ASI) for internet and automated-telephone self-report administration. Psychology of Addictive Behaviors, 19(1), 54–61. Buchan, B. J., Dennis, M. L., Tims, F. M., & Diamond, G. S. (2002). Cannabis use: Consistency and validity of self-report, on-site urine testing and laboratory testing. Addiction, 97(Supp.1), 98–108. Buckley, T. J., Pleil, J. D., Bowyer, J. R., & Davis, J. M. (2001). Evaluation of methyl tert-butyl ether (MTBE) as an interference on commercial breath-alcohol analyzers. Forensic Science International, 123, 111–118. Burchett, D. L., & Ben-Porath, Y. S. (2010). The impact of over-reporting on MMPI-2-RF Substantive scale score validity. Assessment, 4, 497–516. Burck, A. M., Laux, J. M., Harper, H., & Ritchie, M. (2010). Detecting faking good and faking bad with the Substance Abuse Subtle Screening Inventory–3 in a college student sample. Journal of College Counseling, 13, 63–72. Bureau of Justice Statistics. (2010). Federal Justice Statistics, 2010. Washington, DC: U.S. Department of Justice. Butcher, J. N., Dahlstrom, W. G., Graham, J. R., Tellegen, A., & Kaemmer, B. (1989). MMPI-2: Minnesota Multiphasic Personality Inventory–2: Manual for administration and scoring. Minneapolis: University of Minnesota Press. Butcher, J. N., Williams, C. L., Graham, J. R., Archer, R. P., Tellegen, A., Ben-Porath, Y. S., et al. (1992). MMPI-A: Minnesota Multiphasic Personality Inventory–A: Manual for administration scoring, and interpretation. Minneapolis: University of Minnesota Press. Cassidy, C. M., Schmitz, N., & Malla, A. (2007). Validation of the Alcohol Use Disorders Identification Test and the Drug Abuse Screening Test in first episode psychosis. Canadian Journal of Psychiatry, 53(1) 26–33. Chermack, S. T., Roll, J., Reilly, M., Davis, L., Kilaru, U., & Grabowski, J. (2000). Comparison of patient

self-reports and urinalysis results obtained under naturalistic methadone treatment conditions. Drug and Alcohol Dependence, 59(1), 43–49. Cherpitel, C. J. (2000). A brief screening instrument for problem drinking in the emergency room: The RAPS4. Journal on Studies on Alcoholism, 61, 447– 449. Cherpitel, C. J. (2002). Screening for alcohol problems in the U.S. general population: Comparison of the CAGE, and RAPS4, and RAPS4-QF by gender, ethnicity, and services utilization. Alcohol Clinical Experimental Research, 26, 1686–1691. Cherpitel, C. J., Ye, Y., Moskalewicz, J., & Swiatkiewicz, G. (2005). Screening for alcohol problems in two emergency service samples in Poland: Comparison of the RAPS4, CAGE and AUDIT. Drug and Alcohol Dependence, 80, 201–207. Choca, J. P., & Grossman, S. D. (2015). Special section: Theodore Millon’s legacy for personality theory and assessment evolution of the Millon Clinical Multiaxial Inventory. Journal of Personality Assessment, 97(6), 541–549. Clements, R., & Heintz, J. M. (2002). Diagnostic accuracy and factor structure of the AAS and APS scales of the MMPI-2. Journal of Personality Assessment, 79(3), 564–582. Colon, H. M., Robles, R. R., & Sahai, H. (2002). The validity of drug use self-reports among hard core drug users in a household survey in Puerto Rico: Comparison of survey responses of cocaine and heroin use with hair tests. Drug and Alcohol Dependence, 67, 269–279. Connor, J. P., Grier, M., Feeney, G. F., & Young, R. M. (2007). The validity of the Brief Michigan Alcohol Screening Test (bMAST) as a problem drinking severity measure. Journal of Studies on Alcohol and Drugs, 68(5), 771–779. Cook, R. L., Chung, T., Kelly, T. M., & Clark, D. B. (2005). Alcohol screening in young persons attending a sexually transmitted disease clinic: Comparison of AUDIT, CRAFFT, and CAGE instruments. Journal of General Internal Medicine, 20(1), 1–6. Cooper, G. A., Kronstrand, R., & Kintz, P. (2012). Society of hair testing guidelines for drug testing in hair. Forensic Science International, 218(1–3), 20–24. Cottler, L. B., Robins, L. N., & Helzer, J. E. (1989). The reliability of the CIDI-SAM: A comprehensive substance abuse interview. British Journal of Addiction, 84(7), 801–814. Craig, R. J., Kuncel, R., & Olson, R. E. (1994). Ability of drug abusers to avoid detection of substance abuse on the MCMI-II. Journal of Social Behavior and Personality, 9, 95–106. Cremonte, M., Ledesma, R. D., Cherpitel, C. J., & Borges, G. (2010). Psychometric properties of alcohol screening tests in the emergency department in Argentina, Mexico and the United States. Addictive Behaviors, 35(9), 818–825. Dalla-Déa, H. R. F., De Micheli, D., & Formigoni, M. L. (2003). Effects of identification and usefulness of

168

II .   D i a g n o s t ic I ss u e s

the Lie Scale of the Drug Use Screening Inventory (DUSI-R) in the assessment of adolescent drug use. Drug and Alcohol Dependence, 72, 215–223. Dasgupta, A. (Ed.). (2008). Handbook of drug monitoring methods: Therapeutics and drugs of abuse. Totowa, NJ: Humana Press. De Marce, J. M., Lash, S. J., Parker, J. D., Burke R. S., & Grambow, S. C. (2013). Validity of the Structured Clinical Interview for DSM-IV among veterans seeking treatment for substance use disorders. International Journal of Mental Health and Addiction, 11, 546–556. De Micheli, D., & Formigoni, M. L. (2002). Psychometrics properties of the Brazilian version of DUSI (Drug Use Screening Inventory). Alcohol Clinical and Experimental Research, 26(10), 1523–1528. de Wit, H., Uhlenhuth, E. H., Pierri, J., & Johanson, C. E. (1987). Individual differences in behavioral and subjective responses to alcohol. Alcoholism: Clinical and Experimental Research, 11(1), 52–59. Delaney-Black, V., Chiodo, L. M., Hannigan, J. H., Greenwald, M. K., Janisee, J., Patterson, G., et al. (2010). Just Say “I don’t”: Lack of concordance between teen report and biological measures of drug use. Pediatrics, 126(5), 887–893. Dennis, M. L., Chan, Y., & Funk, R. R. (2006). Development and validation of the GAIN Short Screener (GSS) for internalizing, externalizing and substance use disorders and crime/violence problems among adolescents and adults. American Journal on Addiction, 15, 80–91. Dennis, M. L., Feeney, T., & Titus, J. C. (2013). Global Appraisal of Individual Needs—Short Screener (GAINSS): Administration and scoring manual, version 3. Bloomington, IL: Chestnut Health Systems. Dennis, M. L., White, M., Titus, J. C., & Unsicker, J. (2008). Global Appraisal of Individual Needs (GAIN): Administration guide for the GAIN and related measures (version 5). Bloomington, IL: Chestnut Health Systems. DiThomas, M. (2012, October). Drug test promotes workplace safety. Occupational Health and Safety. Retrieved April 9, 2016, from https://ohsonline.com/ articles/2012/10/01/drug-testing-promotes-workplacesafety.aspx. Dolan, K., Rouen, D., & Kimber, J. (2004). An overview of the use of urine, hair, sweat and saliva to detect drug use. Drug and Alcohol Review, 23, 213–217. Donovan, D. M., Kivlahan, D. R., Doyle, S. R., Longabaugh, R., & Greenfield, S. F. (2006). Concurrent validity of the Alcohol Use Disorders Identification Test (AUDIT) and AUDIT zones in defining levels of severity among out-patients with alcohol dependence in the COMBINE study. Addiction, 101(12), 1696–1704. Fals-Stewart, W. (1995). The effect of defensive responding by substance-abusing patients on the Million Clinical Multiaxial Inventory. Journal of Personality Assessment, 64(3), 540–551.

Fals-Stewart, W., O’Farrell, T. J., Freitas, T. T., McFarlin, S. K., & Rutigliano, P. (2000). The timeline followback reports of psychoactive substance use by drugabusing patients: Psychometric properties. Journal of Consulting and Clinical Psychology, 68(1), 134–144. Fantoni-Salvador, P., & Rogers, R. (1997). Spanish versions of the MMPI-2 and PAI: An investigation of concurrent validity with Hispanic patients. Assessment, 4(1), 29–39. Fazari, G. M. (2011). Testing the validity of pupillometer technology against traditional drug screening instruments. Federal Probation a Journal of Correctional Philosophy and Practice, 75(3), 37–44. Feinn, R., Tennen, H., & Kranzler, H. R. (2003). Psychometric properties of the Short Index of Problems as a measure of recent alcohol-related problems. Alcoholism: Clinical and Experimental Research, 27(9), 1436–1441. Feldstein, S. W., & Miller, W. R. (2007). Does subtle screening for substance abuse work?: A review of the Substance Abuse Subtle Screening Inventory (SASSI). Addiction, 102(1), 41–50. First, M. B., Williams, J. B. W., Karg, R. S., & Spitzer, R. L. (2016). Structured Clinical Interview for DSM-5 Disorders, Clinician Version (SCID-CV). Washington, DC: American Psychiatric Press. Forbey, J. D., Ben-Porath, Y. S., & Arbisi, P. A. (2012). The MMPI-2 computerized adaptive version (MMPI2-CA) in a Veterans Administration medical outpatient facility. Psychological Assessment, 24, 628–639. Forbey, J. D., Lee, T. T., Ben-Porath, Y. S., Arbisi, P. A., & Garland, D. (2013). Associations between MMPI2-RF validity scale scores and extra-test measures of personality and psychopathology. Assessment, 20(4), 448–461. Gantner, A. B., Graham, J. R., & Archer, R. R. (1992). Usefulness of the MAC scale in differentiating adolescents in normal, psychiatric, and substance abuse settings. Psychological Assessment, 4, 133–137. Garriott, J. C. (Ed.). (2008). Garriott’s medico-legal aspects of alcohol (5th ed.). Tucson, AZ: Lawyers & Judges Publishing. Geneste, J., Pereira, B., Arnaud, B., Christol, J., Liotier, J., Blanc, O., et al. (2012). CAGE, RAPS4, RAPS4QF and AUDIT screening tests for men and women admitted for acute alcohol intoxication to an emergency department: Are standard thresholds appropriate? Alcohol and Alcoholism, 47(3), 273–281. Ghimire, D. J., Chardoul, S., Kessler, R. C., Axinn, W. G., & Adhikari, B. P. (2013). Modifying and validating the Composite International Diagnostic Interview (CIDI) for use in Nepal. International Journal of Methods in Psychiatric Research, 22(1), 71–81. Gillaspy, J. A., & Campbell, T. C. (2006). Reliability and validity of scores from the Inventory of Drug Use Consequences. Journal of Addictions and Offender Counseling, 27, 17–27. Greene, R. L. (1991). The MMPI-2/MMPI: An interpretive manual. Boston: Allyn & Bacon.

8.  Denial and Misreporting of Substance Abuse 169

Greene, R. L., Weed, N. C., Butcher, J. N., Arredondo, R., & Davis, H. G. (1992). A cross-validation of MMPI-2 substance abuse scales. Journal of Personality Assessment, 58, 405–410. Grilo, C. M., Fehon, D. C., Walker, M., & Martino, S. (1996). A comparison of adolescent inpatients with and without substance abuse using the Millon Adolescent Clinical Inventory. Journal of Youth and Adolescence, 25(3), 379–388. Gripshover, D. L., & Dacey, C. M. (1994). Discriminative validity of the MacAndrew scale in settings with a high base rate of substance abuse. Journal of Studies on Alcohol, 55, 303–308. Grisso, T., & Barnum, R. (2006). Massachusetts Youth Screening Instrument–Version 2: User’s manual and technical report. Sarasota, FL: Professional Resource Press. Gryczynski, J., Kelly, S. M., Mitchell, S. G., Kirk, A., O’Grady, K. E., & Schwartz, R. P. (2014). Validation and performance of the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST) among adolescent primary care patients. Addiction, 110(2), 240–247. Gryczynski, J., Schwartz, R. P., Mitchell, S. G., O’Grady, K. E., & Ondersma, S. J. (2014). Hair drug testing results and self-reported drug use among primary care patients with moderate-risk illicit drug use. Drug and Alcohol Dependence, 141, 44–50. Haber, J., & Baum, L. (2014). MMPI-2-RF scales as predictors of psychiatric diagnosis. South African Journal of Psychology, 44(4), 439–453. Han, E., Chung, H., & Song, J. M. (2012). Segmental hair analysis for 11-nor-D9-tetrahydrocannabinol9carboxylic acid and the patterns of cannabis use. Journal of Analytical Toxicology, 36, 195–200. Haro, J. M., Arbabzadeh-Bouchez, S., Brugha, T. S., De Girolamo, G., Guyer M., Jin, R., et al. (2006). Concordance of the Composite International Diagnostic Interview Version 3.0 (CIDI 3.0) with standardized clinical assessments in the WHO World Mental Health surveys. International Journal of Methods Psychiatry Research, 15, 167–180. Hayes, M. A., McReynolds, L. S., & Wasserman, G. A. (2005). Paper and voice MAYSI-2: Format comparability and concordance with the voice DISC-IV. Assessment, 12(4), 395–403. Helander, A., von Wachenfeldt, J., Hiltunen, A., Beck, O., Liljeberg, P., & Borg, S. (1999). Comparison of urinary 5-hydroxytryptophol, breath-ethanol, and self-report for detection of recent alcohol use during outpatient treatment: A study on methadone patients. Drug and Alcohol Dependence, 56, 33–38. Hesse, M., Guldager, S., & Linneberg, I. H. (2012). Convergent validity of MCMI-III clinical syndrome scales. British Journal of Clinical Psychology, 51, 172– 184. Hjorthoj, C. R., Hjorthoj, A. R., & Nordengoft, M. (2012). Validity of timeline follow-back for selfreported use of cannabis and other illicit substanc-

es—Systematic review and meta-analysis. Addictive Behaviors, 37, 225–233. Hoskins, B. (2005). Roadside detection of drug-impaired driving. Paper presented at the WPMO 2005, 7th International Conference on Clinical Forensic Medicine of the World Police Medical Officers, Sydney, Australia Hser, Y., Maglione, M., & Boyle, K. (1999). Validity of self-report of drug use among STD patients, ER patients, and arrestees. American Journal of Drug and Alcohol Abuse, 25(1), 81–91. Humeniuk, R., Ali, R., Babor, T. F., Farrell, M., Formigoni, M. L., Jittiwutikarn, J., et al. (2008). Validation of the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST). Addiction, 103, 1039–1047. Humeniuk, R. E., Henry-Edwards, S., Ali, R. L., Poznyak, V., & Monteiro, M. (2010). The ASSIST linked brief intervention for hazardous and harmful substance use: Manual for use in primary care. Geneva: World Health Organization. Institute of Behavioral Research. (2007). TCU Drug Screen II (TCUDS II). Fort Worth: Texas Christian University, Institute of Behavioral Research. Jaffee, W. B., Trucco, E., Levy, S., & Weiss, R. D. (2007). Is this urine really negative?: A systematic review of tampering methods in urine drug screening and testing. Journal of Substance Abuse Treatment, 33, 33–42. Jones, A. W. (1993). Pharmaco*kinetics of ethanol in saliva: Comparison with blood and breath alcohol profiles, subjective feelings of intoxication and diminished performance. Clinical Chemistry, 39, 1837– 1844. Kaminer, Y. (2008). The Teen-Addiction Severity Index around the globe: The tower of Babel revisited. Substance Abuse, 29(3), 89–94. Kaminer, Y., Bukstein, O., & Tarter, R. E. (1991). The Teen-Addiction Severity Index: Rationale and reliability. International Journal of Addictions, 26, 219– 226. Kaminer, Y., Wagner, E., Plummer, B., & Seifer, R. (1993). Validation of the Teen-Addiction Severity Index (T-ASI): Preliminary findings. American Journal of Addictions, 2, 250–254. Kapur, B. (1994). Drug testing methods and interpretations of test results. In S. MacDonald & P. Roman (Eds.), Drug testing in the workplace (pp.103–120). New York: Plenum Press. Kenaszchuk, C., Wild, T. C., Rush, B. R., & Urbanoski, K. (2013). Rasch model of the GAIN Substance Problem Scale among Canadian adults seeking residential and outpatient addiction treatment. Addictive Behaviors, 38, 2279–2287. Kessler, R. C., Avenevoli, S., Green, J. G., Gruber, M. J., Guyer, M., He, Y., et al. (2009). The National Comorbidity Survey Replication Adolescent Supplement (NCS-A): III. Concordance of DSM-IV/CIDI diagnoses with clinical reassessments. Journal of American Academy of Child and Adolescent Psychiatry, 48(4), 386–399.

170

II .   D i a g n o s t ic I ss u e s

Kessler, R. C., & Ustun, T. B. (2004). The World Mental Health (WMH) Survey Initiative version of the World Health Organization (WHO) Composite International Diagnostic Interview (CIDI). International Journal of Methods in Psychiatric Research, 13(2), 93–121. Kintz, P., Villain, M., & Cirimele, V. (2006). Hair analysis for drug detection. Therapeutic of Drug Monitoring, 28(3), 442–446. Kirisci, L., Hsu, T. C., & Tarter, R. (1994). Fitting a two-parameter logistic item response model to clarify the psychometric properties of the Drug Use Screening Inventory for adolescent alcohol and drug abusers. Alcohol Clinical and Experimental Research, 18, 1335–1341. Knight, J. R., Sherritt, L., Kim, S., Gates, E. C., & Chang, G. (2003). Validity of brief alcohol screening tests among adolescents: A comparison of the AUDIT, POSIT, CAGE, and CRAFFT. Alcoholism: Clinical and Experimental Research, 27(1), 67–73. Knight, J. R., Sherritt, L., Shrier, L., Harris, S. K., & Chang, G. (2002). Validity of the CRAFFT substance abuse screening test among adolescent clinic patients. Archives of Pediatric and Adolescent Medicine, 156(6), 607–614. Knight, K., Hiller, M. L., Simpson, D. D., Broome, K. M. (1998). The validity of self-reported cocaine use in a criminal justice treatment sample. American Journal of Drug and Alcohol Abuse, 24(4), 647–660. Lapham, S. C., C’de Baca, J., Hunt, W. C., & Berger, L. R. (2001). Are drunk-driving offenders referred for screening accurately reporting their drug use? Drug and Alcohol Dependence, 66(3), 243–253. Laux, J. M., Piazza, N. J., Salyer, K., & Roseman, C. P. (2011). The Substance Abuse Subtle Screening Inventory–3 and stages of change: A screening validity study. Journal of Addictions and Offender Counseling, 33, 82–92. Lazowski, L. E., & Geary, B. B. (2016). Validation of the Adult Substance Abuse Subtle Screening Inventory–4 (SASSI-4). European Journal of Psychological Assessment. [Epub ahead of print] Lennox, C., O’Malley, K., Bell, V., Shaw, J., & Dolan, M. (2015). The validity and clinical utility of the Massachusetts Youth Screening Instrument-2 (MAYSI-2) in the UK. Criminal Behaviour and Mental Health, 25(3), 207–219. Liu, H. C., Liu, R. H., & Lin, D. L. (2015). Simultaneous quantitation of amphetamines and opiates in human hair by liquid chromatography—Tandem mass spectrometry. Journal of Analytical Toxicology, 39, 183–191. Lundin, A., Hallgren, M., Balliu, N., & Forsell, Y. (2015). The use of Alcohol Use Disorders Identification Test (AUDIT) in detecting alcohol use disorder and risk drinking in the general population: Validation of AUDIT using schedules for clinical assessment in neuropsychiatry. Alcoholism Clinical Experimental Research, 39(1), 158–165.

MacAndrew, C. (1965). The differentiation of male alcoholic outpatients from nonalcoholic psychiatric outpatients by means of the MMPI. Quarterly Studies of Alcohol, 47, 161–166. MacAndrew, C. (1981). What the MAC scale tells us about alcoholics: An interpretive review. Journal of Studies on Alcohol, 42, 604–625. MacAndrew, C. (1986). Toward the psychometric detection of substance misuse in young men: The SAP scale. Journal of Studies in Alcohol, 47, 161–166. Magura, S. (2010). Validating self-reports of illegal drug use to evaluate National Drug Control Policy: A reanalysis and critique. Evaluation and Program Planning, 33, 234–237. Martens, M. P., Neighbors, C., Dams-O’Connor, K., Lee, C. M., & Larimer, M. E. (2007). The factor structure of a dichotomously scored Rutgers Alcohol Problem Index. Journal of Studies on Alcohol and Drugs, 68(4), 597–606. Martin, C. S., Earleywine, M., Musty, R. E., Perrine, M. W., & Swift, R. M. (1993). Development and validation of the Biphasic Alcohol Effects Scale. Alcohol Clinical and Experimental Research, 17, 140–146. Martino, S., Grilo, C. M., Fehon, D. C. (2000). Development of the Drug Abuse Screening Test for Adolescents (DAST-A). Addictive Behaviors, 25(1), 57–70. McClelland, G. M., Elkington, K. S., Teplin, L. A., & Abram, K. M. (2004). Multiple substance use disorders in juvenile detainees. Journal of the American Academy of Child Adolescent Psychiatry, 43, 1215– 1224. McDonell, M, G., Comtois, K. A., Voss, W. D., Morgan, A. H., & Ries, R. K. (2009) Global Appraisal of Individual Needs Short Screener (GSS): Psychometric properties and performance as a screening measure in adolescents, The American Journal of Drug and Alcohol Abuse, 35(3), 157–160. McLellan, A. T., Cacciola, J. C., Alterman, A. I., Rikoon, S. H., & Carise, D. (2006). The Addiction Severity Index at 25: Origins, contributions and transitions. American Journal of Addiction, 15(2), 113–124. McLellan, A. T., Luborsky, L., Woody, G. E., & O’Brien, C. P. (1980). An improved diagnostic evaluation instrument for substance abuse patients: The Addiction Severity Index. Journal of Nervous and Mental Disease, 168, 26–33. Meldrum, R. C., & Piquero, A. R. (2015). On the potential implications of reports of fictitious drug use for survey research on juvenile delinquency. International Journal of Offender Therapy and Comparative Criminology, 59, 986–1005. Mieczkowski, T. (2002). Does ADAM need a haircut?: A pilot study of self-reported drug use and hair analysis in an arrestee sample. Journal of Drug Issues, 32(1), 97–118. Miller, C. S., Woodson, J., Howell, R. T., & Shields, A. L. (2009). Assessing the reliability of scores produced by the Substance Abuse Subtle Screening Inventory. Substance Use and Misuse, 44(8), 1090–1100.

8.  Denial and Misreporting of Substance Abuse 171

Miller, F. G., & Lazowski, L. E. (1999). The SASSI-3 Manual. Springville, IN: SASSI Institute. Miller, F. G., & Lazowski, L. E. (2001). The Adolescent SASSI-A2 manual: Identifying substance use disorders. Springville, IN: SASSI Institute. Miller, G. A. (1985). The Substance Abuse Subtle Screening Inventory—Revised (SASSI-R) manual. Bloomington, IN: SASSI Institute. Miller, G. A. (1990). The SASSI-Adolescent manual. Bloomington, IN: SASSI Institute. Miller, G. A. (1994). The Substance Abuse Subtle Screening Inventory (SASSI) manual. Bloomington, IN: SASSI Institute. Miller, W. R., Tonigan, J. S., & Longabaugh, R. (1995). The Drinker Inventory of Consequences (DrInC): An instrument for assessing adverse consequences of alcohol abuse (Project MATCH Monograph Series, NIH Publication No. 95-3911). Rockville, MD: U.S. Department of Health and Human Services, Public Health Service, National Institutes of Health, National Institute on Alcohol Abuse and Alcoholism. Millon, T. (1983). Millon Clinical Multiaxial Inventory. Minneapolis, MN: National Computer Systems. Millon, T. (1993). Millon Adolescent Clinical Inventory. Minneapolis, MN: National Computer Systems. Millon, T., Grossman, S., & Millon, C. (2015). Millon Clinical Multiaxial Inventory–IV (MCMI-IV) manual. Bloomington, MN: NCS Pearson. Moeller, K. E., Lee, K. C., & Kissack, J. C. (2008). Urine drug screening: Practical guide for clinicians. Mayo Clinical Procedures, 83(1), 66–76. Montesano, C., Johansen, S. S., & Nielsen, M. K. (2014). Validation of a method for the targeted analysis of 96 drugs in hair by UPLC-MS/MS. Journal of Pharmaceutical and Biomedical Analysis, 88, 295–306. Moore, R. A. (1972). The diagnosis of alcoholism in a psychiatric hospital: A trial of the Michigan Alcoholism Screening Test (MAST). American Journal of Psychiatry, 128, 1565–1569. Moosmann, B., Roth, N., & Auwarter, V. (2015). Finding cannabinoids in hair does not prove cannabis consumption. Scientific Reports, 5, Article 14906. Morgan, J. P. (1984). Problems of mass urine screening for misused drugs. Journal of Psychoactive Drugs, 16(4), 305–317. Musshoff, F., & Madea, B. (2006). Review of biologic matrices (urine, blood, hair) as indicators of recent or ongoing cannabis use. Therapeutic of Drug Monitoring, 28(2), 155–163. Myerholtz, L. E., & Rosenberg, H. (1998). Screening DUI offenders for alcohol problems: Psychometric assessment of the Substance Abuse Subtle Screening Inventory. Journal of Studies of Alcohol, 59, 439–446. National Institute on Alcohol Abuse and Alcoholism (NIAAA). (2015). Alcohol Screening and Brief Intervention for Youth: A practitioner’s guide (NIH Publication No. 11-7805). Rockville, MD: Author. Newcombe, D. A., Humeniuk, R. E., & Ali, R. (2005). Validation of the World Health Organization Alco-

hol, Smoking and Substance Involvement Screening Test (ASSIST): Report of results from the Australian site. Drug and Alcohol Review, 24, 217–226. Nochajski, T. H., & Wieczorek, W. F. (1998). Identifying potential drinking-driving recidivists: Do nonobvious indicators help? Journal of Prevention and Intervention in the Community, 17(1), 69–83. Oesterle, T. S., Hitschfeld, M. J., Lineberry, T. W., & Schneekloth, T. D. (2015). CRAFFT as a substance use screening instrument for adolescent psychiatry admissions. Journal of Psychiatry Practice, 21(4), 259–266. O’Farrell, T. J., & Langenbucher, J. (1988). Time-line drinking behavior interview. In M. Hersen & A. Bellack (Eds.), Dictionary of behavioral assessment techniques (pp.477–479). Elmsford, NY: Pergamon Press. Ortiz, N. R., & Spohn, C. (2014). Mitigating the effect of a criminal record at sentencing: Local life circ*mstances and substantial assistance departures among recidivists in federal court. Criminal Justice Policy Review, 25(1), 3–28. Otto, R. K., & Hall, J. E. (1988). The utility of the Michigan Alcoholism Screening Test in the detection of alcoholics and problems drinkers. Journal of Personality Assessment, 52, 499–505. Otto, R. K., Lang, A. R., Megargee, E. I., & Rosenblatt, A. I. (1988). Ability of alcoholics to escape detection by the MMPI. Journal of Consulting and Clinical Psychology, 56, 452–457. Palmer, R. F., Dwyer, J. H., & Semmer, N. (1994). A measurement model of adolescent smoking. Addictive Behaviors, 19, 477–489. Peters, R. H., LeVasseur, M. E., & Chandler, R. K. (2004). Correctional treatment for co-occurring disorders: Results of a national survey. Behavioral Sciences and the Law, 22(4), 563–584. Peters, R. H., Greenbaum, P. E., Steinberg, M. L., Carter, C. R., Ortiz, M. M., Fry, B. C., et al. (2000). Effectiveness of screening instruments in detecting substance use disorders among prisoners. Journal of Substance Abuse Treatment, 18, 349–358. Petzel, T., Johnson, J., & McKillip, J. (1973). Response bias in drug surveys. Journal of Consulting and Clinical Psychology, 40, 427–439. Pinto, M., & Grilo, C. M. (2004). Reliability, diagnostic efficiency, and validity of the Millon Adolescent Clinical Inventory: Examination of selected scales in psychiatrically hospitalized adolescents. Behaviour Research and Therapy, 42, 1505–1519. Pokorny, A. D., Miller, B. A., & Kaplan, H. B. (1972). The Brief MAST: A shortened version of the Michigan Alcoholism Screening Test. American Journal of Psychiatry, 129, 342–345. Pradhan, B., Chappuis, F., Baral, D., Karki, P., Rijal, P., Hadengue A., et al. (2012). The Alcohol Use Disorders Identification Test (AUDIT): Validation of a Nepali version for the detection of alcohol use disorders and hazardous drinking in medical settings. Substance Abuse Treatment, Prevention and Policy, 7, 42.

172

II .   D i a g n o s t i c I s s u e s

Pragst, F., & Balikova, M.A. (2006). State of the art in hair analysis for detection of drug and alcohol abuse. Clinica Chimica Acta, 370, 17–49. Reinert, D. F., & Allen, J. P. (2002). The Alcohol Use Disorders Identification Test (AUDIT): A review of recent research. Alcohol Clinical and Experimental Research, 26(2), 272–279. Richards, H. J., & Pai, S. M. (2003). Deception in prison assessment of substance abuse. Journal of Substance Abuse Treatment, 24(2), 121–128. Richman, J., & Noriega, R. S. (2002, December 12–15). The sensitivity and specificity of infrared pupillometry measurements in identifying drug impairment in a county probation program. Presened at the Scientific Program American Academy of Optometry, San Diego, CA. Robins, L. N., Heizer, J. F., Croughan, J., & Ratcliff, K. S. (1981). National Institute of Mental Health diagnostic interview schedule: Its history, characteristics, and validity. Archives of General Psychiatry, 38(4), 381–389. Robinson, S. M., Sobell, L. C., Sobell, M. B., & Leo, G. I. (2014). Reliability of the Timeline Followback for cocaine, cannabis, and cigarette use. Psychology of Addictive Behaviors, 25(1), 154–162. Rockett, I. R. H., Putnam, S. L., Jia, H., & Smith, G. S. (2006). Declared and undeclared substance use among emergency department patients: A population-based study. Addiction, 101(5), 706–712. Rogers, R. (2001). Handbook of diagnostic and structured interviewing. New York: Guilford Press. Rogers, R., & Mitchell, C. N. (1991). Mental health experts and the criminal courts: A handbook for lawyers and clinicians. Toronto: Carswell. Ross, H. E., Gavin, D. R., & Skinner, H. A. (1990). Diagnostic validity of the MAST and the Alcohol Dependence Scale in the assessment of DSM-III alcohol disorders. Journal of Studies on Alcohol, 51, 506–513. Saunders, J. B., Aasland, O. G., Babor, T. F., De La Fuente, J. R., & Grant, M. (1993). Development of the Alcohol Use Disorders Identification Test (AUDIT): WHO collaborative project on early detecting of persons with harmful alcohol consumption–II. Addiction, 88, 791–804. Selzer, M. L. (1971). Michigan Alcoholism Screening Test: The quest for a new diagnostic instrument. American Journal of Psychiatry, 127, 1653–1658. Selzer, M. L., Vinokur, A., & van Rooijen, M. A. (1975). A self-administered Short Michigan Alcohol Screening Test (SMAST). Journal of Studies on Alcohol, 36, 117–126. Shaffer, D., Fisher, P., Lucas, C. P., Dulcan, M. K., & Schwab-Stone, M. E. (2000). NIMH Diagnostic Interview Schedule for Children Version IV (NIMH DISC-IV): Description, differences from previous versions, and reliability of some common diagnoses. Journal of the American Academy of Child and Adolescent Psychiatry, 39, 28–38. Shields, A. L., & Caruso, J. C. (2003). Reliability gener-

alization of the Alcohol Use Disorders Identification Test. Educational and Psychological Measurement, 63(3), 404–413. Siegel, S. (2005). Drug tolerance, drug addiction, and drug anticipation. American Psychological Society, 14(6), 296–300. Simpson, G. (1987). Accuracy and precision of breathalcohol measurements for a random subject in postabsorptive state. Clinical Chemistry, 32, 261–268. Skinner, H. A. (1982). The Drug Abuse Screening Test. Addictive Behaviors, 7, 363–371. Sobell, M. B., Sobell, L. C., & Vanderspek, R. (1979). Relationships among clinical judgment, self-report, and breath-analysis measurements of intoxication in alcoholics. Journal of Consulting and Clinical Psychology, 47, 204–206. Stein, J. A., Smith, G. M., Guy, S. M., & Bentler, P. M. (1993). Consequences of adolescent drug use on young adult job behavior and job satisfaction. Journal of Applied Psychology, 78(3), 463–474. Stein, L. A. R. (2016, July). SBIRT for adolescents in juvenile justice and outpatient settings. Invited presentation at Improving Care of Substance Use, including Nicotine Use for Adolescents: Effective Approaches to Assessing, Treating and Engaging Teens, Part II (S. Becker, Chair). The New England Addiction Technology Transfer Center, SAMHSA, Dartmouth Medical Center, Brattleboro, VT. Stein, L. A. R., & Graham, J. R. (2001). Use of the MMPI-A to detect substance abuse in a juvenile correctional setting. Journal of Personality Assessment, 77, 508–523. Stein, L. A. R., Graham, J. R., Ben-Porath, Y. S., & McNulty, J. L. (1999). Using the MMPI-2 to detect substance abuse in an outpatient mental health setting. Psychological Assessment, 11, 94–100. Stinchfield, R., & Winters, K. C. (2003). Predicting adolescent drug abuse treatment outcome with the Personal Experience Inventory (PEI). Journal of Child and Adolescent Substance Abuse, 13(2), 103–120. Stucky, B. D., Orlando, M., & Ramchad, R. (2014). A psychometric assessment of the GAIN Individual Severity Scale (GAIN-GISS) and short screeners (GAIN-SS) among adolescents in outpatient treatment programs. Journal of Substance Abuse Treatment, 46,(2), 165–173. Substance Abuse and Mental Health Services Administration. (2012). Clinical Drug Testing in Primary Care (Technical Assistance Publication TAP 32, HHS ­Publication No. SMA 12-4668). Rockville, MD: Author. Substance Abuse and Mental Health Services Administration. (2014). Results from the 2013 National Survey on Drug Use and Health: Summary of National Findings (NSDUH Series H-48, HHS Publication No. SMA 14-4863). Rockville, MD: Author. Suwannachom, N., Thananchai, T., Junkuy, A., O’Biren, T. E., & Sribanditmongkol, P. (2015). Duration of detection of methamphetamine in hair

8.  Denial and Misreporting of Substance Abuse 173

after abstinence. Forensic Science International, 254, 80–86. Tahir, N. M., & Bakri, R. (2006). Eye scan for detection of recent drug abuse (Technology review). Putrajaya, Malaysia: Health Technology Assessment Unit, Medical Development Division, Ministry of Health. Tarter, R. E., & Hegedus, A. M. (1991). The Drug Use Screening Inventory: Its applications in evaluation and treatment of alcohol and other drug use. Alcohol Health and Research World, 15, 65–75. Tellegen, A., & Ben-Porath, Y. (2008). MMPI-2-RF technical manual. Minneapolis: University of Minnesota Press. Tenore, P. L. (2010). Advanced urine toxicology testing. Journal of Addictive Disease, 29(4), 436–448. Tiet, Q. Q., Finney, J. W., & Moos, R. H. (2008). Screening psychiatric patients for illicit drug use disorders and problems. Clinical Psychology Review, 28, 578–591. Tonigan, J. S., & Miller, W. R. (2002). The inventory of drug use consequences (InDUC): Test–retest stability and sensitivity to detect change. Psychology of Addictive Behaviors, 16(2), 165–168. Tourangeau, R., & Yan, T. (2007). Sensitive questions in surveys. Psychological Bulletin, 133(5), 859–883. Trichter, J. G., McKinney, W. T., & Pena, M. (1995). DWI demonstrative evidence: Show and tell the easy way. Voice, 24(3), 34–47. University of Pennsylvania VA Center for Studies of Addiction (UPVACSA). (1990). Addiction Severity Index, manual and question by question guide (Supported by grants from NIDA and VA). Philadelphia: Treatment Research Institute. Watson, I. D., Bertholf, R., Hammett-Stabler, C., Nicholes, B., Smith, B., George, S., et al. (2006). Drugs and ethanol. In J. H. Nichols (Ed.), Laboratory medicine practice guidelines: Evidence-based practice for point-ofcare testing (pp.63–75). Washington, DC: National Academy of Clinical Biochemistry. Weed, N. C., Butcher, J. N., McKenna, T., & Ben-Porath, Y. S. (1992). New measures for assessing alcohol and drug abuse with the MMPI-2: The APS and AAS. Journal of Personality Assessment, 58, 389–404. Weed, N. C., Butcher, J. N., & Williams, C. L. (1994). Development of MMPI-alcohol/drug problem scales. Journal of Studies on Alcohol, 55, 296–302. White, H. R., & Labouvie, E. W. (1989). Towards the assessment of adolescent problem drinking. Journal of Studies on Alcohol, 50, 30–37. Williams, R. J., & Nowatzki, N. (2005). Validity of adolescent self-report of substance use. Substance Use and Misuse, 40, 299–311. Winters, K. C. (1991). Personal Experience Screening Questionnaire (PESQ) manual. Los Angeles: Western Psychological Services. Winters, K. C. (1992). Development of an adolescent alcohol and other drug abuse screening scale: Personal Experiences Screening Questionnaire. Addictive Behaviors, 17, 479–490.

Winters, K. C., Anderson, N., Bengston, P., Stinchfield, R. D., & Latimer, W. W. (2000). Development of a parent questionnaire for use in assessing adolescent drug abuse. Journal of Psychoactive Drugs, 32(1), 3–13. Winters, K. C., & Henly, G. A. (1989). The Personal Experience Inventory (PEI) test and manual. Los Angeles: Western Psychological Services. Winters, K. C., & Kaminer, Y. (2008). Screening and assessing adolescent substance use disorders in clinical populations. Journal of American Academy Child Adolescent Psychiatry, 47(7), 740–744. Winters, K. C., Latimer, W. W., Stinchfield, R. D., & Henly, G. A. (1999). Examining psychosocial correlates of drug involvement among drug clinic-referred youth. Journal of Child and Adolescent Substance Abuse, 9(1), 1–17. Winters, K. C., Stinchfield, R. D., & Henly, G. A. (1993). Further validation of new scales measuring adolescent alcohol and other drug abuse. Journal of Studies on Alcohol, 54, 534–541. Winters, K. C., Stinchfield, R. D., Henly, G. A., & Schwartz, R. H. (1991). Validity of adolescent selfreport of alcohol and other drug involvement. International Journal of the Addictions, 25, 1379–1395. Winters, K. C., Stinchfield, R., & Latimer, W. W. (2004). Clinical assessment of adolescent drug abuse with the Personal Experience Inventory (PEI). In M. F. Maruish (Ed.), The use of psychological training for treatment planning and outcomes assessment: Vol. 2. Instruments for children and adolescents (3rd ed., pp.371–409). Mahwah, NJ: Erlbaum. Woodland, M. H., Andretta, J. R., Moore, J. A., Bennett, M. T., Worrell, F. C., & Barnes, M. E. (2014). MACI scores of African American males in a forensic setting: Are we measuring what we think we are measuring? Journal of Forensic Psychology Practice, 5, 418–437. Wooley, C. N., Rogers, R., Fiduccia, C. E., & Kelsey, K. (2012). The effectiveness of substance use measures in the detection of full and partial denial of drug use. Assessment, 20(6), 670–680. Xiang, P., Shen, M., & Drummer, O. H. (2015). Review: Drug concentrations in hair and their relevance in drug facilitated crimes. Journal of Forensic Legal Medicine, 36, 126–135. Ying, Y., Wu, C., & Chang, K. (2013). The effectiveness of drinking and driving policies for different alcoholrelated fatalities: A quantile regression analysis. International Journal of Environmental Research and Public Health, 10, 4628–4644. Yudko, E., Lozhkina, O., & Fouts, A. (2007). A comprehensive review of the psychometric properties of the Drug Abuse Screening Test. Journal of Substance Abuse Treatment, 32(2), 189–198. Zaldivar, F., Molina, A. M., Rios, F., & Montes, J. M. (2009). Evaluation of alcohol and other drug use and the influence of social desirability: Direct and camouflaged measures. European Journal of Psychological Assessment, 25(4), 244–251.

CHAPTER 9

Psychopathy andDeception NathanD.Gillard,PhD

Deception has remained a key characteristic of prototypical psychopathy throughout history. Cleckley’s (1941, 1976) influential description of psychopathy included “untruthfulness and insincerity,” while Hare (1985) has included “pathological lying” on the original Psychopathy Checklist (PCL) and Psychopathy Checklist— Revised (PCL-R; Hare, 1991, 2003). Descriptions of psychopathy also include various core features that require deceptive practices, including superficial charm, manipulativeness, and shallowness. Psychopaths are also described as being naturally exploitative of others, with particularly strong skills at conning. Modern conceptualizations of psychopathy often include antisocial and criminal behaviors, which often require deception during the crime itself and the subsequent avoidance of being apprehended. Beyond the fabrications and general deceit of ordinary criminals, psychopaths, almost by definition, use conscious distortions and manipulations across multiple domains of their lives, leaving no relationship unaffected (Cleckley, 1976). This chapter focuses on both the theoretical and actual use of deception in psychopathy. Despite the clear conceptual connection noted by influential authors such as Kraepelin (1915), Schneider (1923), Cleckley (1941, 1976), and Hare (1991, 2003), very few studies have examined the frequency with which psychopaths use deception or their success at doing so. Perhaps more surprising,

the existing studies indicate that psychopaths are not actually any more successful at being deceptive than nonpsychopaths, even though they may do so more frequently (Clark, 1997; Lykken, 1978; Patrick & Iacono, 1989; Raskin & Hare, 1978). This chapter considers psychopathy in relation to malingering, positive impression management, and general deception. The effect of psychopathic traits on the validity of interview-based and selfreport clinical and risk assessment measures is also addressed, with some initial findings on the effectiveness of existing validity scales. Finally, areas for future research are discussed.

EARLY CONCEPTUALIZATIONS OFPSYCHOPATHY Starting in the late 19th and early 20th century, the term psychopathy emerged as representing a “group of persons showing abnormality expressed mainly in the character and intensity of the emotional reactions” (Partridge, 1930, p.63). This overly inclusive clinical description was commonplace for the time and Partridge pointed out that at least 13 separate terms emerged to describe overlapping conceptualizations of psychopathy. Perhaps the most influential early definitions were those of Kraepelin (1915) and Schneider (1923). Kraepelin (1915) defined psychopathy mostly in behavioral terms, similar to the modern conceptu-

174

9. Psychopathy and Deception 175

alization of antisocial personality disorder (American Psychiatric Association, 2013). Accordingly, those with psychopathic traits continuously violated societal rules and the rights of others. They used aggression and deceit frequently and were often destructive. In contrast, Schneider (1923) attributed psychopathy to emotional deficits and characterological disorder; he noted that such individuals either suffer personally or make others suffer due to their abnormal personalities. He believed psychopathic individuals displayed blunted affect and an inability to experience inhibitory emotions.1 Writing near the end of his career, he bluntly stated that they “lack capacity for shame, decency, remorse, and conscience. They are ungracious, cold, surly, and brutal in crime” (Schneider, 1950/1958, p.126). These contrasting definitions present a conceptual debate that, to some extent, has continued until modern day.

DECEPTION INTHEDEFINITION OFPSYCHOPATHY For many years, the conclusion that “everybody lies” was widely accepted, primarily due to studies that indicated noncriminal participants told an average of two lies per day (DePaulo, Kashy, Kirkendol, Wyer, & Epstein, 1996; see also Vrij, 2000). More recently, challenges to this conclusion arose due to wide individual differences in the frequency of lying. Serota, Levine, and Boster (2010) found that the often replicated mean of “two lies per day” was skewed by a small number of frequent liars. Social psychology studies have examined situations that increase or decrease the willingness to deceive, whereas research in clinical psychology has examined the characteristics of frequent liars. While a discussion of all situational and characterological factors is beyond the scope of this chapter, studies (Kashy & DePaulo, 1996; Halevy, Shalvi, & Verschuere, 2014) have documented psychopathic traits as being particularly salient to deception. Deception represents a common occurrence in many different settings, especially when the incentives are high (Frank & Ekman, 2004). Nowhere is this observation more true than in forensic settings. Likewise, psychopathy has been assumed to be a particularly influential factor in the frequency and success of deception. Despite the large independent literatures on psychopathy and deception, few studies have investigated the association between these two concepts. The following sections

review the theoretical and empirical evidence that does exist. As previously noted, deception plays a pivotal role in defining the core features of psychopathy. Not only have leading theorists and researchers (Cleckley, 1976; Hare, 2003) listed deception directly as a defining feature of psychopathy, but it can also be considered a core component of many other psychopathy features. For example, Cleckley (1976, p.338) listed “untruthfulness and insincerity” as one of 16 core components of psychopathy. Also included in this list are descriptions of superficial charm, unreliability, and “inadequately motivated antisocial behavior” (p.338). Each of these characteristics requires purposeful deception. Core features of a Cleckley-defined psychopath stem from broad emotional deficits. Like Schneider (1958), Cleckley (1976) understood psychopathy to consist of primarily innate deficits in emotions, while criminal behavior was a secondary product of these deficits. Cleckley described psychopaths as having a remarkable disregard for truth when recollecting the past, speaking of the current situation, and making promises about the future. They are “at ease and unpretentious” (p.342) when making promises or lying and can be especially convincing in their portrayal. According to Cleckley, the usual signs that others notice in a clever liar are not evident, which means that the psychopath’s lies are often undetected. Cleckley believed the very concept of honesty has only intellectual meaning to the psychopath, and is devoid of the positive emotion that most nonpsychopaths have attached to honesty and trustworthiness. To Cleckley, a male psychopath would “keep his word” only if doing so was to his benefit. In his description of the psychopath’s antisocial behavior, Cleckley (1976) further expounded on circ*mstances in which deceit may occur, noting a willingness to lie for “astonishingly small stakes.” Over the last three decades, Robert Hare has pioneered the modern conceptualization of psychopathy. Hare (1991, 2003) covered some of the same basic concepts as Cleckley, though he used different terms and provided somewhat dissimilar descriptions in his PCL-R scoring. First, “pathological lying” is the PCL-R item most directly related to deception. It describes a likelihood of characteristic lying and deceit when interacting with others. The psychopath’s readiness and ease of lying are described as being “remarkable,” and if caught, he or she is not anxious but simply changes the story as needed (Hare, 2003, p.37).

176

II .   D i a g n o s t ic I ss u e s

Unlike Cleckley, Hare believed psychopaths might freely discuss and even brag about their lies. Hare also included a description of “conning/manipulative” behavior, which shares elements with the previously discussed PCL-R item and Cleckley’s items on untruthfulness and superficial charm. In particular, this item measures deceit used to cheat or manipulate others, whether in extravagant or simple ways. Like Cleckley, Hare describes a willingness to lie in many different circ*mstances (see Table 9.1). Aside from the interpersonal and affective uses of deceit, Hare (2003) believed psychopathy included antisocial behavioral traits to a greater extent than did Cleckley (1976). These behaviors do not always directly involve lies, but some deceit is necessary to successfully carry out most crimes. Although subsequent models of psychopathy now exist (Harpur, Hakstian, & Hare, 1988; Cooke & Michie, 2001; Hare, 2003), each retains deception as a key element. Forensic experts, community members, and offenders alike recognize the importance of deception to psychopathy, though they differ in their assessment of the centrality of this characteristic. Rogers and colleagues (Rogers, Dion, & Lynett, 1992; Rogers, Duncan, Lynett, & Sewell, 1994; Rogers, Salekin, Sewell, & Cruise, 2000) conducted three studies on the prototypical features of psychopathy and DSM-III antisocial personality disorder (ASPD). Groups of community volunteers, forensic experts, and adult offenders all recognized the importance of deceit in these disorders, but important differences also existed. Community volunteers rated “no regard for the truth” and “pathological lying” as highly prototypical descriptors of adult ASPD, with only “lack of remorse” and “unlawful behavior” ranked higher (see Rogers et al., 1992, Table 1, p.683). In comparison, forensic experts rated “no regard for the truth” even higher than did laypersons (Rogers et al., 1994, Table 1, p.478). In sharp contrast, adult offenders ranked untruthfulness as less important and aggression as more important than did experts and members of the community. In both childhood and adulthood, adult offenders placed relatively little emphasis on interpersonal factors as compared to experts and community members. As one possible reason for this difference, inmates may lack insight into the importance of deceit in relation to antisocial behavior more generally. Then again, offenders may see deception as being so common that it does not differentiate psychopaths from nonpsychopaths.

FREQUENCY OFMALINGERING BYPSYCHOPATHS Similar to general deception, there is a relative dearth of studies assessing psychopathy and malingering. Existing studies indicate little or no association between the two (e.g., Edens, Buffington, & Tomicic, 2000; Poythress, Edens, & Lilienfeld, 1998; Poythress, Edens, & Watkins, 2001). For instance, Gacono, Meloy, Sheppard, Speth, and Roske (1995) attempted to examine the presence of psychopathic traits in a small sample of insanity acquittals. They found those with higher levels of psychopathic traits were more likely to admit to previous malingering. They suggested this relationship may indicate that psychopaths have a higher willingness to use deception or a greater proficiency at it. However, this study was flawed in several ways, including (1) the use of self-reported malingering and retrospective data, and (2) providing the incentive to admit feigning to facilitate being released from the hospital (Rogers & Cruise, 2000). Additionally, the study utilized only patients with highly violent and antisocial histories, increasing the likelihood of high PCL-R scores. Kucharski, Duncan, Egan, and Falkenbach (2006) found mixed evidence that criminal defendants with psychopathic traits were able to feign while avoiding detection. While scores on the Minnesota Multiphasic Personality Inventory–2 (MMPI-2) validity scales were generally high to extremely high (means ranging from 83.73 to 95.41 for the F family of scales), scores on PAI indices exhibited less extreme elevations. Furthermore, the high psychopathy group did not differ significantly from the low and moderate psychopathy groups on two of three PAI feigning scales (Rogers Discriminant Function [RDF] and Malingering Index [MAL]). In a known-groups comparison with the Structured Interview of Reported Symptoms (SIRS; Rogers, Bagby, & Dickens, 1992) as the external criterion, it classified roughly twothirds of those in the high psychopath group as feigning, while close to one-fifth of the moderate group (21.6%) was detected. Rogers (1990) has argued that the idea that malingering occurs more frequently in psychopathic and antisocial individuals is a methodological artifact. Most malingering studies are conducted in criminal forensic settings. Similarly, most psychopathy studies are also conducted in correctional or forensic settings. Therefore, both concepts occur at increased rates in the same setting (Rog-

9. Psychopathy and Deception 177

TABLE 9.1.  Comparison of Hare Facets and Corresponding Cleckley Traits

Hare (1991)

Cleckley (1976) Facet 1: Interpersonal

  1. Glibness/superficial charm

  1. Superficial charm and good “intelligence”

  2. Grandiose self-worth

  9. Pathological egocentricity

  4. Pathological lying **

  5. Untruthfulness and insincerity **

  5. Conning/manipulative ** Facet 2: Affective   6. Lack of remorse or guilt

  6. Lack of remorse or shame

  7. Shallow affect

10. General poverty in major affective reactions

  8. Callous/lack of empathy

  9. Incapacity for love

16. Failure to accept responsibility for own actions *

  4. Unreliability * Facet 3: Lifestyle

  3. Need for stimulation/proneness to boredom

No Cleckley equivalent

  9. Parasitic lifestyle

No Cleckley equivalent

13. Lack of realistic, long-term goals

16. Failure to follow any life plan

14. Impulsivity (no mention of alcohol)

13. Fantastic and uninviting behavior with drink and sometimes without

15. Irresponsibility *

  4. Unreliability * Facet 4: Antisocial

10. Poor behavioral controls

13. Fantastic and uninviting behavior with drink andsometimes without

12. Early behavioral problems

No Cleckley equivalent

18. Juvenile delinquency *

No Cleckley equivalent

19. Revocation of conditional release

No Cleckley equivalent

20. Criminal versatility

No Cleckley equivalent Unloaded items

11. Promiscuous sexual behavior

15. Sex life impersonal, trivial, and poorly integrated *

17. Many short-term marital relationships

No Cleckley equivalent No Hare equivalent   2. Absence of delusions and other signs of irrational thinking   3. Absence of “nervousness” or psychoneurotic manifestations   7. Inadequately motivated antisocial behavior   8. Poor judgment and failure to learn by experience 11. Specific loss of insight 14. Suicide rarely carried out

Note. * denotes items associated with deception; ** denotes items directly tapping deception. From Gillard (2013, p.22). Reprinted with permission.

178

II .   D i a g n o s t ic I ss u e s

ers, 1990). While few studies exist to resolve this question directly, Rogers found similar portions of ASPD-diagnosed individuals in groups of malingerers (20.8%) and genuine patients (17.7%). Rogers and Robinson (2016) recognized that the problematic relationship between psychopathy and malingering may be troubling to practicing clinicians who must decide whether the former should be used as a determining factor in decisions about the latter. They suggest clinicians weigh the substantial pitfalls of utilizing psychopathy in this regard against the potential benefit. Therefore, the presence of psychopathy might be used as a screen to indicate that increased scrutiny for malingering is needed but it should never be used as evidence of increased likelihood (Rogers & Robinson, 2016). Looked at another way, ASPD and psychopathy may be seen as common factors rather than distinguishing factors in those who malinger.

FREQUENCY OFDECEPTION BYPSYCHOPATHS Conceptual descriptions indicate psychopaths lie frequently, but an interesting question involves psychopaths’ actual (i.e., empirically tested) likelihood of lying compared to nonpsychopaths placed in similar circ*mstances. Many of the early “studies” of deception in psychopaths were based on clinical case studies and anecdotal reports (e.g., Cleckley, 1941; Hare, 1991). The limited number of empirical studies to date has yielded inconsistent findings on the likelihood that psychopaths lie in a variety of contexts. For self-reports of deception, those with psychopathic traits tend to report higher rates of lying, with at least three studies lending support. Kashy and DePaulo (1996) used a daily diary paradigm (i.e., a daily log of social interactions and the lies told during them) and found such traits were correlated with lying in a normal population.2 Similarly, Halevy et al. (2014) found that in a community sample, psychopathic traits as measured by the Youth Psychopathic Traits Inventory (YPI) were positively correlated with the number of lies told during the 24-hour period preceding the study (r = .31). Seto, Khattar, Lalumiere, and Quinsey (1997) found a moderate connection between PCL-R scores and sexual (r = .36) and nonsexual (.49) deceptive tactics. However, the significance of sexual deceptive practices was erased when accounting for general deceptiveness, suggesting the occurrence of deception is similar across domains.

Observers raise salient concerns about the generalizability of these self-report studies. Given the grandiosity often seen in psychopaths, is it possible they would brag about deception rather than reporting their actual deceptive practices? More generally, should we assume that psychopaths are being honest about their dishonesty, either by raising or lowering its reported frequency? Seto and colleagues (1997) found a moderately large negative correlation between impression management and psychopathy, which suggests these offenders were not concerned with presenting themselves in a positive light, though this does not necessarily equate to honesty. However, their finding may also indicate that those high on impression management suppressed their psychopathy scores. Seeking to fill the void in the literature, Halevy et al. (2014) conducted a second study, offering participants from their earlier self-report study an opportunity to cheat for financial gain. Specifically, they instructed participants to roll dice a number of times, with earnings determined by the number rolled. The number of lies reported in the initial study was correlated with the amount they earned in the dice rolling task (r = .39) when there was zero chance of being detected. The finding was interpreted as evidence that self-reported lies were a relatively accurate measure of the frequency and willingness to lie. As one of the only real-world examinations of lying in psychopaths, Porter, Birt, and Boer (2001) reviewed the correctional records of psychopathic and nonpsychopathic murderers. They found those scoring higher on psychopathic traits were twice as likely to change the details of their version of the crime during incarceration. Cooper and Yuille (2007) hypothesized the primary use of laboratory studies is responsible for the disconnection between empirical findings and the classic description of psychopaths as deceptive. In such studies, participants with psychopathy are research volunteers with poor intrinsic motivation to lie. As Feeley and deTurck (1998) pointed out, laboratory studies use “sanctioned lies” (i.e., experimental instructions asking participants to lie). Sanctioned lies are quite different, cognitively and affectively, from unsanctioned lies. Psychopaths may have no need to lie in laboratory studies, thereby making it appear they lie less frequently than is found in field studies or anecdotal reports. As discussed earlier, clinical case studies that led to the core characteristics of psychopathy indicate frequent use of lying (Cleckley, 1941, 1976; Hare, 1991, 2003). On this point, Cooper and

9. Psychopathy and Deception 179

Yuille (2007) suggest that although anecdotal evidence is not ideal, it may offer superior evidence to laboratory studies regarding psychopaths’ lying. Regardless of the frequency of deception, classic psychopathic descriptions also include successful deception (i.e., lying without detection), which will now be discussed.

BEYOND LIKELIHOOD: SKILLEDDECEPTION ANDPSYCHOPATHICTRAITS Descriptions of psychopaths suggest they are not only more likely to engage in deceptive and manipulative practices but also more skilled at doing so (Hare, 2003). The first part of this assumption has garnered some support based on self-report and laboratory tasks. Regarding psychopaths’ actual success, or skill, at deception, the literature has again been heavy on theory and anecdotal evidence and low on empirical findings. Until recently, the few available studies have focused on malingering. However, new and exciting general deception studies have recently emerged. Despite inconsistent results, a general trend is observed toward poor performance at deception by psychopathic individuals.

Avoiding theDetection ofFeigning Researchers have found psychopaths appear to be either no more or only marginally more successful at malingering than nonpsychopaths. For example, Kropp (1994) found psychopaths were no better than other inmates at simulating mental illness in an analogue design study using the SIRS (Rogers et al., 1992). While psychopaths were generally no better at malingering, Kropp (1994) noted a disproportionately high number of psychopaths in the small group of successful feigners that eluded detection on the SIRS. At least two additional studies have supported the early findings of Kropp (1994). First, Edens et al. (2000) found no connection between psychopathic traits and undergraduates’ ability to feign psychosis and psychopathy on the MMPI-2 and Psychopathic Personality Inventory (PPI; Lilienfeld & Andrews, 1996). Specifically, psychopathic traits were not related to the level of feigned symptoms on the two measures or the ability to avoid detection on the Deviant Responding (DR) scale of the PPI, as the DR scale was generally effective at detecting feigning across groups (area under the

curve [AUC] = .98). Using male prison inmates, Poythress, Edens, and Watkins (2001) similarly reported no significant correlations between PPI scores and three feigning measures: the Structured Inventory of Malingered Symptomatology (SIMS; Widows & Smith, 2005), the SIRS (Rogers et al., 1992), and the three commonly used feigning indicators (Positive Impression Management [PIM], MAL, and RDF) on the PAI (Morey, 1991). Methodologically, none of these studies used known-groups design (see Rogers, Chapter 1, this volume) with independently classified malingerers. Therefore, it is possible that psychopaths’ apparent lack of skill for feigning reflects their limited motivation when engaged in laboratory studies. Alternatively, malingering is quite different than other forms of deception and requires skills that may neutralize any advantage psychopaths otherwise have at lying. For instance, a basic understanding of detection strategies and how to avoid elevating them has been found to be helpful to successful malingerers (Rogers, Dolmetsch, & Cavanaugh, 1983). Psychopaths’ use of charm and grandiosity may be useful when deceiving others in interpersonal situations but not be beneficial when dealing with such structured methods of detection. As an example, the PPI DR scale was significantly more effective than the corresponding PIM scale (Unlikely Virtues scale; see Edens Buffington, Tomicic, & Riley, 2001).

Avoiding Detection byPolygraph Two independent research groups have found psychopaths’ deception can be detected by polygraph tests just as effectively as they detect nonpsychopaths. For the first group, Raskin and Hare (1978) conducted a mock crime study with psychopathic and nonpsychopathic offenders. They instructed all participants, guilty and innocent, to deny the crime. Both denial-of-guilt and genuine groups were detected at similarly high rates (> 90% correct classifications). Patrick and Iacono’s (1989) results agreed for the denial-of-guilt group but not the genuine group. Specifically, psychopaths and nonpsychopaths were detected at exactly the same percentage (i.e., 91.7%) but an unacceptably large percentage of genuine responders were misclassified as deniers (41.7%). Their study differed from that of Raskin and Hare (1978) in that a threat (i.e., a list of poor performers would be posted for all to see) was used as the motivation instead of a reward. Both studies generally agree with the

180

II .   D i a g n o s t ic I ss u e s

aforementioned research from malingering studies indicating psychopaths are no more successful than nonpsychopaths when structured or standardized detection methods are used.

Avoiding Detection byHumanJudges Classic descriptions of psychopathy describe deception occurring in a wide range of situations, even when there are no obvious advantages. Cleckley (1976), in particular, thought lies by psychopaths were told without the typical signs of lying, making psychpaths more difficult to identify. However, research suggests psychopaths’ skills are not necessarily so clear-cut. Two studies (Cogburn, 1993; Billings, 2004) have used naive judges (i.e., untrained college students) to rate the believability of offender statements. In one of the earliest and more cited studies, Cogburn (1993) found psychopaths, as classified on the PCL-R, exhibited no improvements in “successful” deception when compared to nonpsychopaths. Furthermore, naive judges rated psychopaths as less credible regardless of the actual veracity of their statements. This finding contradicts Cleckley’s (1976) assertion regarding psychopaths’ particular adeptness at appearing honest and avoiding typical signs exhibited by liars. In contrast to Cogburn (1993), Billings (2004) found that individuals with higher psychopathy scores on both the Psychopathy Checklist: Screening Version (PCL:SV; Hart, Cox, & Hare, 1995) and PPI were better able to deceive naive judges when making false statements. Klaver, Lee, Spidel, and Hart (2009) bolstered Cogburn’s (1993) earlier conclusions by studying raters’ observations of psychopaths’ storytelling behavior. Specifically, the undergraduate participants viewed video recordings of psychopathic and nonpsychopathic offenders telling true and false stories. They then rated the credibility of the statements and the nonverbal signs of deception, such as appearing to think hard and nervousness. Psychopaths were not successful at lying. In fact, even worse than in Cogburn (1993), psychopaths’ deceptions were detected more often than those of nonpsychopaths. Moreover, like Cogburn, psychopaths were generally viewed as less credible. In direct contrast, the same researchers (Lee, Klaver, & Hart, 2008) found psychopaths were more successful at avoiding detection when only verbal indicators were rated. Specifically, psychopaths used more appropriate details when lying, suggesting a focused attempt to appear credible. Still, the psy-

chopathic group was far more likely to be judged as noncredible when actually telling the truth. The body of research suggests psychopaths are somewhat proficient at lying in a verbal manner but are unable to successfully control their nonverbal behaviors. This disparity shows their apparent adeptness at using verbal skills, while many behavioral indicators are neglected by research. Given the scripted, analogue design of these studies, an alternative hypothesis is that real-world lying would be easier to accomplish for psychopaths. Additionally, the aforementioned studies suggest psychopaths are often less able to appear credible when telling the truth compared to nonpsychopaths. Although empirically untested, limited emotional expression may account for this observation under the honest condition, while this same characteristic may be beneficial when telling lies. Thus, clinicians must use caution when evaluating the verbal clues of psychopaths that may be useful for detecting deception in the general public. The interpersonal facet of psychopathy in particular may facilitate deception. Lee et al. (2008) found Factor 1 was associated with credibility when discussing false information, while Factor 2 was not. Thus, an arrogant and grandiose behavioral presentation may distract the listener from lies that, contentwise, are not particularly sophisticated. Theoretically, for undetected lying, it would seem advantageous to appear confident and avoid displaying affective arousal related to the deception. On this matter, it has been suggested that psychopaths have very little emotional investment in the words they use. For instance, Williamson, Harpur, and Hare (1991) presented affective and neutral words and nonwords to offenders. Nonpsychopaths were able to distinguish emotional words from nonwords very efficiently; the authors theorized that the affective content facilitated their cognitive processes. In contrast, psychopaths failed to show this same pattern. The lack of emotional investment by psychopaths may free them to deceive without regard for the affective significance of the topic. As Pegay (1943, as cited by Hare, 1991) stated, psychopaths can talk about the deepest of topics and “[pull] the words from their overcoat pocket” (p.124). Both behavioral and verbal changes occur when psychopaths are placed in experimental situations requiring deception. Regarding verbal behaviors exhibited while lying, Louth, Williamson, Alpert, Pouget, and Hare (1998) found differences in the volume of psychopaths’ voices when compared to

9. Psychopathy and Deception 181

nonpsychopaths. When discussing deceptive neutral, negative, and positive topics, psychopathic inmates spoke more quietly than other inmates. While not examining the volume of speech, Klaver, Lee, and Hart (2007) found differences in nonverbal behavior while speaking in psychopathic and nonpsychopathic offenders. Besides speaking faster, psychopathic offenders engaged in more blinking and head movements when lying. For male adolescents with behavior problems, Rimé, Bouvy, Leborgne, and Rouillon (1978) concluded that psychopaths appeared more invested during the interview than nonpsychopaths (i.e., they leaned forward more and looked at the interviewer longer). Unfortunately, it is not possible to conclude whether psychopaths intentionally chose these verbal and nonverbal changes.

DECEPTION ONRISK ASSESSMENT ANDPSYCHOPATHYMEASURES The assessment of psychopathy frequently occurs as part of forensic evaluations, especially those involving future violent risk. Psychopathy has consistently been found to be one of the strongest single predictors of violent and general risk (Steadman, 2000), and has been included as part of multiple risk assessment measures. They include structured professional judgment (SPJ) guides and actuarial measures. The Historic Clinical Risk–20 (HCR20; Webster, Douglas, Eaves, & Hart, 1997a) and Violence Risk Assessment Guide (VRAG; Harris, Rice, & Quinsey, 1993) even include PCL-R and/ or PCL:SV scores as part of their appraisal. Three large meta-analyses produced encouraging results for actuarial measures when compared to the stand-alone PCL-R. Gendreau, Goggin, and Smith (2002), Yang, Wong, and Coid (2010), and Singh, Grann, and Fazel (2011) found roughly equivalent results for the most frequently used general (i.e., nonsexual) violent risk measures. Singh et al. found slightly higher predictive validity for the Structured Assessment of Violence Risk in Youth (SAVRY; Borum, Bartel, & Forth, 2006) and lower predictive validity for the Level of Service Inventory (LSI; Andrews & Bonta, 1995) and the PCL-R. These meta-analyses continue the long line of studies showing little to no difference among well-validated, published risk assessment measures. Examiners should be careful to assess how well their population matches validation samples, as the latter meta-analysis found higher pre-

dictive validity when the study sample was similar to the initial validation sample.

Intentional Minimization onRisk AssessmentMeasures There appears to be an implicit assumption held by many clinicians that risk assessment procedures are not greatly affected by deception. However, Gillard and Rogers (2015; summarized later in this chapter), as well as general research on the ease of impression management, strongly question this tacit assumption. Most structured clinical judgments rely on both interviews and records, which have been assumed to protect against falsification (Webster, Douglas, Eaves, & Hart, 1997b). While collateral sources are always encouraged, these records are often unreliable due to missing information and a lack of truly objective information (i.e., records based on previous interviews are similarly susceptible to manipulation). As an illustration, the number and type of previous criminal convictions can now be fairly reliably gleaned from searchable databases due to advances in technology. However, records regarding mental health history, substance abuse, interpersonal relationships, and treatment noncompliance (all included on many risk assessment measures) are often painfully incomplete due to a lack of centralized records and the fact that such records only exist if treatment is sought. When records do exist, practitioners should use caution in assessing their reliability; many of these records are simply based on earlier interviews with the same offender. Reports completed for the court, such as presentence investigations, often contain information primarily based on collateral interviews with family members regarding the offender’s background, education, and personality. These records are often accepted as fact, but they should be seen as susceptible to the same biases as any direct interview. These cautions should not dissuade the use of records, but they should encourage practitioners to ensure the accuracy of clinical data found in past records. Three studies have been conducted using offenders to examine intentional minimization on interview-based and self-report measures, with mixed results. The Self-Appraisal Questionnaire (SAQ; Loza, 2005), a self-report risk measure, was utilized in all three studies. Loza, Loza-Fanous, and Heseltine (2007) compared responses under two conditions: (1) presumably genuine (i.e., confiden-

182

II .   D i a g n o s t ic I ss u e s

tiality-guaranteed instructions) and (2) presumably intentional minimization (i.e., psychological evaluations being used to determine early release). Contrary to expectations, the scores in the real evaluation group were slightly higher on scales measuring substance use and past criminal conduct. However, the results are unreliable due to the use of differential prevalence design, a weak design that assumes motivation may be presumed by the referral issue (see Rogers, Chapter 1, this volume). A second study by the same authors (Loza et al., 2007) used a “simulated” differential prevalence design, with participants first receiving “research only” instructions, then a second condition (i.e., simulated “release evaluation”). High correlations between conditions might indicate that the SAQ is resistant to minimization. However, their results are difficult to interpret due to the previously noted differential prevalence design. Furthermore, the simulation component could not be verified given the absence of a manipulation check. Thus, neither study directly addresses the vulnerability of the SAQ to intentional minimization. Recently, Gillard and Rogers (2015) examined the effects of PIM on both interview-based (HCR-20) and self-report (SAQ) risk assessment measures, as well as the Psychological Inventory of Criminal Thinking Styles (PICTS; Walters, 2001), a measure of offenders’ cognitive styles. Using a jail sample, all measures were administered in a repeated-measures simulation design. Offenders were able to substantially lower both HCR-20 and SAQ scores utilizing PIM. Greater suppression was made to the Historical subscale than to Clinical and Risk items (d = 1.32 vs. 1.06 and 0.60, respectively). Likely due to the ease of changing historical information, as well as their transparency, offenders may easily deny items such as previous violence. Scores on the SAQ were modified in a proportionally similar manner, with effect sizes between 0.96 and 1.34 for all subscales except Associates, a measure of interactions with other antisocial individuals. The presence of psychopathic traits, especially Factor 1 characteristics, was associated with greater change during PIM on the interview-based HCR-20 (Gillard & Rogers, 2015). Interpersonal and affective traits alone were found to predict approximately one-third of the variability in HCR-20 Total scores (R2 = .34) even after Factor 2 traits were accounted for using hierarchical regression. These findings did not hold for the SAQ. Similarly, the PICTS, also a self-report measure, did not exhibit strong changes during the impression

management stage, regardless of the presence of psychopathic traits. As a possible explanation, affective and interpersonal traits were more strongly related to deception during interview than selfreport due to the interaction with the examiner.

Impression Management onMeasures ofPsychopathy The PCL and PCL-R have been considered the “gold standard” of psychopathy assessment for over two decades (Patrick, 2006). Based on both conceptual grounds and unresolved concerns about deception, researchers such as Ray et al. (2013) have recommended self-report measures not be used in clinical or forensic practice as a proxy for the PCL-R. However, self-report measures of psychopathy have increased in number and popularity over the last 20 years. Such scales provide an alternative approach to the PCL-R and may be advantageous in the following ways: • The allowance of group administration that can be completed in time-limited situations. • An increased sensitivity to changes as a result of time, treatment, and other interventions (Edens et al., 2001) that cannot be accomplished with the relatively trait-based PCL-R. • Elimination of the heavy reliance on collateral data sources. As previously discussed, collateral reports may provide the illusion of reliability but are susceptible to many of the same pitfalls as direct interviews. Despite self-reports’ being prone to manipulation by responders, they have received surprisingly little research attention. On a positive note, the solutions to safeguard self-reports against impression management (i.e., validity scales utilizing well-constructed detection strategies) may be relatively easy to incorporate compared to the solutions for detecting deception during interviews (i.e., training interviewers to spot verbal deception and controlling for unreliable collateral data). To explore the susceptibility of self-report measures to deception, Edens et al. (2001) asked undergraduates to positively manage scores on the PPI and Marlowe–Crowne Social Desirability Scale (MCSDS; Crowne & Marlowe, 1960). The researchers found that regardless of psychopathic traits and the simulation scenario, simulators were able to lower PPI scores (d range from 0.37 to 0.48). Those with higher psychopathic traits had more significant decreases in their PPI scores (mean d

9. Psychopathy and Deception 183

= 1.22 vs. 0.17). The PPI Unlikely Virtues validity scale was unable to achieve high levels of detection with a false-positive rate of 25%. Readers should note, however, that interpretation of the Edens et al. (2001) results is complicated by the use of a college sample, with the “high” psychopathy group obtaining an honest mean score (M = 396), which is comparable to a youthful prison population (M = 392; Poythress et al., 1998). As a reevaluation of Edens et al. (2001), Edens (2004) found that only certain traits were lowered during PIM. Research has found that the PPI can be separated into two higher-order factors. PPI-I consists of the Social Potency, Fearlessness, and Stress Immunity subscales, while PPI-II consists of the Blame Externalization, Impulsive Nonconformity, Carefree Nonplanfulness, and Machiavellian Egocentricity subscales. The reanalysis found that PPI-II items were lowered significantly more during PIM (d = –0.86), while PPI-I scores actually increased (d = 0.53). As discussed by Edens, these results make conceptual sense, as some PPI-I traits may seem advantageous. Unfortunately, Edens did not evaluate which genuinely occurring psychopathic traits appear to aid PIM the most. Kelsey (2014) provided the only known examination of PIM using a self-report questionnaire with an offender population. As part of a repeatedmeasures simulation design, participants reduced their PPI-R scores well below community levels. Predictably, the finding was most pronounced for those with high psychopathy. Thus, while those with genuinely high levels of psychopathy are able to lower scores to “normal” levels, they are often unable to elude detection. Kelsey (2014) compared PPI-R results to the Self-Report Psychopathy (SRP-4) scale and Levenson Self-Report Psychopathy Scale (LSRP) and found that both were also vulnerable to PIM. Kelsey (2014) also provided initial results for the detection of minimized psychopathy traits. As an initial screen, the PPI-R Virtuous Responding (VR) scale can potentially identify individuals engaging in PIM (cutoff score of > 40) and those with likely genuine responding not engaging in PIM (cutoff score of < 25). These scores currently leave a substantial group of examinees who would require further evaluation. Further work on the PPI-R VR scale cutoff scores is needed before it can be widely used in correctional/forensic settings. Given the vulnerability of self-report psychopathy measures to PIM, Rogers et al. (2002) evaluated whether an interview-based PCL measure

showed similar vulnerability. Using the PCL-YV in a simulation design, adolescent offenders were able to lower scores by 44.2%. Without a record review, the generalizability of these results to realworld forensic evaluations is limited. However, as noted throughout this chapter, whether such a review would decrease the effects of PIM has yet to be studied, likely due to the pervasive belief that such records protect the integrity of results.

Impression Management onClinicalMeasures The potential for psychopaths to manipulate assessment results can extend beyond psychopathy and risk to general clinical measures. Unfortunately, only two published studies have examined this important issue. Book, Holden, Starzyk, Wasylkiw, and Edwards (2006) studied undergraduates who successfully portrayed socially desirable personality traits (i.e., were not detected by the validity scale cutoff scores) on the Holden Psychological Screening Inventory (HPSI; Holden, 1996). Successful deceivers had modestly higher primary, secondary, and total psychopathy scores on the LSRP compared to those detected as faking (Cohen’s d of 0.46, 0.36, and 0.50, respectively). In line with previously reported studies, the positive advantage for psychopathy disappeared when participants were asked to malinger “serious psychological problems.” MacNeil and Holden (2006) also analyzed successful PIM on the HPSI. While initial results did not find psychopathy to be advantageous, three particular PPI scales were most relevant for those who avoided detection: Machiavellian Egocentricity, Blame Externalization, and Stress Immunity. Conceptually, these scales—measuring emotional coldness, aggressiveness, and a willingness to engage in selfish acts—could aid in the ability to deceive. Their connection to successfully avoiding formal detection strategies is less obvious. For example, while emotional coldness has been found to aid verbal lying, it does not assist on written multiple-choice assessment measure.

CONCLUSIONS The inclusion of deception in the description of those with psychopathic personality predates the term psychopathy itself. Perhaps because of the centricity of this connection, little research has been devoted to psychopaths’ frequency of and skill at deception. With the increased popularity of selfreport psychopathy and risk assessment measures,

184

II .   D i a g n o s t ic I ss u e s

research on this connection becomes even more imperative. Decisions based on these measures can have far-reaching clinical and forensic consequences. For example, classification as a psychopath is often associated with poor treatment outcomes and sometimes leads to the unwarranted assumption that other symptoms are disingenuous. The significant ramifications of risk assessment findings include extended detention and loss of freedom (or conversely, the avoidance of certain consequences). Thus, the effect that psychopathic traits have on deception, and on the manipulation of assessment results in particular, is a very high priority. Research conducted so far has consistently found the following: 1.  Little, if any, connection between psychopathy and the likelihood of feigning. The most acceptable explanation for malingering is that it is adaptational in nature: It occurs when an individual has (a) substantial investment in (b) an adversarial setting and (c) sees few alternatives (Rogers & Cavanaugh, 1983). Thus, while both malingering and psychopathy occur more commonly in forensic settings, the connection is an artifact, and neither should be used as evidence of the other. 2.  No increased skill at malingering for psychopaths. Avoiding modern detection strategies for feigned mental disorder and cognitive impairment is a skill that requires specific knowledge, unlikely to be aided by psychopathic traits. 3.  Increased likelihood of general deception. This constitutes the most direct support for Cleckley’s (1976) description of “untruthfulness and insincerity” and Hare’s (1985) descriptions of “pathological lying.” In both self-report and experimental designs, psychopathic traits are associated with more frequent deception. However, this likelihood should not be equated with increased skill in avoiding detection. 4.  No increased ability to avoid detection on polygraph tests for psychopaths. Although neurological and autonomic variables are related to psychopathy, current findings indicate no increased ability to avoid polygraph detection. Research into other aspects of the deception– psychopathy connection is less conclusive. The following issues clearly need further investigation: •• The type and effectiveness of verbal, nonverbal, and interpersonal strategies psychopaths use to de-

ceive. Results are mixed but suggest greater fluency with verbal strategies. However, multiple studies found those with higher psychopathic traits were rated as less credible when telling stories. Of particular interest is the finding that psychopaths might be incorrectly perceived to be lying when telling the truth. •• Psychopaths’ skill at using PIM on interviewbased and self-report measures. Initial results indicate psychopaths are able to dramatically decrease scores on psychopathy, risk assessment, and at least one general clinical measure to a greater extent that nonpsychopaths. More research is also needed to establish whether psychopaths lower traits to a believable level that would meet their intended goal without being caught by validity scales. As an initial finding, Kelsey (2014) found that those with high psychopathy traits were no better at avoiding detection that those with lower traits. •• Development of validity scales on risk- and psychopathy-specific measures. Additional research is also needed on effective validity scales included on self-report risk and psychopathy measures. Currently, those that do exist (e.g., Unlikely Virtues on the PPI) often misclassify genuine responders at an unacceptably high rate (i.e., false-positive rate of 25%).

NOTES 1.  Schneider used a wide definition for psychopathy more akin to the modern term personality disorder. Of his 10 typologies, the affectionless psychopath is most similar to the modern conceptualization that is the topic of this chapter (Crowhurst & Coles, 1989). 2. While Kashy and DePaulo (1996) do not reference psychopathy specifically, they measure a number of traits common to psychopaths, including Machiavellianism and social adroitness.

REFERENCES American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: Author. Andrews, D. A., & Bonta, J. (1995). The Level of Service Inventory—Revised (LSI-R). Toronto: Multi-Health Systems. Billings, F. (2004). Psychopathy and the ability to deceive. Dissertation Abstracts International, 65, 3-B. Book, A. S., Holden, R. R., Starzyk, K. B., Wasylkiw, L., & Edwards, M. J. (2006). Psychopathic traits and experimentally induced deception in self-report assess-

9. Psychopathy and Deception 185

ment. Personality and Individual Differences, 41(4), 601–608. Borum, R., Bartel, P., & Forth, A. (2006). Manual for the Structured Assessment for Violence Risk in Youth (SAVRY). Odessa, FL: Psychological Assessment Resources. Clark, C. R. (1997). Sociopathy, malingering, and defensiveness. In R. Rogers (Ed.), Clinical assessment of malingering and deception (2nd ed., pp.68–84). New York: Guilford Press. Cleckley, H. (1941). The mask of sanity: An attempt to reinterpret the so-called psychopathic personality. Oxford, UK: Mosby. Cleckley, H. (1976). The mask of sanity (5th ed.). St. Louis, MO: Mosby. Cogburn, R. (1993). A study of psychopathy and its relation to success in interpersonal deception. Dissertation Abstracts International, 54, 2191B. Cooke, D. J., & Michie, C. (2001). Refining the construct of psychopath: Towards a hierarchical model. Psychological Assessment, 13(2), 171–188. Cooper, B. S., & Yuille, J. C. (2007). Psychopathy and deception. In H. Hervé & J. C. Yuille (Eds.), The Psychopath: Theory, Research, and Practice (pp.487– 503). Mahwah, NJ: Erlbaum. Crowhurst, B., & Coles, E. M. (1989). Kurt Schneider’s concepts of psychopathy and schizophrenia: A review of the English literature. Canadian Journal of Psychiatry, 34(3), 238–243. Crowne, D. P., & Marlowe, D. (1960). A new scale of social desirability independent of psychopathology. Journal of Consulting Psychology, 24, 349–354. DePaulo, B. M., Kashy, D. A., Kirkendol, S. E., Wyer, M. M., & Epstein, J. A. (1996). Lying in everyday life. Journal of Personality and Social Psychology, 70(5), 979–995. Edens, J. F. (2004). Effect of response distortion on the assessment of divergent facets of psychopathy. Assessment, 11, 109–112. Edens, J. F., Buffington, J. K., & Tomicic, T. L. (2000). An investigation of the relationship between psychopathic traits and malingering on the Psychopathic Personality Inventory. Assessment, 7(3), 281–296. Edens, J. F., Buffington, J. K., Tomicic, T. L., & Riley, B. D. (2001). Effects of positive impression management on the Psychopathic Personality Inventory. Law and Human Behavior, 25(3), 235–256. Feeley, T. H., & deTurck, M. A. (1998). The behavioral correlates of sanctioned and unsanctioned deceptive communication. Journal of Nonverbal Behavior, 22(3), 189–204. Frank, M. G., & Ekman, P. (2004). Nonverbal detection of deception in forensic contexts. In W. T. O’Donohue & E. R. Levensky (Eds.), Handbook of forensic psychology: Resource for mental health and legal professionals (pp.635–653). New York: Elsevier Science. Gacono, C. B., Meloy, J., Sheppard, K., Speth, E., & Roske, A. (1995). A clinical investigation of malin-

gering and psychopathy in hospitalized insanity acquittees. Bulletin of the American Academy of Psychiatry and the Law, 23(3), 387–397. Gendreau, P., Goggin, C., & Smith, P. (2002). Is the PCL-R really the “unparalleled” measure of offender risk?: A lesson in knowledge accumulation. Criminal Justice and Behavior, 29(4), 397–426. Gillard, N. D. (2013). Denial of risk: The effects of intentional minimization on risk assessments for psychopathic and nonpsychopathic offenders. Unpublished doctoral dissertation, University of North Texas, Denton, TX. Gillard, N. D., & Rogers, R. (2015). Denial of risk: The effects of positive impression management on risk assessments for psychopathic and nonpsychopathic offenders. International Journal of Law and Psychiatry, 42–43, 106–113. Halevy, R., Shalvi, S., & Verschuere, B. (2014). Being honest about dishonesty: Correlating self-reports and actual lying. Human Communication Research, 40(1), 54–72. Hare, R. D. (1985). The Psychopathy Checklist. Unpublished manuscript, University of British Columbia, Vancouver, BC, Canada. Hare, R. D. (1991). The Hare Psychopathy Checklist—Revised. Toronto: Multi-Health Systems. Hare, R. D. (2003). The Hare Psychopathy Checklist— Revised, 2nd Edition. Toronto: Multi-Health Systems. Harpur, T. J., Hakstian, A., & Hare, R. D. (1988). Factor structure of the Psychopathy Checklist. Journal of Consulting and Clinical Psychology, 56(5), 741–747. Harris, G. T., Rice, M. E., & Quinsey, V. L. (1993). Violent recidivism of mentally disordered offenders: The development of a statistical prediction instrument. Criminal Justice and Behavior, 20(4), 315–335. Hart, S. D., Cox, D. N., & Hare, R. D. (1995). Psychopathy Checklist: Screening Version. Toronto: MultiHealth Systems. Holden, R. R. (1996). Holden Psychological Screening Inventory (HPSI). Toronto: Multi-Health Systems. Kashy, D. A., & DePaulo, B. M. (1996). Who lies? Journal of Personality and Social Psychology, 70(5), 1037– 1051. Kelsey, K. R. (2014). The vulnerability of self-report measures of psychopathy to positive impression management: A simulation study with inmates. Unpublished doctoral dissertation, University of North Texas, Denton, TX. Klaver, J. R., Lee, Z., & Hart, S. D. (2007). Psychopathy and nonverbal indicators of deception in offenders. Law and Human Behavior, 31(4), 337–351. Klaver, J. R., Lee, Z., Spidel, A., & Hart, S. D. (2009). Psychopathy and deception detection using indirect measures. Legal and Criminological Psychology, 14(1), 171–182. Kraepelin, E. (1915). Psychiatrie: Ein Lehrbuch [Psychiatry: A textbook] (8th ed., Vol. 4). Leipzig, Germany: Barth. Kropp, P. (1994). The relationship between psychopathy and malingering of mental illness. Unpublished doc-

186

II .   D i a g n o s t ic I ss u e s

toral dissertation, Simon Fraser University, Burnaby, BC, Canada. Kucharski, L., Duncan, S., Egan, S. S., & Falkenbach, D. M. (2006). Psychopathy and malingering of psychiatric disorder in criminal defendants. Behavioral Sciences and the Law, 24(5), 633–644. Lee, Z., Klaver, J. R., & Hart, S. D. (2008). Psychopathy and verbal indicators of deception in offenders. Psychology, Crime and Law, 14(1), 73–84. Lilienfeld, S. O., & Andrews, B. P. (1996). Development and preliminary validation of a self-report measure of psychopathic personality traits in noncriminal populations. Journal of Personality Assessment, 66(3), 488–524. Louth, S. M., Williamson, S., Alpert, M., Pouget, E. R., & Hare, R. D. (1998). Acoustic distinctions in the speech of male psychopaths. Journal of Psycholinguistic Research, 27(3), 375–384. Loza, W. (2005). Self-Appraisal Questionnaire (SAQ): A tool for assessing violent and nonviolent recidivism. Toronto: Multi-Health Systems. Loza, W., Loza-Fanous, A., & Heseltine, K. (2007). The myth of offenders’ deception on self-report measure predicting recidivism: Example from the Self-Appraisal Questionnaire (SAQ). Journal of Interpersonal Violence, 22(6), 671–683. Lykken, D. T. (1978). The psychopath and the lie detector. Psychophysiology, 15(2), 137–142. MacNeil, B., & Holden, R. (2006). Psychopathy and the detection of faking on self-report inventories of personality. Personality and Individual Differences, 41(4), 641–651. Morey, L. C. (1991). The Personality Assessment Inventory professional manual. Odessa, FL: Psychological Assessment Resources. Partridge, G. E. (1930). Current conceptualizations of psychopathic personality. American Journal of Psychiatry, 10, 53–99. Patrick, C. J. (2006). Back to the future: Cleckley as a guide to the next generation of psychopathy research. In C. J. Patrick (Ed.), Handbook of psychopathy (pp.605–617). New York: Guilford Press. Patrick, C. J., & Iacono, W. G. (1989). Psychopathy, threat, and polygraph test accuracy. Journal of Applied Psychology, 74(2), 347–355. Porter, S., Birt, A., & Boer, D. P. (2001). Investigation of the criminal and conditional release profiles of Canadian federal offenders as a function of psy­ chopathy and age. Law and Human Behavior, 25(6), 647–661. Poythress, N. G., Edens, J. F., & Lilienfeld, S. O. (1998). Criterion-related validity of the Psychopathic Personality Inventory in a prison sample. Psychological Assessment, 10(4), 426–430. Poythress, N. G., Edens, J. F., & Watkins, M. (2001). The relationship between psychopathic personality features and malingering symptoms of major mental illness. Law and Human Behavior, 25(6), 567–582. Raskin, D. C., & Hare, R. D. (1978). Psychopathy and

detection of deception in a prison population. Psychophysiology, 15(2), 126–136. Ray, J. V., Hall, J., Rivera-Hudson, N., Poythress, N. G., Lilienfeld, S. O., & Morano, M. (2013). The relation between self-reported psychopathic traits and distorted response styles: A meta-analytic review. Personality Disorders, 4(1), 1–14. Rimé, B., Bouvy, H., Leborgne, B., & Rouillon, F. (1978). Psychopathy and nonverbal behavior in an interpersonal situation. Journal of Abnormal Psychology, 87(6), 636–643. Rogers, R. (1990). Models of feigned mental illness. Professional Psychology: Research and Practice, 21(3), 182–188. Rogers, R., Bagby, R. M., & Dickens, S. E. (1992). Structured Interview of Reported Symptoms professional manual. Odessa, FL: Psychological Assessment Resources. Rogers, R., & Cavanaugh, J. L. (1983). “Nothing but the truth”... a reexamination of malingering. Journal of Psychiatry and Law, 11(4), 443–459. Rogers, R., & Cruise, K. R. (2000). Malingering and deception among psychopaths. In C. B. Gacono (Ed.), The clinical and forensic assessment of psychopathy: A practitioner’s guide (pp.269–284). Mahwah, NJ: Erlbaum. Rogers, R., Dion, K. L., & Lynett, E. (1992). Diagnostic validity of antisocial personality disorder: A prototypical analysis. Law and Human Behavior, 16(6), 677–689. Rogers, R., Dolmetsch, R., & Cavanaugh, J. L. (1983). Identification of random responders on MMPI protocols. Journal of Personality Assessment, 47(4), 364– 368. Rogers, R., Duncan, J. C., Lynett, E., & Sewell, K. W. (1994). Prototypical analysis of antisocial personality disorder: DSM-IV and beyond. Law and Human Behavior, 18(4), 471–484. Rogers, R., & Robinson, E. V. (2016). Psychopathy and response styles. In C. B. Gacono (Ed.), The clinical and forensic assessment of psychopathy: A practitioner’s guide (2nd ed., pp.217–230). New York: Routledge/ Taylor & Francis Group. Rogers, R., Salekin, R. T., Sewell, K. W., & Cruise, K. R. (2000). Prototypical analysis of antisocial personality disorder: A study of inmate samples. Criminal Justice and Behavior, 27(2), 234–255. Rogers, R., Vitacco, M. J., Jackson, R. L., Martin, M., Collins, M., & Sewell, K. W. (2002). Faking psychopathy?: An examination of response styles with antisocial youth. Journal of Personality Assessment, 78(1), 31–46. Schneider, K. (1923). Die psychopathischen personlichkeiten [The psychopathic personalities] (M. W. Hamilton, Trans.). Vienna: Deuticke. Schneider, K. (1958). Psychopathic personalities (M. W. Hamilton, Trans.). London: Cassell & Company. (Original work published 1950) Serota, K. B., Levine, T. R., & Boster, F. J. (2010). The

9. Psychopathy and Deception 187

prevalence of lying in America: Three studies of selfreported lies. Human Communication Research, 36(1), 2–25. Seto, M. C., Khattar, N. A., Lalumière, M. L., & Quinsey, V. L. (1997). Deception and sexual strategy in psychopathy. Personality and Individual Differences, 22(3), 301–307. Singh, J. P., Grann, M., & Fazel, S. (2011). A comparative study of violence risk assessment tools: A systematic review and meta-regression analysis of 68 studies involving 25,980 participants. Clinical Psychology Review, 31(3), 499–513. Steadman, H. J. (2000). From dangerousness to risk assessment of community violence: Taking stock at the turn of the century. Journal of the American Academy of Psychiatry and the Law, 28(3), 265–271. Vrij, A. (2000). Detecting lies and deceit. Chichester, UK: Wiley. Walters, G. D. (2001). The Psychological Inventory of Criminal Thinking Styles (PICTS) professional manual. Allentown, PA: Center for Lifestyle Studies.

Webster, C. D., Douglas, K. S., Eaves, D., & Hart, S. D. (1997a). HCR-20: Assessing risk for violence (Version 2). Burnaby, British Columbia, Canada: Simon Fraser University, Mental Health, Law, and Policy Institute. Webster, C. D., Douglas, K. S., Eaves, D., & Hart, S. D. (1997b). Assessing risk of violence to others. In C. D. Webster & M. A. Jackson (Eds.), Impulsivity: Theory, assessment, and treatment (pp.251–277). New York: Guilford Press. Widows, M., & Smith, G. P. (2005). Structured Inventory of Malingered Symptomatology (SIMS) and professional manual. Odessa, FL: Psychological Assessment Resources. Williamson, S., Harpur, T. J., & Hare, R. D. (1991). Abnormal processing of affective words by psychopaths. Psychophysiology, 28(3), 260–273. Yang, M., Wong, S. P., & Coid, J. (2010). The efficacy of violence prediction: A meta-analytic comparison of nine risk assessment tools. Psychological Bulletin, 136(5), 740–767.

CH A P TER 10

The Malingering ofPosttraumatic Disorders PhillipJ.Resnick,MD SaraG.West,MD ChelseaN.Wooley,PhD

EXPLORING THEDISORDERS The History ofPosttraumatic StressDisorder Posttraumatic stress disorder (PTSD) is a relatively new term to psychiatry; however, the concept has existed for over 100 years. In the 1880s, terms such as nervous shock and posttraumatic neurosis were coined to describe the psychological phenomena resulting from exposure to trauma (Adamou & Hale, 2003; Hausotter, 1996; Sparr, 1990). In 1889, Dr. Clevenger proposed the idea that similar diagnoses, including “railroad spine” and “compensation neurosis,” were related to an accidental concussion resulting in abnormalities in the central nervous system (Hall & Chapman, 2005; Thomann & Rauschmann, 2003, 2004). In the 20th century’s many wars, the psychological sequelae of battle were explored. During World War I, this reaction to trauma was called “shell shock,” and in World War II, the term battle fatigue was used. The first DSM used the diagnosis, gross stress reaction (American Psychiatric Association, 1952) and DSM-II (American Psychiatric Association, 1968) used adjustment reaction to adult life. Societal awareness intensified with the return of the veterans from the Vietnam War. The term posttraumatic stress disorder was introduced with the publication of DSM-III (American Psychiatric Association, 1980).

DSM-IV (American Psychiatric Association, 1994) led to an alteration in the criteria used to diagnose PTSD. Criterion A, which describes the traumatic event, was changed from the DSM-III’s objective standard (an event that would be markedly distressing to almost anyone) to a subjective standard (an event that the victim found personally distressing). This broadening of the definition led to a 39% increase in the number of individuals who met diagnostic criteria for PTSD (Breslau & Kessler, 2001). DSM-IV was also responsible for the introduction of the diagnosis “acute stress reaction,” which is a time-limited precursor to PTSD involving dissociative symptoms (American Psychiatric Association, 1994). In 2013, the American Psychiatric Association made major revisions in the criteria for PTSD with the publication of DSM-5. The first notable change was the movement of PTSD from the Anxiety Disorders category to the new category of Trauma- and Stress-Related Disorders in DSM5. Other diagnoses in this new category include reactive attachment disorder, disinhibited social engagement disorder, acute stress disorder, and adjustment disorder. The common link shared by these conditions is that they are all precipitated by stress (Friedman, Resick, Bryant, & Brewin, 2011). DSM-5’s Criterion A redefined the precipitating stressor required for the diagnosis of PTSD. Sexu-

188

10. The Malingering of Posttraumatic Disorders 189

al violence replaced a vague item about physical integrity. DSM-5 specified actual or threatened death of a loved one, and the event must have been violent or accidental. Finally, DSM-IV’s A2 criterion requiring extreme emotional response was eliminated, because certain groups, such as military personnel or those with a traumatic brain injury, may experience these responses differently yet still develop PTSD (Friedman et al., 2011). The largest change associated with Criterion B (intrusion) is that the dreams are now described as being reflective of the content or the affect related to the trauma. Levin, Kleinman, and Adler (2014) posited that this change is consistent with Res­ nick’s (2004) observation that recurrent dreams with invariable content may serve as evidence for malingering. In DSM-IV, Criterion C consisted of both avoidance and numbing symptoms. These two symptom clusters have been divided into Criteria C and D in DSM-5. Individuals must have at least one symptom in each of these categories to qualify for the diagnosis. Criterion D has also been reworked to include more specific alterations in cognition and mood, in addition to numbing. Criterion E (hyperarousal) was expanded to include reckless or self-destructive behavior, as well as aggression. It reflects the findings of Calhoun et al. (2012), who studied 185 patients with PTSD (one-third of whom were veterans). They demonstrated that 58% of the sample engaged in aggressive behavior and 17% were reckless. Finally, the subtype “with dissociative symptoms” was added to DSM-5, while the acute and chronic specifiers in DSM-IV were eliminated. These significant changes to the diagnosis generated multiple publications speculating on the effects that such alterations would have. McNally (2009) reported that the allowance of additional indirect means of exposures to a trauma would create “a bracket creep, expanding those eligible for the diagnosis.” In 2009, Elhai, Ford, Ruggiero, and Frueh suggested that, though Criterion C would be divided into two more specific symptom clusters, the more stringent requirements would have little impact on the prevalence of the disorder. Following the release of DSM-5, a study of 2,953 participants, Kilpatrick et al. (2013) demonstrated a 4% drop in those who met Criterion A from 93.7% in DSM-IV to 89.7% in DSM-5, as well as a 1.2% decrease in the lifetime prevalence from DSM-IV to DSM-5 criteria. The authors attributed this decrease to the elimination of the indirect exposure to the nonviolent death of a loved one from Criterion A and the failure to have one of

two avoidance symptoms defined by Criterion C in DSM-5. Breslau, Davis, Andreski, and Peterson (1991) described five risk factors for exposure to traumatic events: low education, male gender, early conduct problems, extraversion, and a family history of psychiatric disorders or substance problems. Since only a minority of individuals develops PTSD in response to a trauma, Davidson (1993) identified 11 pretrauma characteristics that cause an individual to be more vulnerable to PTSD (see Table 10.1); these are linked to backgrounds, childhood antecedents, and prior psychiatric issues. A metaanalysis (N = 476) indicated that peritraumatic dissociation is the single best predictor for the development of PTSD (Ozer, Best, Lipsey, & Weiss, 2003). Other factors included prior trauma, prior psychological maladjustment, family history of psychopathology, perceived life threat during the trauma, lack of posttraumatic social support, and peritraumatic emotional responses. After the diagnosis of PTSD was introduced in 1980, the psychiatric community began investigating the disorder in a group of people universally recognized to have been exposed to significant atrocities, specifically, Holocaust survivors. Kuch and Cox (1992) identified 124 Jewish Holocaust survivors deemed to be free of major psychiatric illnesses; 46% met DSM-III criteria for PTSD (American Psychiatric Association, 1980). Thus, although all these survivors experienced significant trauma, less than half of them had symptoms TABLE 10.1.  Pretrauma Characteristics That Increase the Likelihood of Developing PTSD

A. Background 1. Female gender 2. History of psychiatric illness in first-degree relatives B. Childhood antecedents 1. Parental poverty 2. Separation or divorce of parents before the age of10 3. Trauma in childhood (may be of a sexual nature) 4. Behavioral disorder in childhood or adolescence 5. Poor self-confidence in adolescence C. Prior psychiatric problems 1. Prior psychiatric disorders 2. Introversion 3. Life stress prior to the trauma 4. High neuroticism Note. From Davidson (1993).

190

II .   D i a g n o s t ic I ss u e s

of PTSD at the time of evaluation. Further research on the children of Holocaust survivors indicated that they did not experience more traumatic events than a demographically similar comparison group. However, they still had a greater prevalence of current and lifetime PTSD (Yehuda, Schmeidler, Wainberg, Binder-Brynes, & Duvdevani, 1998). It suggests that a biological component predisposes people to the development of PTSD. In recent years, the neurobiology of PTSD has been more clearly delineated. Dysfunction in both the hippocampus and the amygdala has been linked to PTSD. Increased stress is believed to lead to increased systemic glucocorticoids, which in turn interfere with the biochemical process of long-term potentiation in the hippocampus (McEwen, 1997). This process is hypothesized to be the origin of the memory disturbance noted in PTSD. A number of studies compared the size of hippocampi in those with PTSD to normal controls and the results have been varied (Grossman, Buchsbaum, & Yehuda, 2002). The amygdala is associated with processing fear and the accompanying autonomic responses. It appears that an overactive amygdala may play a role in the inability to extinguish classical fear conditioning and the hyperarousal symptoms that are salient features of PTSD. Diagnosing PTSD may prove difficult due to the high degree of comorbid psychopathology. For instance, in a large study of treatment-seeking outpatients with a primary diagnosis of PTSD, Brown, Campbell, Lehman, Grisham, and Mancill (2001) noted that 92% had another active psychiatric diagnosis. For the clinician, the challenge lies in determining the comorbidities’ presence or absence prior to the trauma, their course following the trauma, and their contribution to the patient’s current symptoms.

Malingering Malingering is defined by the American Psychiatric Association (2013, pp.726–727) as “the intentional production of false or grossly exaggerated physical or psychological symptoms motivated by external incentives.” This external gain may take the form of financial rewards, relief of responsibilities at a job or at home, avoiding military service, or evading criminal responsibility (American Psychiatric Association, 2013). This definition did not change with the publication of DSM-5. Malingering may be divided into three categories that depict the nature of the symptoms. Pure

malingering is the conscious feigning of a disorder that does not exist at all. Partial malingering is the fraudulent, conscious exaggeration of present symptoms or the allegation that prior genuine symptoms are still present. False imputation refers to the individual’s intentional attribution of actual symptoms to a different cause. For example, individuals who are aware that they developed PTSD due to a prior trauma may falsely ascribe the symptoms to a more recent car accident in order to gain financial compensation. However, it is important recognize that some individuals fail to recognize that consecutive events do not necessarily have a causal relationship (Collie, 1971), and this failure must be differentiated from malingering. Partial malingering is difficult to identify because the individual can, from personal experience, accurately describe the symptoms (Wooley & Rogers, 2015). Differential diagnosis for malingering must also consider factitious disorder and conversion disorder (Table 10.2). Factitious disorders are similar to malingering in that the diagnosis requires the conscious production of false symptoms. They differ, however, in the motivation for the symptom production. Unlike malingering, those with factitious disorder intentionally produce symptoms “even in the absence of obvious external rewards” (American Psychiatric Association, 2013, p.324). What is notably absent in DSM-5 is reference to the sick role, which served as a criterion in DSM-IV (American Psychiatric Association, 2000). Conversion disorder differs from malingering in that the individual is unaware of the origin of his symptoms. In contrast to individuals who malinger, those with factitious disorder and conversion disorder are ill and may be eligible for compensation. In a forensic setting, malingering should be suspected in all examinees, since external rewards

TABLE 10.2.  A Comparison of Malingering, Factitious Disorder, and Conversion Disorder

Intentional production of symptoms

External reward

Awareness of purpose

Malingering

Yes

Yes

Yes

Factitious disorder

Yes

No

No

Conversion disorder

No

No

No

10.  The Malingering of Posttraumatic Disorders 191

are almost always available. Individuals deemed to be disabled may in fact demonstrate no disability even shortly after the determination has been made. While some cases were legitimate, others likely reflected a feigned initial impairment. In a survey of clinicians, Mittenberg, Patton, Canyock, and Condit (2002) found that symptom exaggeration or malingering occurred in about 30% of personal injury and disability cases. Evaluators should be aware that disreputable attorneys may coach their clients about psychological symptoms following trauma, especially in the case of a major accident with multiple plaintiffs (Rosen, 1995). Legally, malingering constitutes fraud, but convictions for perjury requires the trier of fact be convinced beyond a reasonable doubt that a conscious decision to lie under oath occurred. There is frequently not enough evidence to support a conviction. Often, the trier of fact will simply not grant an award to a plaintiff believed to be malingering. Clinicians are hesitant to classify an individual as malingering for several reasons (Burges & McMillan, 2001). First, a wide range of diagnoses must be ruled out prior to classifying someone as a malingerer (Pollack, 1982). Second, a false accusation of malingering may lead to stigmatization and subsequent inability to receive appropriate care (Kropp & Rogers, 1993). Third, the clinician may fear litigation or even physical assault due to labeling someone as a liar. One of us (P. J. R.) has been involved in three cases of clinicians who were sued for defamation of character due to insufficiently supported labels of malingering. It is easier to defend the term feigning, because there is no obligation to prove that the conduct is motivated by an external incentive (see Rogers, Chapter 1, this volume, for the distinction between feigning and malingering). The public’s hostility toward suspected malingerers is understandable given that their undeserved financial gain would be associated with another’s undeserved financial loss (Braverman, 1978). This link may be one of the reasons Trimble (1981) noted that monetary compensation for a posttraumatic disorder is far less than that for physical injury despite the fact that the limitations on an individual’s life caused by psychological symptoms may be greater. There are only two ways that a malingerer may be identified with certainty. The first occurs when an individual participates in activities that he or she has claimed to be incapable of doing. For example, a man involved in an auto accident may

claim that he is unable to drive, yet may be seen driving to the store. The second occurs when the individual confesses to malingering. Given that neither option occurs often, the classification of malingering must involve the integration of multiple pieces of clinical evidence obtained from a thorough investigation and psychological assessment.

MALINGERING OFPTSD PTSD is easy to fake. The diagnosis is based almost entirely on the individual’s subjective report of symptoms, which are difficult to independently verify. Furthermore, in an effort to educate the public, the diagnostic criteria are widely available in print and online, allowing unscrupulous individuals to familiarize themselves with what PTSD symptoms to falsely report. Studies have indicated that participants naive to the criteria of PTSD could achieve the diagnosis on a checklist when asked to do so 86–94% of the time (Burges & McMillan, 2001; Lees-Haley & Dunn, 1994; Slovenko, 1994). The primary motivation for malingering PTSD is financial gain. Once PTSD was included in the DSM, the number of personal injury lawsuits doubled over the next decade (Olson, 1991). Workers’ compensation claims associated with stress-related disorders rose rapidly, and insurance company costs linked to these claims soon outstripped those related to physical injuries (de Carteret, 1994). By the late 1990s, 14% of all workers’ compensation claims were based on stress-related disorders (Guriel & Fremouw, 2003). While financial gain usually serves as the incentive, malingered PTSD to be classified as “disabled” may also serve as a means to “save face” by not admitting to more stigmatizing causes of disability, such as lack of social skills or substance use. The malingerer may gain sympathy and support where none existed before (Keiser, 1968). Finally, malingered PTSD may also be a last-ditch attempt to gain compensation when a physical injury claim is unsuccessful. The adversarial nature of the court system may serve as a breeding ground for symptom exaggeration or partial malingering. Plaintiffs with legitimate disabilities may enter into the process with no intention of lying. However, after exposure to an aggressive deposition by the defense attorney, they may become angry or worry that they will not receive appropriate compensation for their damages (Enelow, 1971). Therefore, plaintiffs may

192

II .   D i a g n o s t ic I ss u e s

exaggerate their symptoms in order to obtain what they believe they rightfully deserve. While prominent in civil litigation, PTSD has a smaller impact in the criminal courts. Appelbaum, Jick, and Grisso (1993) found that only 0.3% of defendants used PTSD as the basis for a not guilty by reason of insanity (NGRI) plea. Berger, McNeil, and Binder (2012) noted that a successful insanity plea based on PTSD is usually related to dissociative phenomena that would cause defendants to misperceive their circ*mstances and act in a way that would be reasonable in this context. Moskowitz (2004) points out that it may be difficult to determine if the dissociation caused the violence, or conversely, if the violent act spawned the dissociative episode. Individuals involved in the criminal courts and diagnosed with PTSD were more likely to be found competent to stand trial, have a jury trial, and be found guilty than those with other diagnoses (Guriel & Freemouw, 2003). However, a diagnosis of PTSD can result in a reduction of charges or mitigation of penalties (Pitman & Sparr, 1998). Defendants may suggest that PTSD-related aggression or recklessness impacted the process of forming criminal intent (mens rea). Courts have recently been more willing to recognize this, especially in reference to veterans (Porter v. McCollum, 2009), often leading to reduced sentences (Grey, 2012).

Use ofPsychological Testing toDetect FeignedPTSD PTSD has been characterized as a relatively easy disorder to feign (Demakis & Elhai, 2011; Koch, O’Neill, & Douglas, 2005; Wetter & Deitsch, 1996). As a result, the use of psychological testing in medicolegal settings has increased (Butcher & Miller, 1999), due to their effectiveness in assessing various forms of response bias (Rogers & Granacher, 2011). Both feigned and genuine PTSD can produce similar elevations on PTSD-relevant scales on psychological tests, which makes it difficult to distinguish between the two presentations (e.g., Elhai, Gold, Sellers, & Dorfman, 2001; Resnick, 1997). Rogers, Payne, Correa, Gillard, and Ross (2009) proposed that individuals with PTSD may experience intensified symptoms and impairment that elevate both clinical and validity scales. In addition, persons with genuine PTSD often present with highly variable symptom profiles, making it difficult to differentiate from feigning profiles (Foa, Riggs, & Gershuny, 1995; Guriel & Fre-

mouw, 2003). As noted, PTSD is also a highly comorbid disorder, especially with major depressive disorder, anxiety disorders, and substance use disorders (Kessler, 2000). Therefore, feigners are difficult to differentiate from their genuine counterparts due to the complex nature of PTSD. To be successful, feigners must elevate the appropriate symptom scales with significant severity, while avoiding detection on validity scales that may alert professionals to their deception. In the next section we briefly discuss the effectiveness of several popular psychological measures used to differentiate between genuine and feigned PTSD. Studies included in this section were used only if they had both a genuine PTSD and feigning group (i.e., simulators or known group). Cohen’s ds are used to assess the magnitude of the differences between genuine and feigned PTSD groups with the effect-size descriptors discussed by Rogers (Chapter 2, this volume). In addition, feigning scale scores are examined to determine their accuracy for feigning and genuine PTSD. Due to the far-reaching consequences of being classified as “malingering,” the minimization of false positives is emphasized to avoid misclassification of genuine responders as feigners. Minnesota Multiphasic PersonalityInventory–2

The Minnesota Multiphasic Personality Inventory–2 (MMPI-2; Butcher, 2001) continues to be one of the most commonly used personality inventories for the assessment of feigned PTSD and for some, the most effective test for detecting malingered PTSD (Demakis & Elhai, 2011). Three main detection strategies are used on the MMPI-2 to detect feigned PTSD: Quasi-Rare Symptoms, Rare Symptoms, and Erroneous Stereotypes. Table 10.3 summarizes eight simulation studies that have examined the effectiveness of the MMPI-2 to detect feigned PTSD. Utilizing the quasi-rare detection strategy, prior research has shown that the F scale is often significantly elevated among individuals with PTSD, frequently above established cutoff scores for feigning (Franklin, Repasky, Thompson, Shelton, & Uddo, 2002; Garcia, Franklin, & Chambliss, 2010; Jordon, Nunley, & Cook, 1992; Munley, Bains, Bloem, & Busby, 1995). Specifically, Greene (2000) found the F scale to sometimes be elevated in individuals with complicated histories and numerous symptom complaints, which often occur in those diagnosed with PTSD. As shown in Table 10.3, individuals with genuine PTSD in these eight studies

10. The Malingering of Posttraumatic Disorders 193

TABLE 10.3.  Effect Sizes and Scale Elevations for the MMPI-2 Validity Scales

Quasi-rare symptoms

Rare symptoms

Fb

F-K

Fp

Bury & Bagby (2002) a Eakin et al. (2006) Elhai et al. (2000) Elhai et al. (2001) Elhai et al. (2002) Elhai et al. (2004) Marshall & Bagby (2006) Wetter et al. (1993)

  1.06   0.89   0.93   1.10   0.91   0.53   1.17   1.52

  1.37   1.04

 0.97

  1.38   1.13

  1.24   0.97   1.01   1.41   1.31   1.21   1.53

 1.60

M effect size

  1.01

  1.13

  1.26

  1.24

  1.07

  1.17

  0.13

M PSTD scale elevation M PTSD SD

 84.98  22.33

 82.77  23.83

 5.19 10.53

 66.61  19.75

65.90 15.92

 77.71  14.45

 82.00  14.68

M feigning PTSD scale elevation M feigning PTSD SD

102.35  19.97

107.67  19.08

20.47 14.48

 89.74  23.85

84.85 23.68

 96.64  18.24

 83.30  14.46

PTSD screening cutoff (M + 1 SD) PTSD screening cutoff (M + 1.5 SD) PTSD screening cutoff (M + 2 SD)

107.31 118.48 129.64

106.60 118.52 130.43

15.72 20.99 26.25

 86.36  96.24 106.11

81.82 89.78 97.74

 92.16  99.39 106.61

 96.68 104.02 111.36

 1.10  1.37   0.75

Fptsd

Erroneous stereotypes

F

Ds

FBS

  1.39   0.84   0.87   1.03

  0.02  –0.09   0.47

 0.81  1.32   1.73

Note. All effect sizes are reflected as Cohen’s d values. Total patients with PTSD in review (N = 839); total feigners in review (N = 257). Mean PTSD patient-feigner elevations and SDs based only on samples in this review. aOnly the symptom-coached simulator group was used to provide uniform comparisons with other studies.

demonstrated marked elevations (M = 84.98T) on the F scale. These elevated scores may be related to the development of the scale, as the F scale simply measures divergence from normality but does not necessarily distinguish genuine from feigned presentations (Rogers, Sewell, Martin, & Vitacco, 2003). Developed similarly, the Fb scale also demonstrated marked elevations (M = 82.77T) among patients with genuine PTSD with moderate effect sizes (mean d = 1.13). The F-K scale demonstrated slightly improved effect sizes (mean d = 1.26), but it is important to note that the F scale and F-K index are highly correlated (Nichols, 2011). In fact, Bury and Bagby (2002) found that the F-K does not add incremental predictive capacity over the F family of scales and discouraged its use in assessing feigning psychological symptoms. The Fp and Fptsd scales outperformed the F and Fb scales, with lower clinical elevations for patients with genuine PTSD, with averages around 65T. The Fp scale specifically has gained popularity as a primary scale for the assessment of feigning on the MMPI-2 due to the relative consistency of cutoff scores and low probability of false positives (Rogers et al., 2003). The Fp was specifically de-

signed to assess the differences between genuine and feigned disorders, and appears to be effective with PTSD. As discussed in the previous edition of this book, a cutoff score of Fp > 8 (raw score) is suggested, which produces a relatively small falsepositive rate of 9%. Elhai and colleagues (2004) recognized that the F family scales were not specific to the detection of feigned PTSD, so the Infrequency-Posttraumatic Stress Disorder (Fptsd) scale was developed to differentiate between genuine and feigned combat PTSD. The Fptsd scale was developed empirically with infrequently endorsed items (< 20%) from a sample of male combat veterans with PTSD (N = 940). It includes 20 items from the F scale and 12 more items that reflect family or social problems, antisocial behavior, morally righteous attitudes, and self-injurious behavior. Two evaluative studies have shown its possible potential (mean d = 1.07; Elhai et al., 2004; Marshall & Bagby, 2006). Interestingly, genuine PTSD samples scored below 70T on average. Due to the number of similar items between the Fp and Fptsd, Marshall and Bagby examined but were unable to establish any incremental validity for Pptsd over the Fp. They

194

II .   D i a g n o s t ic I ss u e s

concluded that the standard validity scales may be able to detect malingering across a wide spectrum of psychiatric disorders. The Fake Bad (FBS) and Ds scales both capitalize on the erroneous stereotypes detection strategy, but show markedly different effect sizes. Gough’s Ds scale appears more effective with an average moderate effect size (mean d = 1.17). The FBS scale was designed for only personal injury cases (LeesHaley, 1992) and its effectiveness to detect feigned PTSD outside of that setting has been questioned due to its narrow focus (Arbisi & Butcher, 2004; Rogers et al., 2003). While only three studies examined the FBS, it does not appear to be effective at differentiating between feigned and genuine PTSD (mean d = 0.13) (see Table 10.3). While most studies involve pure malingerers, two studies have examined the effectiveness of the MMPI-2 to detect feigned PTSD in populations that have personal knowledge of PTSD. The MMPI-2 was slightly more effective in the detection of partial malingerers over pure malingerers using all three main detection strategies (Arbisi, Ben-Porath, & McNulty, 2006; Efendov, Sellbom, & Bagby, 2008). For example, the F scale produced very large effect sizes for these two studies (mean d = 1.62). Utilizing the rare symptom strategy, the Fp produced the largest effect sizes (mean d = 2.15), with the genuine PTSD sample averaging below 60T on the Fp in both samples.

MMPI-2 RestructuredForm

The MMPI-2 Restructured Form (MMPI-2-RF; BenPorath & Tellegen, 2008; Tellegen & Ben-Porath, 2011) includes revised versions of the MMPI-2 validity scales designed to detect feigning. Although a newly designed measure, the MMPI-2-RF validity scales have already shown utility in several studies of the detection of overreported symptoms even when individuals are coached about the presence and purpose of validity scales (Sellbom & Bagby, 2010). To date, only three studies have examined the ability of these scales to differentiate between genuine and feigned PTSD (see Table 10.4). Similar to the MMPI-2, the revised F Scale (F-r) has been found to be markedly elevated among individuals with PTSD (Arbisi, Polusny, Erbes, Thuras, & Reddy, 2011; Goodwin, Sellbom, & Arbisi, 2013; Marion, Sellbom, & Bagby, 2011). As seen in Table 10.4, the F-r scale for patients with genuine PTSD averaged 80.25T, with a large standard deviation of 24.26, making any PTSD screening cutoff scores extremely high. As a contrast, the Fp-r has demonstrated only slight elevations (M = 63.57T) for genuine PTSD but extreme elevations for feigned PTSD (M = 102.56T). Goodwin et al. (2013) and Marion et al. (2011) also examined the effectiveness of the Fp-r with sophisticated feigners; both studies produced good results (ds = 1.01 and 1.21, respectively). These studies revealed that

TABLE 10.4.  Effect Sizes and Scale Elevations for the MMPI-2 RF Feigning Scales

Quasi-rare symptoms

Rare symptoms

Erroneous stereotypes

F-r

Fp-r

Fs

FBS-r

Goodwin et al. Marion et al. (2011) a Mason et al. (2013)

  1.20   0.79   1.86

1.62 1.02 1.95

1.32 1.09 1.93

  0.74   0.50   1.01

M effect size

  1.28

1.53

1.45

  0.75

M PSTD scale elevation M PTSD SD

 80.25  24.26

63.57 15.47

72.78 21.67

 73.86  15.50

M feigning PTSD scale elevation M feigning PTSD SD

112.28  30.10

102.56 38.01

106.02 27.41

 84.96  14.56

PTSD screening cutoff (M + 1 SD) PTSD screening cutoff (M + 1.5 SD) PTSD screening cutoff (M + 2 SD)

104.51 116.64 128.77

94.45 105.285 116.12

 89.36  97.11 104.86

(2013) a

79.04 86.775 94.51

Note. All effect sizes are reflected as Cohen’s d values. Total patients with PTSD in review (N = 191); total feigners in review (N = 115). Mean PTSD patient-feigner elevations and SDs based only on samples in this review. aOnly the symptom-coached simulator group was used to provide uniform comparisons with other studies.

10. The Malingering of Posttraumatic Disorders 195

there is a 97.3% likelihood that a person is feigning PTSD when scores exceeded a cutoff score of Fp-r ≥ 100, with a false-positive rate of only 2.7%. This cutoff score is recommended by Ben-Porath (2012). However, a cutoff score of Fp-r ≥ 90 was recommended when base rates are expected to be high (i.e., .30 or above), such as PTSD feigning among combat veterans (Goodwin et al., 2013; Marion et al., 2011). The Infrequent Somatic Complaints (Fs), which is a new scale, includes items endorsed by less than 25% of the normative sample and a large sample of medical and chronic pain patients. While the Fs scale has been shown to differentiate between feigned and genuine PTSD groups (mean d = 1.45), this scale was not designed for this purpose. This finding suggests that PTSD feigners may be more likely to endorse symptoms related to somatic and pain complaints than those with genuine PTSD, but that the Fs scale should not be used in isolation for the detection of feigned PTSD. Overall, the MMPI-2-RF validity scales show promise in their ability to detect general feigning, but more research is necessary to understand their effectiveness in identifying feigned PTSD. The rare symptom scales (Fp-r) have demonstrated the

best results thus far, with genuine PTSD samples scoring relatively low on this scale. Personality AssessmentInventory

The Personality Assessment Inventory (PAI; Morey, 2007) has gained wide acceptance in forensic practice (Kucharski, Toomey, Fila, & Duncan, 2007). Evidence suggests it has been useful at assessing feigned PTSD due to its response validity scales and ability to assess a variety of symptoms related to the disorder (Morey, 1996; Mozley, Miller, Weathers, Beckham, & Feldman, 2005). In fact, the Veteran’s Affairs Healthcare System relies on the PAI to obtain valid, clinically relevant patient information (Calhoun, Collie, Clancy, Braxton, & Beckham, 2010). The Negative Impression Management (NIM) validity scale has been shown to be the most effective PAI feigning scale (see Table 10.5) in differentiating genuine and feigned PTSD groups, with average moderate effect size (mean d = 0.98). Importantly, many genuine PTSD responders did not have elevated NIM scores (mean = 63.41T). However, NIM elevations have also been associated with severe psychological impairment (Thomas,

TABLE 10.5.  Effect Sizes and Scale Elevations for the PAI Validity Scales

Rare symptoms NIM

NDSa

Spurious patterns of psychopathology MAL

RDF

 0.04

 0.36

 0.96  1.35

 0.21  0.92

  0.78

  0.50

Eakin et al. (2006) Liljequest et al. (1998) Scragg et al. (2000) Thomas et al. (2012) b

 0.19  1.06  1.17  1.48

M effect size

  0.98

M PSTD scale elevation M PTSD SD

63.41 14.74

57.93 12.86

59.10 14.19

52.74 11.07

M feigning PTSD scale elevation M feigning PTSD SD

84.21 88.82

93.61 25.60

70.74 18.31

59.56 10.74

PTSD screening cutoff (M + 1 SD) PTSD screening cutoff (M + 1.5 SD) PTSD screening cutoff (M + 2 SD)

78.15 85.52 92.89

70.79 77.22 83.65

73.29 80.39 87.48

63.81 69.35 74.88

 1.76

Note. All effect sizes are reflected as Cohen’s d values. Total patients with PTSD in review (N = 117); total feigners in review (N = 158). Mean PTSD patient-feigner elevations and SDs based only on samples in this review. aOnly one study has examined the NDS for feigning PTSD. Therefore, all calculations in this column are based solely on the data from Thomas et al. (2012). bOnly the symptom-coached simulator group was used to provide uniform comparisons with other studies.

196

II .   D i a g n o s t ic I ss u e s

Hopwood, Orlando, Weathers, & McDevitt-Murphy, 2012) and has been found to be elevated in inpatient trauma samples (Rogers, Gillard, Wooley, & Ross, 2012). When evaluating cut scores, Calhoun, Earnst, Tucker, Kirby, and Beckham (2000) found that a cutoff score of NIM ≥ 13 in a sample of combat veterans with PTSD allowed 39% of feigners to go undetected and 35% of genuine PTSD responders to be misclassified. Therefore, the NIM score should be considered in conjunction with other data, as NIM scores may be related to high levels of distress or severe psychopathology. Spurious patterns of psychopathology appear to be less effective than the rare symptoms strategy (i.e., NIM), with the Malingering Index (MAL) producing a barely moderate effect size (mean d = 0.76). The RDF produced even smaller effect sizes (mean d = 0.46), with both genuine PTSD and feigners demonstrating low scores (Table 10.5). A newly created feigning scale, the Negative Distortion Scale (NDS), was designed in an attempt to identify feigners in samples with high levels of psychopathology (Mogge, Lepage, Bell, & Ragatz, 2010). Like the NIM scale, the NDS utilizes the rare symptoms detection strategy. Unlike the NIM, the NDS was created from eight clinical scales by using rarely endorsed items by inpatients. Moreover, 60% of the NDS items are from trauma-relevant scales, which could potentially be problematic for identifying feigned PTSD. Of note, Rogers, Gillard, Wooley, and Kelsey (2013) found the NDS to be highly effective (d = 1.81) compared to the other PAI validity indicators in identifying feigned psychopathology within an inpatient sample simulating distress for disability. In the only NDS study of feigned PTSD, Thomas and colleagues (2012) found impressive effect sizes for simulators: (1) symptom-coached (d = 1.76) and (2) coached on both symptoms and validity indicators (d = 1.46). Thomas et al. concluded that a cutoff score of 85T correctly classified 97% of patients with genuine PTSD, while still correctly classifying 64% of the feigners. Further research is necessary to provide more conclusive evidence that the NDS can accurately differentiate between genuine and feigned PTSD. Trauma SymptomInventory–2

The Trauma Symptom Inventory (TSI; Briere, 1995) is a popular self-report measure designed to assess psychological symptoms commonly associated with traumatic experiences. However, the

TSI was criticized for its lack of clinical research, as most studies utilized college samples (Edens, Otto, & Dwyer, 1998; Guriel et al., 2004; Rosen et al., 2006). While simulators in these studies did generally score higher than genuine PTSD samples on the Atypical Response (ATR) scale, the classification rates were not impressive (Elhai, Gray, Kashdan, & Franklin, 2005). The Trauma Symptom Inventory–2 (TSI-2; Briere, 2010), an updated version of the TSI with new scales and norms, includes a revised ATR scale. Specifically, the content of the revised ATR scale includes items that are unlikely to be endorsed by genuine PTSD samples, rather than the bizarre or extreme symptomatology found on the original ATR scale (Gray, Elhai, & Briere, 2010). However, only Gray and colleagues have examined the revised ATR scale. Utilizing undergraduate simulators, the recommended cutoff score of ≤ 7 on the revised ATR produced a substantial falsepositive rate (23%), possibly indicating that the revised ATR may be susceptible to general distress or other comorbid symptoms. Significant concerns continue regarding the ATR in identifying feigned PTSD on the TSI-2. The Detailed Assessment ofPosttraumaticStress

The Detailed Assessment of Posttraumatic Stress (DAPS; Briere, 2001) is a 104-item self-report measure designed to evaluate trauma exposure and posttraumatic response. The DAPS includes eight clinical scales that measures trauma-specific symptoms and posttraumatic stress reactions. Additionally, the DAPS examines several associated features of PTSD: trauma-specific dissociation, substance abuse, and suicidality. Of particular importance, the Negative Bias (NB) scale assesses the tendency of respondents to overendorse unusual or unlikely symptoms. Furthermore, the DAPS includes decision rules that combine to create a provisional diagnosis of PTSD (Briere, 2001). The DAPS has only been empirically investigated in one simulation study by Wooley and Rogers (2015) on feigning PTSD. They discouraged the use of the NB scale to detect feigned PTSD based on its disappointing results. The NB scale was found to have remarkable variability for both patients with genuine PTSD (SD = 41.54) and feigners (SD = 69.51). Not surprisingly, the NB scale’s effect sizes suffered as a result of this variability (mean d = 0.71). The DAPS manual (Briere, 2001)

10. The Malingering of Posttraumatic Disorders 197

indicated that only about one-third (32.8%) of the validation sample was from “clinical settings,” and few details were provided about the severity of psychopathology in this subgroup. Therefore, the NB scale may not be effective in clinical samples, and more research is necessary to know whether the scale is sensitive to clinical distress.

Well-DefinedGroups The laser accuracy of single-point cutoff scores is the traditional approach to classifying individuals as feigning using validity scales, in which clinicians use a one-point difference to differentiate honest responders from feigners. Rogers and Bender (2013) provided an alternative by identifying scores that are “too close to call” and susceptible to being wrongly classified. To identify an indeterminate range, researchers can exclude scores that occur either within ±1 standard error of measurement (SEM) or ± 5T from the designated cutoff score. Previous studies have found false-positive rates of over 50% in some indeterminate ranges (Rogers, Gillard, Wooley, & Kelsey, 2013; Wooley & Rogers, 2015). The use of well-defined groups could improve the effectiveness of these psychological tests in differentiating between feigned and genuine PTSD, while lowering false-positive rates. A common criticism of using an indeterminate group is the number of individuals excluded from classification. In some samples, establishing welldefined groups with ± 1 SEM can result in a large indeterminate group. However, the far-reaching consequences of being wrongly classified as malingering should be a priority. An advantage of examining this issue is that researchers can evaluate the use of well-defined groups with data already collected (Rogers & Granacher, 2011). Many of the measures used to detect feigned PTSD either have a high false-positive rate or are not specifically designed to detect feigned PTSD. A newly designed measure, the Posttrauma Response Set Measure (PRSM) presented at the American Psychology–Law Society in 2015, was created with these concerns in mind (Weaver & Yano, 2015). This measure includes a diagnostic index to assess genuine PTSD, as well as five malingering indices to detect various forms of feigned PTSD. When compared to the MMPI-2 validity scales, the malingering indices of the new measure (areas under the curve [AUCs] from .69 to .89) were found to be superior in differentiating between the honest responders and feigners (Morson, Davis, & Weav-

er, 2016). Specifically, all of the feigning scales on the brief form were significantly better at detecting feigned PTSD than the F and Fp scales on the MMPI-2. The PRSM has yet to be published or to undergo the peer-review process, but it appears to have potential in the detection of feigned PTSD.

Conclusions aboutPsychological Tests andFeignedPTSD The presence of malingering is often a major clinical concern when conducting assessments in a medicolegal setting. Psychological assessments can help clinicians reach more accurate conclusions. This review highlights several important findings that clinicians should consider when deciding what psychological tests to administer. • The rare and quasi-rare detection strategy scales appear to be the most successful, with few exceptions, at differentiating feigned and genuine PTSD, as demonstrated with the PAI, MMPI-2, and MMPI-2-RF. • The selection of measures to use should be based on research-based evidence rather than popularity. While newer measures hold promise, more research is necessary to understand the level of their effectiveness across professional settings. • Decisions about which cutoff scores to use can vary widely based on the primary purpose. For example, lower cutoff scores can be used as a way to screen potential feigners and identify which cases need further investigation. In contrast, more conservative cutoff scores can help avoid misclassification by minimizing false positives and provide evidence for possible feigning. In addition, the prudent use of an indeterminate range could further reduce false positives. Psychological testing should not be used alone to make conclusions about feigned PTSD and should be corroborated with other evidence (Burchett & Bagby, 2014; Rogers, 2008). Official military records, hospital records, police reports, and employment records may help to substantiate reported trauma exposure and may reveal previous traumas that might account for the current symptom presentation (Demakis & Elhai, 2011). Clinical interviews may also provide valuable information on whether the patient is inconsistent across his or her presentation or reporting absurd symptoms. In summary, the use of psychological

198

II .   D i a g n o s t ic I ss u e s

testing to detect feigned PTSD is important, but it is only one component in the assessment process.

MALINGERING OFOTHER TRAUMA‑RELATEDDISORDERS In addition to PTSD, trauma may lead to other conditions, including chronic pain and neurological sequelae. Otis, Pincus, and Keane (2006) noted that 20–34% of people who are seen for chronic pain also have significant PTSD symptoms. Those with PTSD tend to have more intense pain (Geisser, Roth, Bachman, & Eckert, 1996), and their level of disability is greater (Sherman, Turk, & Okifuji, 2000). When no clinical evidence supports a physiological cause of these neurological symptoms, they may be the by-product of the unintentional (conversion disorder) or intentional (malingering) production of symptoms. Tests to detect malingering in organic disease are invalid in conversion disorder (Lipman, 1962; Smith, 1967). The differential diagnosis is further complicated by the fact that an individual with conversion disorder may also malinger. Clinicians’ ability to distinguish between the two disorders rests on their ability to assess consciousness, an incredibly challenging task. It is worthwhile for the clinician to search for the unconscious gain associated with the individual’s symptoms; patients may unwittingly reveal gains of which they themselves are not aware. Additionally, patients with conversion disorder tend to willingly engage in evaluations because they too are eager for an explanation of their symptoms (Trimble, 1981) and are anxious for a cure (Hofling, 1965). Interestingly, a meta-analysis demonstrated that la belle indifférence (a patient’s lack of concern about his or her symptoms), traditionally associated with conversion disorder, now appears to be more common with organic illness (Stone, Boone, Back-Madruga, & Lesser, 2006). Comorbid personality disorders, most often histrionic and dependent personality disorders, may predispose an individual to conversion symptoms. Other factors include having little formal education, low IQ, low socioeconomic status, and limited psychological/medical knowledge, as well as an existing neurological disorder on which the individual may unconsciously base his or her symptoms (Kaplan & Sadock, 2003; Yutzy, 2002). Depression is quite common after a traumatic accident. It may be related to both physical harm

and emotional loss due to an inability to work or play the same role in one’s family dynamic. Symptoms of depression often accompany PTSD and overlap with PTSD symptoms. Tests evaluating depression rely primarily on self-report, so they can easily be manipulated. PTSD requires a careful differentiation between diagnosis and comorbid diagnoses, in addition to feigning. Compensation neurosis is a term that was first coined in 1879 in response to the injured railroad workers (Rigler, 1879). It was defined in 1979 as a “behavior complex associated specifically with the prospect of recompense and in contradistinction to traumatic neurosis and psychiatric illness... precipitated by the stress of illness, accident or injury” (Rickarby, 1979, p.333). Although not recognized in any version of the DSM, it may be offered in court hearings by defense attorneys. In 1961, Miller stoked the controversy by proposing a similar diagnosis following head injuries entitled “accident neurosis.” The clinical features included the following: 1. Subjects’ unshakable conviction in their lack of fitness for work. 2. Inverse relationship between the severity of the injury and the degree of disability. 3. Absolute failure to respond to treatment until compensation had been awarded. Miller’s study showed that 48 of 50 individuals recovered completely within 2 years of settlement of their personal injury claim. One problem with Miller’s proposal is that he used a broad definition of malingering, suggesting that the motivation behind the production of symptoms (conscious vs. unconscious) is of little consequence. Several studies attempted to replicate his findings but have demonstrated the opposite effect; that is, the removal of the factor of litigation (either because the individual was not eligible for compensation or a settlement was reached) did not cause an improvement in the patient’s symptoms (Kelly & Smith, 1981; Mendelson, 1981; Parker, 1977; Thompson, 1965). Moreover, compensation neurosis is a pejorative term (Modlin, 1960) for a disorder that is not supported by the literature and therefore should not be a valid diagnosis. Compared to PTSD, psychosis is infrequently malingered after personal injury. While fraudulent plaintiffs are willing to invest time and energy into crafting a believable story, few are willing to undergo inpatient hospitalization (Davidson, 1952)

10. The Malingering of Posttraumatic Disorders 199

and treatment with potent medications. It is also quite challenging to pretend to be psychotic over a long period of time, which might be necessary, as litigation may span several months to years.

BOX 10.1.  Clues to Malingered Amnesia 1. Overplaying memory deficits 2. Improbable answers to overlearned data 3. Alleged impairment of procedural memory

ASSESSMENT OFMALINGEREDPTSD Conceptual Issues withMalingeredPTSD Ali, Jabeen, and Alam (2015) suggest several reasons for the importance of detecting feigned PTSD. First, feigners misdiagnosed with PTSD may receive unwarranted and even harmful therapies. Second, feigning is clearly disruptive to the therapeutic relationship between malingerers and their therapists. It may then create a ripple effect, with clinicians becoming mistrustful of all their patients. Third, malingering negatively impacts the economy. Fourth and finally, malingering creates inaccuracies in the medical database, which can impact research regarding PTSD (Rosen, 2006). Antisocial Characteristics oftheMalingerer

The presence of antisocial traits (Hollander & Simeon, 2002) and psychopathic traits (Edens, Buffington, & Tomicic, 2000) may arouse suspicions of malingering. In contrast, persons who are consistently contributing members of society are less likely to malinger (Davidson, 1952). Similar to persons with antisocial traits, malingerers often have poor social and occupational functioning prior to the trauma (Braverman, 1978). These may include sporadic employment with long absences from work or previous incapacitating injuries. Assessing MemoryImpairment

Amnesia may play a role in PTSD but it can certainly be feigned as well. It is important to note that some memory distortion may be expected over time in patients with genuine PTSD (Loftus, 1979). In general, malingerers tend to overplay their memory deficits. They may allege an inability to recall overlearned data, including their name, gender, or Social Security number (Brandt, 1992; Levin, Lilly, Papanicolau, & Eisenberg, 1992). Even with a legitimate history of head trauma, procedural memory is rarely impaired. Therefore, malingering should be suspected when examinees states that they can no longer recall how to ride a bike or drive a car. Malingering is also more likely if an individual scores more poorly on those ques-

4. Poor performance on tests labeled “memory testing” 5. Performing worse than chance on memory testing 6. Clear recollection of memory loss examples

tions clearly labeled as “memory testing” or performs worse than chance on forced-choice tests such as the Test of Malingered Memory (TOMM). Some clues to feigned amnesia are presented in Box 10.1.

Behavior duringtheEvaluation Associated withMalingering Professionals can think of a malingerer as an actor who is playing a role, and the performance will reflect the individual’s preparation (Ossipov, 1944). Thus, a lack of knowledge regarding PTSD will likely result in a poor performance during the evaluation. Malingerers may overact their part by giving an excessively dramatic report of their symptoms. Alternatively, they may adopt a globally evasive posture, hesitating to discuss their return to work or the money they stand to gain from the resolution of their case (Powell, 1991). To avoid being specific, the malingerer may answer questions vaguely, than to clinicians’ focused questions. Conversely, successful malingerers often endorse fewer symptoms and avoid those that are unusual or dramatic (Edens et al., 2001). Malingerers may even “go on the offensive” and attempt to control the interview by trying to intimidate evaluators. They may go so far as to accuse the interviewer of suggesting that they are malingering. During the evaluation, the clinician should take careful note of inconsistencies. Internal inconsistencies occur when individuals report severe symptoms, then contradict themselves (e.g., reports of significant memory impairment, then “remembering” several specific instances when they were unable to recall information). Internal inconsistencies also occur if malingerers provide conflicting accounts of their own story to the same evaluator. External inconsistencies occur when an individual’s symptoms stand in contrast to what is observed or

200

I I .   D i a g n o s t ic I ss u e s

learned (e.g., reported poor concentration is contradicted by engaged responding during a 3-hour evaluation). External inconsistencies also occur if individuals report a decreased level of social functioning but friends state that they often engage in social activities. Finally, external inconsistencies occur when there is a discrepancy between the individual’s report and hospital or police records.

Collateral Information andFeigning Collateral information is crucial to validating the examinee’s reported symptoms and the traumatic event. It is helpful to gather collateral data prior to the evaluation, so that evaluators may address any inconsistencies that arise. A review of police reports or witnesses’ accounts of the trauma may provide a more objective view of the events that occurred. Reports written at the time of the incident also help to curtail biased retrospective recall, in which claimants may amplify their memory of the traumatic event (Harvey & Byrant, 2000; Koch et al., 2005). In a similar vein, it is best to insist on seeing the complete set of medical records rather than accept a summary written by the examinee’s treater. Such summaries often favor what the treater perceives to be in the best interest of the patient. Employee files, school records, and tax returns may provide insight into the claimant’s daily functioning prior to the trauma and should be compared to the claimant’s subjective report. Persons most helpful in providing collateral information are those who are close to the claimant but do not stand to gain from the litigation. Information gained from collateral sources helps to establish an individual’s baseline level of functioning prior to the trauma. For example, an individual may claim that the trauma resulted in poor concentration, whereas a coworker may report that the claimant has always been easily distracted. Collateral sources may also be able to provide observational data, such as the claimant’s body movements during dreams, sleep patterns, and the presence of other symptoms consistent with PTSD. Access to these informants often depends on the jurisdiction and whether the expert is employed by the plaintiff or defense attorney. If the evaluator is unable to gain access to relatives, the retaining attorney may sometimes question them in a deposition.

The Interview andMalingeredPTSD The interview, when conducted in a careful and thorough manner, is a key component in differen-

tiating between genuine and malingered PTSD. During the interview, the evaluator must take care not to reveal PTSD criteria or convey any bias regarding its diagnosis. If evaluators adopt a confrontational style, examinees may feel compelled to exaggerate symptoms in order to justify their impairment. Interviewers should initially adopt an open-ended questioning style and avoid leading questions that give clues to correct responses. Because PTSD criteria are well-known, evaluators should insist on a detailed account of the specific symptoms the individual is reporting. Falsified symptoms often have a vague or artificial quality (Pitman, Sparr, Saunders, & McFarlane, 1996), lacking the nuances of personally experienced symptoms. Statements such as “I have nightmares” cannot be accepted at face value. Details, including circ*mstances, degree, frequency, and context, must be explored. For example, malingerers may claim that their nightmares are repetitive and have occurred without variation or decreased frequency over the past several months. Genuine dreams related to trauma often decrease over time. They may, however, increase in frequency when the individual is reminded of the event (e.g., giving a deposition) or experiences new stressors. Fifty percent of nightmares with PTSD show variations on the theme of the traumatic event (Garfield, 1987; Wittman, Schredl, & Kramer, 2007). After a traumatic event, the event may be dreamt almost literally a few times, then gradually include other elements as the event is woven into the rest of the person’s dream life. For instance, a woman who was raped may dream about another situation in which she feels helpless or trapped. A person malingering PTSD is less likely to report variations on the theme of the traumatic event. Posttraumatic nightmares, in contrast to lifetime nightmares unrelated to trauma, are almost always accompanied by considerable body movement (van der Kolk, Blitz, Burr, Sherry, & Hartmann, 1984). Inman, Silver, and Doghramji (1990) found that persons with PTSD-related insomnia, compared to others with insomnia, were more likely to be afraid of the dark and of going to sleep. They were more likely to wake up with the covers torn apart and to talk and shout during their sleep. They more often woke up confused and disoriented, and had difficulty returning to sleep. Nightmares may occur earlier in the night in those with PTSD (Germain & Nielsen, 2003). During an interview with an individual malingering PTSD, the evaluator may note the absence

10. The Malingering of Posttraumatic Disorders 201

of behavioral manifestations of the disorder, such as irritability, inability to focus, or exaggerated startle response. The interviewer may also find that the claimant minimizes other possible causes of his or her symptoms or portrays the pretrauma level of functioning in a overly positive light (Layden, 1966). Careful consideration must be given to the temporal course of symptom development in relationship to the trauma. It may also prove useful to inquire about claimants’ capacity to work versus their ability to enjoy recreational activities. Malingerers may state that they are incapable of remaining employed but acknowledge their active participation in hobbies. Person with genuine PTSD would likely withdraw from both work and recreation. Third parties should not be present during the interview. Informants may alter their stories, consciously or not, based on the claimant’s report. Also, if the evaluator confronts an examinee about suspected malingering, he or she will be far less willing to admit this with others present in the room. In some instances, the evaluator may feel that the situation warrants the use of subterfuge to expose the suspected malingerer. Prior to starting the interview, the evaluator could converse with a colleague about PTSD within earshot of the examinee and make mention of symptoms clearly not associated with the disorder (e.g., grandiosity or rapid speech). The examiner can then see whether the examinee volunteers these symptoms of the disorder.

PsychophysiologicalAssessment Measurement of the body’s responses to indicators of trauma may serve as one of the few objective methods of distinguishing genuine from malingered PTSD. A physiological reaction to cues related the traumatic event would fulfill Criterion B (Pitman, Saunders, & Orr, 1994). Several studies have evaluated the success of evaluators in diagnosing PTSD based on psychophysiological assessment. Blanchard, Kolb, Pallmeyer, and Gerardi (1982) played an audio recording of combat sounds and measured the heart rate, systolic blood pressure, and muscle tension in the forehead of veterans with PTSD and a control group. The two groups responded differently in terms of heart rate, blood pressure, and forehead electromyography (EMG). The investigators reported that they were able to identify patients with genuine PTSD with 95.5% accuracy using the heart rate response.

Lang (1985) improved on the stimulus used to trigger physiological changes in the previous protocol by using script-driven imagery. This stimulus allowed for a more accurate re-creation of an individual’s unique stressors in a given traumatic event. For example, an army medic might have a strong physiological response to the sound of a helicopter. Pitman and colleagues (1994) requested that 16 combat veterans without PTSD simulate the physiological profile of a patient with PTSD and found that 75% failed in this task. Studies have shown that psychophysiological testing may also be useful in predicting the development of PTSD and estimating treatment response. In 2007, O’Donnell, Creamer, Elliott, and Bryant measured the tonic (resting) and phasic (aroused) heart rate of victims 1 week posttrauma, then evaluated them for symptoms of PTSD 12 months later. They noted that high heart rate reactivity scores (phasic minus the tonic heart rate) predicted the development of PTSD. In 2010, Suendermann, Ehlers, Boellinghaus, Gamer, and Glucksman showed trauma survivors evocative images 1 month after their injuries. They also noted that heart rate reactivity predicted PTSD at 1 month and at 6 months, but skin conductance response was not related. In 2015, Wangelin and Tuerk studied 35 male combat veterans, who received prolonged exposure therapy. Those who completed the therapy showed significantly reduced heart rate and skin conductance reactivity to traumatic imagery. In the largest study of the psychophysiology of PTSD, Keane and colleagues (1998) subjected Vietnam veterans to scripts of traumatic events for three groups: current diagnosis of PTSD (N = 778), past diagnosis of PTSD (181) and no diagnosis of PTSD (369). Many veterans with current PTSD exhibited increased heart rate, skin conductance, and muscle tension compared to those with no diagnosis. The veterans with a past diagnosis fell somewhere in the middle. Unfortunately, however, one-third of those with a current PTSD diagnosis did not respond physiologically, which substantially limits the utility of this assessment method. Another limitation of this method is that persons without PTSD may demonstrate physiological reactivity to upsetting events. In the McNally (2006) study, patients claiming to have been abducted by space aliens (i.e., not a Criterion A diagnosis for PTSD) were exposed to a narrative script similar to their own reported experience. Their psychophysiological responses were at, if not

202

I I .   D i a g n o s t ic I ss u e s

above, those of the veterans with PTSD in Keane and colleagues’ (1998) study. Thus, a genuine physiological response cannot confirm the presence of PTSD. In light of the discrepancies in psychophysiological testing in PTSD, Pole (2006) conducted a meta-analysis to determine which variables influence effect sizes between PTSD and control groups. The effect size was greatest between the two groups when (1) the control group has not been exposed to trauma before, (2) the PTSD group had severe symptoms, and (3) members of the PTSD group were diagnosed using stricter DSM-III criteria. The findings are commonsensical. Removing trauma from the control group and increasing the severity of the symptoms in the PTSD group will increase the observed differences. In summary, psychophysiological testing does provide an objective but flawed means of differentiating between genuine and malingered PTSD. A significant minority of those with genuine PTSD do not demonstrate a physiological response, possibly leading to an incorrect classification as a malingerer. Conversely, some individuals without PTSD demonstrate physiological reactivity, possibly leading to an incorrect diagnosis of PTSD. Given this lack of accuracy, this type of psychophysiological evidence should not be admitted as evidence in court.

A MODEL FORTHEDIAGNOSIS OFMALINGEREDPTSD No single piece of data is pathognomonic for malingered PTSD. Rather, the diagnosis requires a careful assessment of all the evidence gathered, including a meticulously detailed history of symptoms, past and present, plus social and occupational functioning and careful corroboration of evidence. A clinical decision model for determining malingered PTSD is presented in Table 10.6. The model requires the establishment of (1) the individual’s motivation for feigning his or her symptoms, (2) the presence of at least two of the more common characteristics associated with malingered PTSD, and (3) strong collateral information supporting malingering. An evaluator may choose to confront the examinee after a thorough investigation has led to this conclusion. Direct accusations of lying are rarely successful. A more fruitful approach involves asking the examinee to clarify the inconsistencies. The clinician may also convey sympathy regard-

ing the reasons the individual feels compelled to exaggerate symptoms. This approach creates an environment in which the examinee may feel more comfortable acknowledging his or her malingering. For example (Inabu & Reid, 1967), it is preferable to say, “I am not sure you have told me the whole truth,” instead of “You have been lying to me.” Conversely, causing the examinee to feel ashamed may produce feelings of anger and opposition, which may then lead to a greater desire to continue the charade. In some extreme instances, these feelings may even lead to physical violence directed at the examiner. Finally, once a malingerer is confronted and chooses to deny his or her actions, subsequent acknowledgment is highly unlikely. One federal appellate court ruled that malingering is equivalent to obstructing justice and could result in a harsher penalty (Knoll & Resnick, 1999). If a criminal defendant is malingering, the examiner may therefore choose to disclose this information in the hope of producing a more ­honest response. However, such a notification could have the opposite effect, causing the defendant to cling more fiercely to the malingered symptoms.

TABLE 10.6.  Clinical Decision Model forEstablishing Malingered PTSD

A. Establish known motivation for malingering B. Characteristics of the malingerer (two or more of the following criteria): 1. Irregular employment or job dissatisfaction 2. Prior insurance claims 3. Capacity for recreation, but not work 4. Lack of nightmares or nightmares that are inconsistent with presentation 5. Antisocial personality traits 6. Evasiveness or contradictions 7. Unwillingness to cooperate or hostile behavior in the evaluation C. Confirmatory evidence of malingering (one or more of the following criteria): 1. Admission of malingering 2. Incontrovertible proof of malingering (e.g., video recording of a man at a party after claiming that his anxiety did not allow him to leave his home) 3. Unambiguous psychometric evidence of malingering 4. Strong corroborative evidence of malingering (e.g., video recording contradicting alleged symptoms)

10. The Malingering of Posttraumatic Disorders 203

A SPECIAL CASE: PTSDANDMALINGERING INTHECOMBATVETERAN PTSD inVeterans Over the past five decades, the United States has been involved many military conflicts exposing our military forces to trauma capable of producing PTSD in areas such as Vietnam, Somalia, the Persian Gulf, Iraq, and Afghanistan. In 1980, the Department of Veterans Affairs (VA) began to accept PTSD as a diagnosis for the purposes of assigning service-connected benefits (Bitzer, 1980). Following this, veterans became exposed to the PTSD diagnostic criteria via (1) distributed literature (Atkinson, Henderson, Sparr, & Deale, 1982), (2) contact with veterans diagnosed with PTSD (Lynn & Belza, 1984), and (3) the Internet. Some Internet sites provide suggestions on how to maximize the chances of receiving service-connected benefits (vva.org/what-we-do/outreach-programs/ ptsd) or how to write “stress letters” to submit to the VA ratings board (ptsdsupport.net). In 2007, Sparr and Pitman reported that PTSD is now the most common psychiatric condition for which veterans seek compensation. Following the publication of the DSM-5, Levin et al. (2014) suggested that the VA may be forced to exclude the A3 criterion (learning of the violent or accidental death of a close friend) for compensable injuries or risk a significant increase in PTSD-related claims. In the differential diagnosis of combat PTSD, professionals must also consider malingering, factitious disorder, ASPD, and genuine PTSD secondary to a non-combat-related cause.

Motivation toMalinger Combat-RelatedPTSD Three main factors can motivate veterans who malinger PTSD: (1) to obtain financial compensation, (2) to reduce criminal culpability, and (3) to obtain admission to a VA hospital. Veterans who successfully malinger PTSD may be well rewarded; those qualifying for a “100% service connection” for PTSD may be eligible to receive up to $40,000 a year for their disability (Burkett, & Whitley, 1998). Once veterans are qualified for PTSD-related disability, there is an ongoing financial incentive to remain disabled regardless of their true status (Mossman, 1994). The Department of Veterans Affairs Office of the Inspector General (2005) surveyed 2,100 veterans with at least 50% service-connected PTSD. It concluded that 25.1% were misdiagnosed and therefore not entitled to their benefits. Extrapolat-

ing to all veterans with service-connected PTSD, an estimated $19.8 billion had been erroneously paid out over a lifetime to those not meeting PTSD diagnostic criteria. In addition, a sample of 92 veterans with alleged PTSD continued to make regular mental health care visits until they received a 100% service connection. Many reported that their symptoms worsened over time despite their treatment. However, when 100% service connection was assigned, 39% of patients’ mental health visits dropped on average by 82%, and some veterans received no mental health care at all (Department of Veterans Affairs Office of the Inspector General, 2005). A diagnosis of PTSD confirmed by the VA may serve as an excusing or mitigating factor in legal cases. Malingered PTSD might play a role in decreasing criminal responsibility for veterans alleging that they experience (1) a dissociative state and resort to survival skills learned in military training (e.g., killing), (2) survivor guilt and want to have themselves killed (e.g., “suicide by cop”) in order to rejoin fallen comrades, and (3) sensation-seeking behavior (e.g., drug offenses) in order relive combat excitement (Wilson, 1981). When assessing the relationship of PTSD to a crime, the evaluator should first consider whether the crime scene recreated a traumatic event, and if so, whether the veteran experienced dissociative symptoms at the time that the crime was committed. Given the easy opportunity for monetary gain, veterans may fraudulently claim that they developed symptoms of PTSD from combat even though they were never exposed to battle (Lynn, & Belza, 1984). Some individuals who never even served in the military have successfully acquired service-connected benefits for PTSD (Burkett & Whitely, 1998). A key step in either supporting or discrediting a veteran’s story involves the collection of collateral data. There are a number of ways that a clinician may attempt to do so.

Collateral Information forClaimed CombatPTSD Military records, including discharge papers (also known as a DD214), often identify stressors to which a veteran has been exposed, but the clinician should be aware that these documents can easily be forged (Burkett & Whitley, 1998). Forgeries may be avoided by obtaining the documents directly from the U.S. Department of Defense (Sparr & Atkinson, 1986). Individuals malingering PTSD may state that their records do not re-

204

II .   D i a g n o s t ic I ss u e s

flect their covert missions or “black ops”; therefore, no evidence of their experiences exist. In such cases, examiners should look for the special training required for these missions in a veteran’s file, and the term classified assignment will appear next to the date of the mission (Burkett & Whitley, 1998). Collaboration with employees at the VA or experienced combat veterans may help to elucidate false claims (Lynn & Belza, 1984). Malingerers may fabricate stories that are as vivid and horrifying as the experiences of true combat veterans (Burkett & Whitley, 1998; Hamilton, 1985). However, they may reveal their dishonesties by incorrectly identifying certain details, including the geography and culture of the area, the military terminology used at the time, and dates related to specific events (Burkett & Whitley, 1998). Family members and friends of combat veterans may be interviewed to determine both the validity of the symptoms and the social functioning of the veteran prior to service in the military. Frueh et al. (2005) performed an archival study of 100 consecutive cases of men who reported combat trauma related to their experiences in Vietnam. Almost all (94%) had received a diagnosis of PTSD. For 41% of the veterans, archival data verified their stories. One-third had served in Vietnam, but the records indicated that they served in positions that would have made exposure to trauma highly unlikely. Five percent had not been in Vietnam or had not served in the military at all. Finally, those with documented combat experience reported twice the rate (28 vs. 12%, respectively) of battlefield atrocities when compared to those with verified combat exposure.

TheInterview Evaluators must recognize the possibility of experiencing powerful emotions and potentially biasing their evaluation. The recounting of horrific combat experiences can produce a highly charged, affect-laden environment that can be stressful for both the evaluator and the examinee (McAllister, 1994). Interviewers should also recognize the strong emotions that may be stirred up in examinees while recounting their combat-related experiences; some veterans may therefore be hesitant to discuss their painful experiences. Evaluators should be cognizant of a potential inclination to diagnose PTSD in a veteran out of a sense of moral obligation (Atkinson et al., 1982; Pankratz, 1985). Other veterans may exaggerate their symptoms in

fear of not receiving the compensation that they truly deserve (Fairbank, McCaffrey, & Keane, 1986).

Clinical Indicators ofMalingered Combat‑RelatedPTSD Indicators evident of malingered combat-related PTSD are primarily based on case reports and anecdotal descriptions. Thus, they should be used only in combination with other evidence to support the classification of malingering. Table 10.7 compares genuine and malingered combat-related PTSD. Those with genuine PTSD usually come for an evaluation because of the encouragement of loved ones, have repeated loss of employment and depression, display outbursts of anger, and sometimes have issues with substance use. They may feel guilty about surviving while fellow soldiers died in battle (Burkett & Whitley, 1998). They often attribute blame to themselves and may not directly relate their current state to their combat experience (Melton, 1984). They are more likely to downplay their combat experiences and current symptoms by saying things such as “Lots had it worse than me.” They may also hesitate to express how emotionally traumatizing the experience of war was for them. They are more likely to provide specific, trauma-related examples of avoidant behavior, such as not going out on very hot, dry days that resemble the weather in the Iraq. Conversely, malingerers often present themselves as “victims” and attribute a multitude of problems directly to their experience in the military. They more commonly present with a fear of “losing control” and harming others (Melton, 1984). This fear may even be used as an excuse to gain admission to a psychiatric unit. They might provide tales of their own improbable heroics during wartime, including escape from a prisoner of war (POW) camp or singlehandedly fighting multiple enemies in hand-to-hand combat. Individuals who can recite the DSM criteria or use psychiatric jargon such as “intrusive recollections” should be regarded with suspicion. Malingerers are more likely to thrust forward their symptoms and diagnosis by making broad, conclusory statements such as “I have PTSD.” They may also present with dramatic, attention-seeking behavior and overplay symptoms such as hyperarousal. The nightmares of those with combat-related PTSD often involve a specific traumatic event that occurs repeatedly as it did in reality, without variation (Atkinson et al., 1982). If they discuss

10. The Malingering of Posttraumatic Disorders 205

TABLE 10.7.  A Comparison of Genuine and Malingered PTSD in Combat Veterans

Salient features

Genuine PTSD

Malingered PTSD

Presentation for evaluation

Secondary to encouragement from others

Individual seeks out the opportunity

Attributing blame

To oneself

To those in a position of authority

Combat experience and current symptoms

Downplays them, reluctant to discuss

Emphasizes them

Affect

Withdrawn

Assertive

Theme of nightmares

Feeling helpless

Being in power and having control

Other

Avoids environmental cues reminiscent of the trauma (hot, dry weather)

Describes fear of “losing control” and hurting others

nightmares, malingerers are more likely to describe situations in which they were powerful and in control versus the feelings of overwhelming helplessness that are usual theme of posttraumatic dreams in combat veterans (Atkinson et al., 1982).

CONCLUSIONS Most clinicians would agree that malingering is difficult to identify, no matter what illness is feigned. It requires an assessment of whether false symptoms are intentionally produced and what the individual gains by having those symptoms. The diagnosis of PTSD is primarily based on a subjective report of psychological symptoms, which makes it an easy disorder to simulate successfully. Though it is a complex and lengthy process, meticulous collateral data collection and careful evaluation of the examinee allow for the differentiation between genuine and malingered PTSD. Psychological testing is very important in confirming suspected malingering. Clinicians charged with the task of evaluating PTSD must be well versed in the phenomenology of the disorder and know the crucial differences between genuine and malingered PTSD. In summary, in order to improve the likelihood of recognizing malingered PTSD, the evaluator should do the following: • Establish the individual’s motivation for malingering. • Collect collateral information from multiple sources prior to the evaluation. • Approach the assessment in an unbiased fashion.

• Ask open-ended questions, while being mindful not to reveal the criteria for PTSD. • Assess the veracity of the symptoms based on the phenomenology of legitimate PTSD. • Utilize psychological testing designed to detect malingering. • Clarify inconsistencies in the examinee’s account. • Provide the examinee the opportunity to acknowledge malingering without shame.

REFERENCES Adamou, M. C., & Hale, A. S. (2003). PTSD and the law of psychiatric injury in England and Wales: Finally coming closer? Journal of the American Academy of Psychiatry and the Law, 31(3), 327–332. Ali, S., Jabeen, S., & Alam, F. (2015). Multimodal approach to identifying malingered posttraumatic stress disorder: A review. Innovations in Clinical Neuroscience, 12(1–2), 12–20. American Psychiatric Association. (1952). Diagnostic and statistical manual of mental disorders. Washington, DC: Author. American Psychiatric Association. (1968). Diagnostic and statistical manual of mental disorders (2nd ed.). Washington, DC: Author. American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disorders (3rd ed.). Washington, DC: Author. American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: Author. American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: Author.

206

I I .   D i a g n o s t ic I ss u e s

Appelbaum, P., Jick, R., & Grisso, T. (1993). Use of posttraumatic stress disorder to support an insanity defense. American Journal of Psychiatry, 150(2), 229–234. Arbisi, P. A., Ben-Porath, Y. S., & McNulty, J. (2006). The ability of the MMPI-2 to detect feigned PTSD within the context of compensation seeking. Psychological Services, 3(4), 249–261. Arbisi, P. A., & Butcher, J. N. (2004). Failure of the FBS to predict malingering of somatic symptoms: Response to critiques by Greve and Bianchini and Lees Haley and Fox. Archives of Clinical Neuropsychology, 19(3), 341–345. Arbisi, P. A., Polusny, M. A., Erbes, C. R., Thuras, P., & Reddy, M. K. (2011). The Minnesota Multiphasic Personality Inventory–2 Restructured Form in National Guard soldiers screening positive for posttraumatic stress disorder and mild traumatic brain injury. Psychological Assessment, 23(1), 203–214. Atkinson, R. M., Henderson, R. G., Sparr, L. F., & Deale, S. (1982). Assessment of Vietnam veterans for post-traumatic stress disorder in Veterans Administration disability claims. American Journal of Psychiatry, 139, 1118–1121. Ben-Porath, Y. S. (2012). Interpreting the MMPI-2-RF. Minneapolis: University of Minnesota Press. Ben-Porath, Y. S., & Tellegen, A. (2008). MMPI-2-RF: Manual for administration, scoring and interpretation. Minneapolis: University of Minnesota Press. Berger, O., McNiel, D. E., & Binder, R. L. (2012). PTSD as a criminal defense: A review of case law. Journal of the American Academy of Psychiatry and the Law, 40(4), 509–521. Bitzer, R. (1980). Caught in the middle: Mentally disabled veterans and the Veterans Administration. In C. R. Figley & S. Leventman (Eds.), Strangers at home: Vietnam veterans since the war (pp.305–323). New York: Brunner/Mazel. Blanchard, E. B., Kolb, L. C., Pallmeyer, T. P., & Gerardi, R. J. (1982). A psychophysiological study of post-traumatic stress disorder in Vietnam veterans. Psychiatric Quarterly, 54, 220–229. Brandt, J. (1992). Detecting amnesia’s impostors. In L. R. Squire & N. Butters (Eds.), Neuropsychology of memory (2nd ed., pp.156–165). New York: Guilford Press. Braverman, M. (1978). Post-injury malingering is seldom a calculated ploy. Annals of the New York Academy of Sciences, 444, 502–503. Breslau, N., Davis, G. C., Andreski, P., & Peterson, E. (1991). Traumatic events and posttraumatic stress disorder in an urban population of young adults. Archives of General Psychiatry, 48(3), 216–222. Breslau, N., & Kessler, R. C. (2001). The stressor criterion in DSM-IV posttraumatic stress disorder: An empirical investigation. Biological Psychiatry, 50(9), 699–704. Briere, J. (1995). Trauma symptom inventory (TSI). Lutz, FL: Psychological Assessment Resources. Briere, J. (2001). Detailed Assessment of Posttraumatic

Stress professional manual. Lutz, FL: Psychological Assessment Resources. Briere, J. (2010). Trauma Symptom Inventory–2 (TSI-2) professional manual. Lutz, FL: Psychological Assessment Resources. Brown, T. A., Campbell, L. A., Lehman, C. L., Grisham, J. R., & Mancill, R. B. (2001). Current and lifetime comorbidity of the DSM-IV anxiety and mood disorders in a large clinical sample. Journal of Abnormal Psychology, 110(4), 585–599. Burchett, D. L., & Bagby, R. M. (2014). Multimethod assessment of response distortion: Integrating data from interviews, collateral records, and standardized assessment tools. In C. J. Hopwood & R. F. Bornstein (Eds.), Multimethod clinical assessment (pp.345–378). New York: Guilford Press. Burges, C., & McMillan, T. (2001). The ability of naïve participants to report symptoms of post-traumatic stress disorder. British Journal of Clinical Psychology, 40(2), 209–214. Burkett, B., & Whitley, G. (1998). Stolen valor: How the Vietnam generation was robbed of its heroes and history. Dallas, TX: Verity Press. Bury, A. S., & Bagby, R. M. (2002). The detection of feigned uncoached and coached posttraumatic stress disorder with the MMPI-2 in a sample of workplace accident victims. Psychological Assessment, 14(4), 472–484. Butcher, J. N. (2001). Minnesota Multiphasic Personality Inventory–2: Manual for administration, scoring, and interpretation. Minneapolis: University of Minnesota Press. Butcher, J. N., & Miller, K. B. (1999). Personality assessment in personal injury litigation. In A. K. Hess & I. B. Weiner (Eds.), The handbook of forensic psychology (pp.104–126). New York: Wiley. Calhoun, P. S., Collie, C. F., Clancy, C. P., Braxton, L. E., & Beckham, J. C. (2010). Use of the PAI in assessment of Posttraumatic Stress Disorder among help-seeking veterans. In M. A. Blais, M. R. Baity, & C. J. Hopwood (Eds.), Clinical applications of the Personality Assessment Inventory (pp.93–111). New York: Routledge. Calhoun, P. S., Earnst, K. S., Tucker, D. D., Kirby, A. C., & Beckham, J. C. (2000). Feigning combat-related posttraumatic stress disorder on the Personality Assessment Inventory. Journal of Personality Assessment, 75(2), 338–350. Calhoun, P. S., Hertzberg, J. S., Kirby, A. C., Dennis, M. F., Hair, L. P., Dedert, E. A., et al. (2012). The effect of draft DSM-5 criteria on posttraumatic stress disorder prevalence. Depression and Anxiety, 29(12), 1032–1042. Clevenger, S. V. (1889). Spinal concussion. London: F. A. Davis. Collie, J. (1971). Malingering and feigned sickness. London: Edward Arnold. Davidson, H. A. (1952). Forensic psychiatry (2nd ed.). New York: Ronald Press.

10. The Malingering of Posttraumatic Disorders 207

Davidson, J. (1993). Issues in the diagnosis of posttraumatic stress disorder. In J. M. Oldham, M. B. Riba, & A. Tasman (Eds.), Review of psychiatry. (pp.141–155). Washington, DC: American Psychiatric Press. de Carteret, J. C. (1994). Occupational stress claims: Effects on workers’ compensation. American Association of Occupational Health Nurses Journal, 42(10), 494–498. Demakis, G. J., & Elhai, J. D. (2011). Neuropsychological and psychological aspects of malingered posttraumatic stress disorder. Psychological Injury and Law, 4(1), 24–31. Department of Veterans Affairs Office of the Inspector General. (2005). Review of state variances in VA disability compensation payments (No. 05-00765-137). Retrieved May 9, 2016, from www.va.gov/oig/publications/efoia. Eakin, D. E., Weathers, F. W., Benson, T. B., Anderson, C. F., & Funderburk, B. (2006). Detection of feigned posttraumatic stress disorder: A comparison of the MMPI-2 and PAI. Journal of Psychopathology and Behavioral Assessment, 28(3), 145–155. Edens, J. F., Buffington, J. K., & Tomicic, T. L. (2000). An investigation of the relationship between psychopathic traits and malingering on the psychopathic personality inventory. Assessment, 7(3), 281–296. Edens, J. F., Guy, L. S., Otto, R. K., Buffington, J. K., Tomicic, T. L., & Poythress, N. G. (2001). Factors differentiating successful versus unsuccessful malingerers. Journal of Personality Assessment, 77(2), 333–338. Edens, J. F., Otto, R. K., & Dwyer, T. J. (1998). Susceptibility of the Trauma Symptom Inventory to malingering. Journal of Personality Assessment, 71(3), 379–392. Efendov, A. A., Sellbom, M., & Bagby, R. M. (2008). The utility and comparative incremental validity of the MMPI-2 and Trauma Symptom Inventory validity scales in the detection of feigned PTSD. Psychological Assessment, 20(4), 317–326. Elhai, J. D., Ford, J. D., Ruggiero, K. J., & Frueh, B. C. (2009). Diagnostic alterations for post-traumatic stress disorder: Examining data from the National Comorbidity Survey Replication and National Survey of Adolescents. Psychological Medicine, 39(12), 1957–1966. Elhai, J. D., Frueh, B. C., Gold, P. B., Gold, S. N., & Hamner, M. B. (2000). Clinical presentations of posttraumatic stress disorder across trauma populations: A comparison of MMPI-2 profiles of combat veterans and adult survivors of child sexual abuse. Journal of Nervous and Mental Disease, 188(10), 708–713. Elhai, J. D., Gold, S. N., Sellers, A. H., & Dorfman, W. I. (2001). The detection of malingered posttraumatic stress disorder with MMPI-2 fake bad indices. Assessment, 8(2), 221–236. Elhai, J. D., Gray, M. J., Kashdan, T. B., & Franklin, C. L. (2005). Which instruments are most commonly used to assess traumatic event exposure and posttrau-

matic effects?: A survey of traumatic stress specialists. Journal of Traumatic Stress, 18, 541–545. Elhai, J. D., Naifeh, J. A., Zucker, I. S., Gold, S. N., Deitsch, S. E., & Frueh, B. C. (2004). Discriminating malingered from genuine civilian posttraumatic stress disorder: A validation of three MMPI-2 infrequency scales (F, Fp, and Fptsd). Assessment, 11(2), 139–144. Elhai, J. D., Ruggiero, K. J., Frueh, B. C., Beckham, J. C., Gold, P. B., & Feldman, M. E. (2002). The Infrequency-Posttraumatic Stress Disorder scale (Fptsd) for the MMPI-2: Development and initial validation with veterans presenting with combat-related PTSD. Journal of Personality Assessment, 79, 531–549. Enelow, A. (1971). Malingering and delayed recovery from injury. In J. Leedy (Ed.), Compensation in psychiatric disability and rehabilitation (pp.42–46). Springfield, IL: Charles C Thomas. Fairbank, J. A., McCaffrey, R. J., & Keane, T. M. (1986). On simulating post-traumatic stress disorder [Letter to the Editor]. American Journal of Psychiatry, 143, 268–269. Foa, E. B., Riggs, D., & Gershuny, B. (1995). Arousal, numbing, and intrusion. American Journal of Psychiatry, 152, 116–120. Franklin, C. L., Repasky, S. A., Thompson, K. E., Shelton, S. A., & Uddo, M. (2002). Differentiating overreporting and extreme distress: MMPI-2 use with compensation-seeking veterans with PTSD. Journal of Personality Assessment, 79(2), 274–285. Friedman, M. J., Resick, P. A., Bryant, R. A., & Brewin, C. R. (2011). Considering PTSD for DSM-5. Depression and Anxiety, 28, 750–769. Frueh, B. C., Elhai, J. D., Grubaugh, A. L., Monnier, J., Kashdan, T. B., Sauvageot, J. A., et al. (2005). Documented combat exposure of US veterans seeking treatment for combat-related post-traumatic stress disorder. British Journal of Psychiatry, 186, 467–472. Garcia, H. A., Franklin, C. L., & Chambliss, J. (2010). Examining MMPI-2 F-family scales in PTSD-diagnosed veterans of Operation Enduring Freedom and Operation Iraqi Freedom. Psychological Trauma: Theory, Research, Practice, and Policy, 2(2), 126–129. Garfield, P. (1987). Nightmares in the sexually abused female teenager. Psychiatric Journal of the University of Ottawa, 12, 93–97. Geisser, M. E., Roth, R. S., Bachman, J. E., & Eckert, T. A. (1996). The relationship between symptoms of post-traumatic stress disorder and pain, affective disturbance and disability among patients with accident and non-accident related pain. Pain, 66(2–3), 207–214. Germain, A., & Nielsen, T. A. (2003). Sleep pathophysiology in posttraumatic stress disorder and idiopathic nightmare sufferers. Biological Psychiatry, 54(10), 1092–1098. Goodwin, B. E., Sellbom, M., & Arbisi, P. A. (2013). Posttraumatic stress disorder in veterans: The utility of the MMPI-2-RF validity scales in detecting over-

208

II .   D i a g n o s t i c I s s u e s

reported symptoms. Psychological Assessment, 25(3), 671–678. Gray, M. J., Elhai, J. D., & Briere, J. (2010). Evaluation of the Atypical Response scale of the Trauma Symptom Inventory-2 in detecting simulated posttraumatic stress disorder. Journal of Anxiety Disorders, 24(5), 447–451. Greene, R. L. (2000). The MMPI-2: An interpretive manual (2nd ed.). Needham Heights, MA: Allyn & Bacon. Grey, B. J. (2012). Neuroscience, PTSD and sentencing mitigation. Cardozo Law Review, 34, 54–105. Grossman, R., Buchsbaum, M. S., & Yehuda, R. (2002). Neuroimaging studies in post-traumatic stress disorder. Psychiatric Clinics of North America, 25(2), 317–340. Guriel, J., & Fremouw, W. (2003). Assessing malingering posttraumatic stress disorder: A critical review. Clinical Psychology Review, 23(7), 881–904. Guriel, J., Yañez, T., Fremouw, W., Shreve-Neiger, A., Ware, L., Filcheck, H., et al. (2004). Impact of coaching on malingered posttraumatic stress symptoms on the M-FAST and the TSI. Journal of Forensic Psychology Practice, 4(2), 37–56. Hall, R. C. W., & Chapman, M. J. (2005). Definition, diagnosis and forensic implications of postconcussional syndrome. Psychosomatics, 46(3), 195–202. Hamilton, J. D. (1985). Pseudo post-traumatic stress disorder. Military Medicine, 150, 353–356. Harvey, A. G., & Bryant, R. A. (2000). Memory for acute stress disorder and posttraumatic stress disorder: A 2-year prospective study. Journal of Nervous and Mental Disease, 188(9), 602–607. Hausotter, W. (1996). Expert assessment of railroad accidents at the beginning of the industrial age—a medical historical excursion with reference to current circ*mstances. Versicherungsmedizin, 48(4), 138–142. Hofling, C. K. (1965). Some psychological aspects of malingering. General Practitioner, 31, 115–121. Hollander, E., & Simeon, D. (2002). Anxiety disorders. In R. E. Hales & S. C. Yudofsky (Eds.), Textbook of clinical psychiatry (4th ed., pp.543–630). Washington, DC: American Psychiatric Press. Inabu, F. E., & Reid, J. E. (1967). Criminal interrogation and confession (2nd ed.). Baltimore: Williams & Wilkins. Inman, D. J., Silver, S. M., & Doghramji, K. (1990). Sleep disturbance in post-traumatic stress disorder: A comparison with non-PTSD insomnia. Journal of Traumatic Stress, 3, 429–437. Jordon, R. G., Nunley, T. V., & Cook, R. R. (1992). Symptom exaggeration in PTSD inpatient population: Response set or claim for compensation. Journal of Traumatic Stress, 4, 633–642. Kaplan, B. J., & Sadock, V. A. (Eds.). (2003). Synopsis of psychiatry (9th ed.). Philadelphia: Lippincott, Williams & Wilkins. Keane, T. M., Kolb, L. C., Kaloupek, D. G., Orr, S. P.,

Blanchard, E. B., Thomas, R. G., et al. (1998). Utility of psychophysiological measurement in the diagnosis of posttraumatic stress disorder: Results from a Department of Veteran Affairs Cooperative Study. Journal of Consulting and Clinical Psychology, 66(6), 914–923. Keiser, L. (1968). The traumatic neurosis. Philadelphia: Lippincott. Kelly, R., & Smith, B. N. (1981). Post traumatic syndrome: Another myth discredited. Journal of the Royal Society of Medicine, 74(4), 275–278. Kessler, R. C. (2000). Posttraumatic stress disorder: The burden to the individual and to society. Journal of Clinical Psychiatry, 61(Suppl. 5), 4–12. Kilpatrick, D. G., Resnick, H. S., Milanak, M. E., Miller, M. W., Keyes, K. M., & Friedman, M. J. (2013). National estimates of exposure to traumatic events and PTSD prevalence using DSM-IV and DSM-5 criteria. Journal of Traumatic Stress, 26(5), 537–547. Knoll, J. L., Resnick, P. J. (1999). U.S. v. Greer: Longer sentences for malingerers. Journal of the American Academy of Psychiatry and the Law, 27(4), 621–625. Koch, W., O’Neill, M., & Douglas, K. S. (2005). Empirical limits for the forensic assessment of PTSD litigants. Law and Human Behavior, 29(1), 121–149. Kropp, P. R., & Rogers, R. (1993). Understanding malingering: Motivations, method and detection. In M. Lewis & C. Saarini (Eds.), Lying and deception in everyday life (pp.201–216). New York: Guilford Press. Kuch, K., & Cox, B. J. (1992). Symptoms of PTSD in 124 survivors of the Holocaust. American Journal of Psychiatry, 149(3), 337–340. Kucharski, L. T., Toomey, J. P., Fila, K., & Duncan, S. (2007). Detection of malingering of psychiatric disorder with the Personality Assessment Inventory: An investigation of criminal defendants. Journal of Personality Assessment, 88(1), 25–32. Lang, P. J. (1985). The cognitive psychophysiology of emotion: Fear and anxiety. In A. H. Tuma & J. Maser (Eds.), Anxiety and the anxiety disorder (pp.131–170). Hillsdale, NJ: Erlbaum. Layden, M. (1966). Symptoms separate hysteric, malingerer. Psychiatric Progress, 1, 7. Lees-Haley, P. R. (1992). Efficacy of MMPI-2 validity scales and MCMI-II modifier scales for detecting spurious PTSD claims: F, F-K, fake bad scale, ego strength, subtle-obvious subscales, DIS, and DEB. Journal of Clinical Psychology, 48(5), 681–688. Lees-Haley, P. R., & Dunn, J. T. (1994). The ability of naive subjects to report symptoms of mild brain injury, post-traumatic stress disorder, major depression and generalized anxiety disorder. Journal of Clinical Psychology, 50(2), 252–256. Levin, A. P., Kleinman, S. B., & Adler, J. S. (2014). DSM-5 and posttraumatic stress disorder. Journal of the American Academy of Psychiatry and the Law, 42, 146–158. Levin, H. S., Lilly, M. A., Papanicolau, A., & Eisenberg, H. M. (1992). Posttraumatic and retrograde amnesia

10. The Malingering of Posttraumatic Disorders 209

after closed head injury. In L. R. Squire & N. Butters (Eds.), Neuropsychology of memory (2nd ed., pp.290– 308). New York: Guilford Press. Liljequist, L., Kinder, B. N., & Schinka, J. A. (1998). An investigation of malingering posttraumatic stress disorder in the Personality Assessment Inventory. Journal of Personality Assessment, 71(3), 322–336. Lipman, F. D. (1962). Malingering in personal injury cases. Temple Law Quarterly, 35, 141–162. Loftus, E. F. (1979). The malleability of human memory: Information introduced after we view an incident can transform memory. American Scientist, 67(3), 312–320. Lynn, E. J., & Belza, M. (1984). Factitious posttraumatic stress disorder: The veteran who never got to Vietnam. Hospital and Community Psychiatry, 35, 697–701. Marion, B. E., Sellbom, M., & Bagby, R. M. (2011). The detection of feigned psychiatric disorders using the MMPI-2-RF overreporting validity scales: An analog investigation. Psychological Injury and Law, 4(1), 1–12. Marshall, M. B., & Bagby, R. M. (2006). The incremental validity and clinical utility of the MMPI-2 Infrequency Posttraumatic Stress Disorder Scale. Assessment, 13(4), 417–429. Mason, L. H., Shandera-Ochsner, A. L., Williamson, K. D., Harp, J. P., Edmundson, M., Berry, D. T., et al. (2013). Accuracy of MMPI-2-RF validity scales for identifying feigned PTSD symptoms, random responding, and genuine PTSD. Journal of Personality Assessment, 95(6), 585–593. McAllister, T. W. (1994). Mild traumatic brain injury and the postconcussive syndrome. In J. M. Silver, S. C. Yudofsky, & R. E. Hales (Eds.), Neuropsychiatry of traumatic brain injury (pp.357–392). Washington, DC: American Psychiatric Press. McEwen, B. S. (1997). Possible mechanisms for atrophy of the human hippocampus. Molecular Psychiatry, 2(3), 255–262. McNally, R. J. (2006). Applying biological data in the forensic and policy arenas. Annals of the New York Academy of Sciences, 1071, 267–276. McNally, R. J. (2009). Can we fix PTSD on DSM-V? Depression and Anxiety, 26, 597–600. Melton, R. (1984). Differential diagnosis: A common sense guide to psychological assessment. Vet Center Voice Newsletter, 1, 1–12. Mendelson, G. (1981). Persistent work disability following settlement of compensation claims. Law Institute Journal (Melbourne), 55(6), 342–345. Miller, H. (1961). Accident neurosis. British Medical Journal, 1, 919–925. Mittenberg, W., Patton, C., Canyock, E. M., & Condit, D. C. (2002). Base rates of malingering and symptom exaggeration. Journal of Clinical and Experimental Neuropsychology, 24(8), 1094–1102. Modlin, H. (1960). Neurosis and trauma, roundtable presentation. Washington, DC: American Psychiatric Association Press.

Mogge, N. L., Lepage, J. S., Bell, T., & Ragatz, L. (2010). The negative distortion scale: A new PAI validity scale. Journal of Forensic Psychiatry and Psychology, 21(1), 77–90. Morey, L. C. (1996). An Interpretive Guide to the Personality Assessment Inventory (PAI). Odessa, FL: Psychological Assessment Resources. Morey, L. C. (2007). The Personality Assessment Inventory professional manual (2nd ed.). Odessa, FL: Psychological Assessment Resources. Morson, R., Davis, K. D., & Weaver, C. M. (2016, August). Malingered PTSD detection: Comparing a new PTSD measure and validity scales of the MMPI-2. Paper presented at the Annual Conference of the American Psychology Association, Denver, CO. Moskowitz, A. (2004). Dissociation and violence: A review of the literature. Trauma, Violence, and Abuse, 5(1), 21–46. Mossman, D. (1994). At the VA, it pays to be sick. Public Interest, 114, 35–47. Mozley, S. L., Miller, M. W., Weathers, F. W., Beckham, J. C., & Feldman, M. E. (2005). Personality Assessment Inventory (PAI) profiles of male veterans with combat-related posttraumatic stress disorder. Journal of Psychopathology and Behavioral Assessment, 27(3), 179–189. Munley, P. H., Bains, D. S., Bloem, W. D., & Busby, R. M. (1995). Post-traumatic stress disorder and theMMPI-2. Journal of Traumatic Stress, 8(1), 171– 178. Nichols, D. S. (2011). Essentials of MMPI-2 assessment, 2nd edition. Hoboken, NJ: Wiley. O’Donnell, M. L., Creamer, M., Elliott, P., & Bryant, R. (2007). Tonic and phasic heart rate as predictors of posttraumatic stress disorder. Psychosomatic Medicine, 69(3), 256–261. Olson, W. (1991). The litigation explosion. New York: Dutton. Ossipov, P. V. (1994). Malingering: The simulation of psychosis. Bulletin of the Menninger Clinic, 8, 31–42. Otis, J. D., Pincus, D. B., & Keane, T. M. (2006). Comorbid chronic pain and posttraumatic stress disorder across the lifespan: A review of theoretical models. In G. Young, A. W. Kane, & K. Nicholson (Eds.), Psychological knowledge in court: PTSD, pain, and TBI (pp.242–268). New York: Springer. Ozer, E. J., Best, S. R., Lipsey, T. L., & Weiss, D. S. (2003). Predictors of posttraumatic stress disorder and symptoms in adults: A meta-analysis. Psychological Bulletin, 129(1), 52–73. Pankratz, L. (1985, May). The spectrum of factitious posttraumatic stress disorder. Presentation to the American Psychiatric Association, Dallas, TX. Parker, N. (1977). Accident litigants with neurotic symptoms. Medical Journal of Australia, 2(10), 318– 322. Pitman, R. K., Saunders, L. S., & Orr, S. P. (1994, April). Psychophysiological testing for posttraumatic stress disorder. Trial, pp.22–26.

210

II.  Diagnos t ic Issues

Pitman, R., & Sparr, L. (1998). PTSD and the law. PTSD Research Quarterly, 9(2), 1–6. Pitman, R. K., Sparr, L. F., Saunders, L. S., & McFarlane, A. C. (1996). Legal issues in posttraumatic stress disorder. In B. A. van der Kolk, A. C. McFarlane, & L. Weisaeth (Eds.), Traumatic stress: The effects of overwhelming experience on mind, body, and society (pp.378–397). New York: Guilford Press. Pole, N. (2006). Moderators of PTSD-related psychophysiological effect sizes: Results from a meta-analysis. Annals of the New York Academy of Sciences, 1071(1), 422–424. Pollack, S. (1982). Dimensions of malingering. In B. H. Gross & L. E. Weinburger (Eds.), New directions for mental health services: The mental health professional and the legal system (pp.63–75). San Francisco: Jossey-Bass. Porter v. McCollum, 558 U.S. 30 (2009). Powell, K. E. (1991). The malingering of schizophrenia. Unpublished doctoral dissertation, University of Southern California, Los Angeles, CA. Resnick, P. J. (1997). Malingering of posttraumatic disorders. In R. Rogers (Ed.), Clinical assessment of malingering and deception (2nd ed., pp.84–103). New York: Guilford Press. Resnick, P. J. (2004). Guidelines for evaluating malingering in PTSD. In R. I. Simon (Ed.), Posttraumatic stress disorder in litigation: Guidelines for forensic assessment (pp.187–205). Washington, DC: American Psychiatric Publishing. Rickarby, G. A. (1979). Compensation-neurosis and the psycho-social requirements of the family. British Journal of Medical Psychology, 52(4), 333–338. Rigler, J. (1879). Uber die Folgen der Verletzungen auf Eisenbahnen [Concerning the consequences of injuries occurring on railroads]. Berlin: Reimer. Rogers, R. (2008). Detection strategies in malingering and defensiveness. In R. Rogers (Ed.), Clinical assessment of malingering and deception (3rd ed., pp.14–38). New York: Guilford Press. Rogers, R., & Bender, S. D. (2013). Evaluation of malingering and deception. In I. B. Weiner & R. K. Otto (Eds.), Handbook of psychology: Vol. 11. Forensic psychology (2nd ed., pp.517–540). Hoboken, NJ: Wiley. Rogers, R., Gillard, N. D., Wooley, C. N., & Kelsey, K. (2013). Cross-validation of the PAI Negative Distortion Scale for feigned mental disorders: A research report. Assessment, 20(1), 36–42. Rogers, R., Gillard, N. D., Wooley, C. N., & Ross, C. A. (2012). The detection of feigned disabilities: The effectiveness of the PAI in a traumatized inpatient sample. Assessment, 19(1), 77–88. Rogers, R., & Granacher, R. P., Jr. (2011). Conceptualization and assessment of malingering. In E. Y. Drogin, R. M. Dattilio, R. L. Sadoff, & T. G. Gutheil (Eds.), Handbook of forensic assessment: Psychological and psychiatric perspectives (pp.659–678). Hoboken, NJ: Wiley.

Rogers, R., Payne, J. W., Correa, A. A., Gillard, N. G., & Ross, C. A. (2009). A study of the SIRS with severely traumatized patients. Journal of Personality Assessment, 91(5), 429–438. Rogers, R., Sewell, K. W., Martin, M. A., & Vitacco, M. J. (2003). Detection of feigned mental disorders: A meta-analysis of the MMPI-2 and malingering. Assessment, 10(2), 160–177. Rosen, G. M. (1995). The Aleutian Enterprise sinking and posttraumatic stress disorder: Misdiagnosing in clinical and forensic settings. Professional Psychology: Research and Practice, 26(1), 82–87. Rosen, G. M. (2006). DSM’s cautionary guideline to rule out malingering can protect the PTSD data base. Journal of Anxiety Disorders, 20(4), 530–535. Rosen, G. M., Sawchuk, C. N., Atkins, D. C., Brown, M., Price, J. R., & Lees-Haley, P. R. (2006). Risk of false positives when identifying malingered profiles using the Trauma Symptom Inventory. Journal of Personality Assessment, 86(3), 329–333. Scragg, P., Bor, R., & Mendham, M. C. (2000). Feigning post-traumatic stress disorder on the PAI. Clinical Psychology and Psychotherapy, 7(2), 155–160. Sellbom, M., & Bagby, R. M. (2010). Detection of overreported psychopathology with the MMPI-2 RF form validity scales. Psychological Assessment, 22(4), 757–767. Sherman, J. J., Turk, D. C., & Okifuji, A. (2000). Prevalence and impact of posttraumatic stress disorder-like symptoms on patients with fibromyalgia syndrome. Clinical Journal of Pain, 16(2), 127–134. Slovenko, R. (1994). Legal aspects of post-traumatic stress disorder. Psychiatric Clinics of North America, 17, 439–446. Smith, B. H. (1967). A handbook of tests to unmask malingering. Consultant, 7, 41–47. Sparr, L. (1990). Legal aspects of posttraumatic stress disorder: Uses and abuses. In M. Wolf & A. Mosnaim (Eds.), Posttraumatic stress disorder: Etiology, phenomenology and treatment (pp.238–264). Washington, DC: American Psychiatric Press. Sparr, L. F., & Atkinson, R. M. (1986). Posttraumatic stress disorder as an insanity defense: Medicolegal quicksand. American Journal of Psychiatry, 143, 608–613. Sparr, L. F., & Pitman, R. K. (2007). PTSD and the law. In M. J. Frieman, T. M. Keane, & P. A. Resick (Eds.), Handbook of PTSD: Science and practice (pp.449– 468). New York: Guilford Press. Stone, D. C., Boone, K. B., Back-Madruga, C., & Lesser, I. M. (2006). Has the rolling uterus finally gathered moss?: Somatization and malingering of cognitive deficit in six cases of “toxic mold” exposure. Clinical Neuropsychology, 20(4), 766–785. Suendermann, O., Ehlers, A., Boellinghaus, I., Gamer, M., & Glucksman, E. (2010). Early heart rate responses to standardized trauma-related pictures predict posttraumatic stress disorder: A prospective study. Psychosomatic Medicine, 72(3), 301–308.

10. The Malingering of Posttraumatic Disorders 211

Tellegen, A., & Ben-Porath, Y. S. (2011). MMPI-2-RF, Minnesota Multiphasic Personality Inventory–2 Restructured Form: Technical manual. Minneapolis: University of Minnesota Press. Thomann, K. D., & Rauschmann, M. (2003). “Posttraumatic stress disorder”—historical aspects of a “modern” psychiatric illness in the German language areas. Medizinhistorisches Journal, 38(2), 103–138. Thomann, K., & Rauschmann, M. (2004). Whiplash injury and “railway spine.” Versicherungsmedizin, 56(3), 131–135. Thomas, K. M., Hopwood, C. J., Orlando, M. J., Weathers, F. W., & McDevitt-Murphy, M. E. (2012). Detecting feigned PTSD using the personality assessment inventory. Psychological Injury and Law, 5(3–4), 192–201. Thompson, G. N. (1965). Post-traumatic psychoneurosis—a statistical survey. American Journal of Psychiatry, 121(11), 1043–1048. Trimble, M. (1981). Post-traumatic neurosis from railway spine to whiplash. New York: Wiley. van der Kolk, B. A. (1984). Post-traumatic stress disorder: Psychological and biological sequelae. Washington, DC: American Psychiatric Association Press. van der Kolk, B., Blitz, R., Burr, W., Sherry, S., & Hartmann, E. (1984). Nightmares and trauma: A comparison of nightmares after combat with lifelong nightmares in veterans. American Journal of Psychiatry, 141, 187–190. Wangelin, B. C., & Tuerk, P. W. (2015). Taking the pulse of prolonged exposure therapy: Physiological reactivity to trauma imagery as an objective measure of treatment response. Depression and Anxiety, 32(12), 927–934.

Weaver, C. M., & Yano, K. (2015, March). Development and introduction of a new measure of PTSD malingering. Symposium presented at the annual conference of the American Psychology–Law Society, San Diego, CA. Wetter, M. W., Baer, R. A., Berry, D. T., Robison, L. H., & Sumpter, J. (1993). MMPI-2 profiles of motivated fakers given specific symptom information: A comparison to matched patients. Psychological Assessment, 5(3), 317–323. Wetter, M. W., & Deitsch, S. E. (1996). Faking specific disorders and temporal response consistency on the MMPI-2. Psychological Assessment, 8(1), 39–47. Wilson, J. P. (1981). Cognitive control mechanisms in stress response syndromes and their relation to different forms of the disorder. Unpublished manuscript, Cleveland State University, Cleveland, OH. Wittman, L., Schredl, M., & Kramer, M. (2007). Dreaming in posttraumatic stress disorder: A critical review of phenomenology, psychophysiology and treatment. Psychotherapy and Psychosomatics, 76(1), 25–39. Wooley, C. N., & Rogers, R. (2015). The effectiveness of the Personality Assessment Inventory with feigned PTSD: An initial investigation of Resnick’s model of malingering. Assessment, 22(4), 449–458. Yehuda, R., Schmeidler, J., Wainberg, M., BinderBrynes, K., & Duvdevani, T. (1998). Vulnerability to posttraumatic stress disorder in adult offspring of Holocaust survivors. American Journal of Psychiatry, 155(9), 1163–1171. Yutzy, S. (2002). Somatoform disorders. In R. E. Hales & S. C. Yudofsky (Eds.), Textbook of clinical psychiatry (4th ed.). Washington, DC: American Psychiatric Press.

C H A P T E R 11

Factitious Disorders inMedical andPsychiatric Practices GregoryP.Yates,MA MazheruddinM.Mulla,MA,MPH JamesC.Hamilton,PhD MarcD.Feldman,MD

The scholarly study of medical and psychiatric deception is rooted in hundreds of carefully documented cases. These reports provide convincing evidence that some patients intentionally exaggerate, lie about, simulate, aggravate, or induce illness, injury, or impairment, either in themselves or in persons under their care (Yates & Feldman, 2016; Sheridan, 2003). The existence of such cases has been widely accepted for over a century, and these clinical phenomena have been codified as psychological disorder since DSM-III (American Psychiatric Association, 1980). Indeed, most clinicians have stories to tell about “Munchausen’s syndrome,” “frequent flyers,” “peregrinating problem patients,” and “hospital hoboes” encountered in their practice. Such cases of intentional feigning are organized into factitious disorder (FD) and malingering. Malingering is not a disorder per se, but rather a condition that may be of clinical concern. Patients with FD deliberately falsify symptoms associated with physical or psychological illness, injury, or impairment in themselves or others (American Psychiatric Association, 2013). The terms factitious disorder imposed on self (FDIOS) Allan Cunnien, MD, now deceased, made valuable contributions to the first two editions; out of respect, his name was continued on the third edition.

and factitious disorder imposed on another (FDIOA) distinguish the two types. Patients with FD may engage in a variety of deceptive behaviors in order to exaggerate or entirely feign the appearance of a medical problem. Common forms of deceptive illness behavior may include exaggerating the severity of a genuine medical problem, falsely reporting symptoms, acting as if symptoms are present when they are not, or interfering with medical tests or test results. For example, individuals may (1) falsify imaging results to indicate the presence of a tumor, (2) describe thoughts of suicidality following a divorce when they have never in fact been married, or (3) tamper with blood or urine samples to simulate evidence of a disease. In some cases, individuals may actually induce illness, injury, or impairment in themselves or another through behaviors such as injection of a harmful substance or mutilation of the skin with sharp instruments or corrosive chemicals. Little empirical research exists on FD despite considerable professional interest in the disorder. The vast majority of published articles are case reports, which are prone to publication bias and should not be taken as a substitute for empirical research (Yates & Feldman, 2016). Nonetheless, case studies are an important source of clinical guidance (Jenicek, 2001). Case studies, in aggregate, have provided the basis for

212

11.  Factitious Disorders in Medical and Psychiatric Practices 213

key recommendations for the detection (Kenedi et al., 2011; Kinns, Housley, & Freedman, 2013) and management (Eastwood & Bisson, 2008) of FD, as well as a vital evidence base for this chapter.

DIAGNOSTIC CLASSIFICATION OFFD Modern scientific recognition of FD is thought to have begun with Asher’s (1951) introduction of the term Munchausen’s syndrome to describe a chronic pattern of feigned illness behavior, although selfinflicted dermatoses were acknowledged as early as 1948 in international diagnostic systems (World Health Organization, 1948). Asher (1951) coined the term to describe what is now thought to be a rare, archetypal, and untreatable form of FD. The lifestyles of these patients revolved around hospitalizations, surgeries, and contentious battles with physicians. Subsequently, Meadow (1977) coined the term Munchausen syndrome by proxy to describe a parallel condition in which parents or other caretakers falsify health information or produce factitious disease in children, principally to garner emotional satisfaction. FD was first included as a formal diagnostic category in the third edition of the Diagnostic and Statistical Manual of Mental Disorders (DSMIII; American Psychiatric Association, 1980). It discussed deliberately feigned symptoms (Criterion A) and sick-role motivation (Criterion B) in the absence of external incentives (Criterion C). These criteria were maintained until the release of DSM-5 (American Psychiatric Association, 2013), as were the initial subtypes: predominantly psychological signs and symptoms, predominantly physical signs and symptoms, combined psychological and physical symptoms, and FD not otherwise specified. DSM-5 includes several notable changes in FD classification. FD no longer has its own chapter but is instead grouped under somatic symptom and related disorders (American Psychiatric Association, 2013). The rationale for this change is that somatic symptoms are typically the predominant feature in FD, and the preponderance of cases is encountered in medical rather than psychiatric settings. The subtypes of FD listed previously have been replaced by two new subtypes: “FD imposed on self” and “FD imposed on another” (American Psychiatric Association, 2013). These changes also attempted to address the blanket use of the term Munchausen’s syndrome to describe these cases.

The criteria for FD have also changed in DSM5, whereas the key diagnostic features remain largely the same. In particular, motivation to assume the sick role (in oneself or through another) has been removed as a requirement for DSM-5 diagnosis (American Psychiatric Association, 2013). The revised criteria now include (1) feigning (e.g., symptoms and behaviors) or the covert production of illness; (2) pretending to be ill, impaired, or injured; and (3) deception that occurs without apparent external motivation. It also includes an exclusion criterion regarding other disorders. These changes were intended to shift the focus of diagnostic assessment away from drawing inferences about underlying motivation. Instead, the revised criteria place greater reliance on more objective measures, such as the identification of intentional deception. Evidence of intentional deception is used in DSM-5 as the basis for distinguishing FD from behaviorally similar problems. For example, somatic symptom disorders and functional somatic syndromes are presumed to lack this intentionally deceptive quality. In these cases, medically unexplained illness behavior is regarded as the unintended result of faulty automatic cognitive processes. Rather than being distinct categories, some have argued that all these conditions exist along a continuum of consciousness or voluntariness (Nadelson, 1979). The deceptive illness behavior in FD is thought to be the product of conscious motivation to achieve a particular goal. Cases in which this behavior is attributed to external goals are classified as malingering, whereas the diagnosis of FD is given if the deception is enacted principally for emotional gratification. In practice, most cases of excessive illness behavior involve the presence of both external rewards and emotional gratification. The classification of malingering is addressed extensively by Rogers (Chapter 1) and in subsequent chapters of this volume. Regardless, its conceptual relation to FD is a pervasive theme in this chapter.

FDIOS Factitious MedicalDisorders Medical patients with FDIOS may appear in primary care settings, emergency departments, or any number of secondary care specialties. They may even present in palliative care settings (Weis, Gully, & Marks, 2016). The clinical presentation of factitious medical disorders is highly variable.

214

I I .   D i a g n o s t ic I ss u e s

By definition, medical falsification can come in multiple forms: (1) exaggerated symptoms or medical history, (2) outright lies about symptoms or medical history, (3) simulations of medical illnesses through the production of compelling signs or symptoms, (4) manipulations to prolong or exacerbate an existing illness, or (5) actual selfinduction of disease (American Psychiatric Association, 2013). Rarely does a factitious medical patient use only one method of deception. Rather, typical patients with FDIOS employ multiple forms of deception, such as a simulated sign of disease (e.g., rashes), feigned symptoms (e.g., joint pain), and a false history (e.g., tick bites) to create the full medical picture of a known disease (e.g., Lyme disease). Medical patients with FDIOS may be encountered, for example, in (1) dermatology (Fliege et al., 2007), with rashes (Levitz & Tan, 1988), burns (Maurice, Rivers, Jones, & Cronin, 1987), infections (Farrier & Mansel, 2002), or nonhealing wounds (McEwen, 1998); (2) endocrinology, with dysregulation of thyroid hormones (Ashawesh, Murthy, & Fiad, 2010), blood sugar (Alinejad & Oettel, 2011), or insulin (Gordon & Sansone, 2013); and (3) neurology (Bauer & Boegner, 1996), with seizures (Romano, Alqahtani, Griffith, & Koubeissi, 2014), weakness (Papa­ dopoulos & Bell, 1999), sensory deficits (Barnett, Vieregge, & Jantschek, 1990), or paralysis (Feldman & Duval, 1997). In general practice, bleeding problems, such as coughing up blood (Kokturk, Ekim, Aslan, Kanbay, & Acar, 2006), blood in the urine (Lazarus & Kozinn, 1991), and anemia (Hirayama et al., 2003) are often encountered, as are infections, including sepsis (Lazarus & Kozinn, 1991). The voluminous case literature on factitious medical disorders suggests that almost any medical problem can be falsified, although patients with FDIOS may particularly favor endocrinology, cardiology, and dermatology services (Yates & Feldman, 2016). In the age of the Internet, patients with FDIOS are now able to research complex diagnoses, learn how to forge laboratory reports, and even order pharmaceutical drugs online. Because the diagnosis of FDIOS relies on conclusive evidence of intentional medical deception (American Psychiatric Association, 2013), the disorder is more likely to be uncovered in cases that include simulation or self-induction of medical signs. Common examples observed by staff members include patients who are caught tampering with blood (Kurtz, Harrington, Matthews, & Nabarro, 1979) or urine samples (Nasser, Israelit, Muhammad, &

Basis, 2009). Other evidence includes possession of syringes (Tausche et al., 2004), drugs (Saiyasombat & Satyanarayan, 2012), or mechanical devices (Martins, Vieira, & de Fátima Ávila, 2005) often used for self-mutilation (Feily et al., 2009). Cases of FDIOS that involve only false histories and feigned or exaggerated symptoms are likely to go undiscovered.

Factitious PsychologicalDisorders Patients with FDIOS may also feign or produce psychiatric and/or behavioral ailments. Surprisingly, this category is gaining in complexity as developments in neuropsychiatry continue to blur the lines between central nervous system dysfunction and features that are primarily “psychological” or “emotional” in origin. As Parker (1996, p.38) wrote, “The distinction of psychological from physical is becoming increasingly artificial in the face of neurophysiologic advances pointing to the interplay between psychology and neurobiology.” Thus, on the one hand, psychological disorders are gaining in medical legitimacy; on the other hand, few definitive tests exist to establish the presence or absence of psychological disorders. These features combine to increase the appeal of feigned psychological disorders for those seeking to enact the sick role. Carney and Brown (1983) estimated that psychiatric presentations of FDIOS may account for up to 40% of all factitious presentations, even though many cases are probably missed due to inherent difficulties in objective assessment (Popli, Masand, & Dewan, 1992). Examples of particularly difficult factitious presentations include the following: (1) alcohol abuse (Caradoc-Davies, 1988; Mitchell & Francis, 2003), (2) hallucinations (Bahali & Ipek, 2013; Gregory & Jindal, 2006; Yildiz & Torun, 2003), (3) suicidal or homicidal ideation (Gregory & Jindal, 2006; Thompson & Beckson, 2004), and (4) pain disorders (Mailis-Gagnon, Nicholson, Blumberger, & Zurowski, 2008; Callegari, Bortolaso, & Vender, 2006). Rare or obscure psychological disorders may be simulated, including Stockholm syndrome (Spuijbroek, Blom, Braam, & Kahn, 2012) and dissociative identity disorder (Feldman, Rosenquist, & Bond, 1997; Friedl & Draijer, 2000; Thomas, 1997; for a useful review, see Boysen & VanBergen, 2014); cult brainwashing (Coons & Grier, 1990); and various paraphilias including pedophilia (Porter & Feldman, 2011), zoophilia, and exhibitionism (Hanson, Berlin, Malin, Fedoroff, & McGuire, 1992).

11.  Factitious Disorders in Medical and Psychiatric Practices 215

Factitious psychological symptoms are often more pronounced when the physician and hospital staff members are present. Discrepancies tend to arise between what patients describe and their actual appearance or behavior. The classic findings of a well-defined mental disorder are unlikely in factitious psychological disorders. Typically, patients’ symptoms will represent their stereotyped understanding of mental illness (e.g., exaggerated pangs of sorrow rather than emotional blunting in feigned major depression). Likewise, medications indicated for the factitious condition may appear inexplicably ineffective. Patients with factitious psychological disorders are likely to be unusually receptive to psychiatric hospitalization. This may distinguish them from patients with factitious physical disorders, who are likely to recoil at the suggestion that their problems may be psychological. Parker (1996, p.41) explained that a “common pattern for the patient with FD and physical symptoms is to leave the hospital against medical advice when referred to the psychiatric unit; the patient with factitious psychological symptoms, however, seeks hospitalization on the psychiatric unit.” She also noted that “if the patient is willing to provide valid psychological data, traits of psychopathic deviation, paranoia, hysteria, depression, and hypochondriasis may be present” (p.41). However, this conclusion is based on very small numbers of patients who may have had sufficient expertise to avoid detection by the validity indicators on the psychodiagnostic tests. Other authors (e.g.,