How to Obtain Third Party Information

bundet

80 Third Party Information in Forensic Assessment day-to-day contact with an individual or law enforcement ofﬁcers who have conducted thorough investigations of a cer- tain crime series may prove to be more relevant in some cases. For example, orderlies, cafeteria staff, physical thera- pists, home help assistants, and others in similar roles may have important observations, particularly in cases of malin- gered psychosis and exaggerated claims of psychic distress and impairment. In other instances, skilled police investiga- tors may describe commonalties across a series of offenses that is clearly inconsistent with an impulsive, unplanned offense. The importance of any professional source thus arises from the collateral’s proximity to certain behavior rather than the source’s professional status. TPI personal documentation in Table 5.3, such as state- ments made by litigants, witnesses, or alleged victims, has the advantage of already existing in written form and, in many instances, being available to both sides through recip- rocal discovery. Often, however, such documents provide only limited information relevant to the questions being assessed by the forensic clinician. A review of all available documents prior to scheduling collateral interviews can be advantageous for several reasons, therefore. First, the infor- mation available in collateral documents can provide a con- text for the interviews and help shape the questioning. Second, and more speciﬁcally, when a collateral interviewee has difﬁculty recalling an event, the forensic clinician can use third party documents to provide details that might help to facilitate such recall. Of course, the forensic clinician doing this must be extremely careful to avoid providing information that might affect the nature of the interviewee’s description of “sensitive” information; generally, such memory prompts should be entirely limited to nonsensitive details such as date, time, location, and the like. (Sensitive information refers to information that is directly relevant to the forensic capacities being evaluated, and usually includes thoughts, feelings, be- havior, and skills. Nonsensitive information can typically be distinguished when there is a focus on a particular event or time period; nonsensitive details in such cases include date, time, location, and activities unrelated to the legally relevant events or forensic capacities being assessed.) Third, when in- formation provided by collateral informants is not consistent with that contained in third party records, the forensic clini- cian can attempt to clarify the reasons for such inconsistency. It is particularly important to determine whether such incon- sistency seems to result from recall problems or bias. In certain cases, it is important to review personal docu- mentation that has been created by the defendant or plaintiff. Personal diaries can contain information about events that can be highly relevant to criminal adjudication or civil litiga- tion. Collections have been found to be of central importance in the investigation and evaluations of certain types of repet- itive sex offenders (Warren, Hazelwood, & Dietz, 1996). In cases of alleged serial murder, videotapes, photographs, and pornographic drawings have been located by police investi- gators and can be used by the forensic evaluator to assess sex- ual preferences, relevant interactions between coperpetrators, and commonalities in the preparation for and perpetration of particular crimes. Letters between spouses or romantic part- ners can be important in determining the nature of the rela- tionship and any particular events that may have preceded the violent behavior. Moreover, ofﬁce notations or reading mate- rial can be relevant in assessing the risk and needs of particu- lar individuals in a workplace violence context. Professional documentation can be considered on two levels. As sources of behavioral observations, such docu- ments can be quite valuable, particularly when they are de- tailed in their description of relevant behavior. However, professional documents sometimes reﬂect the conclusions of the writer in the very areas being evaluated in the present FMHA. Forensic clinicians are responsible for drawing their own conclusions and should not be overly inﬂuenced by con- clusions drawn by other professionals. Unless there is some reason to regard observations by other professionals as inac- curate, it is reasonable to accept such observations. However, conclusions (such as those regarding diagnosis or speciﬁc forensic capacities) should not be accepted as accurate, al- though they should be recorded as documented in the records. TPI documentation can provide valuable information in both criminal and civil cases when unusual defenses or issues are raised. For example, it was once not uncommon for crim- inal defendants who were Vietnam veterans to report that they committed a certain offense while experiencing ﬂash- backs related to Post Traumatic Stress Disorder (PTSD). In some such cases, while the individual’s presentation was quite credible, it was determined through collateral sources that they had never actually served in Vietnam or had served in a capacity in which they were not exposed to combat. Without conﬁrmation of signiﬁcant trauma, either through military records or third party interview, the viability of a PTSD diagnosis and the associated defense in such cases was greatly diminished. How to Obtain Third Party Information An important question in obtaining collateral interviews is whether such interviews should be conducted in person, over the telephone, or through written questions submitted through an attorney, through the mail, or even via e-mail. There is virtually no research guidance to assist in answering Collecting and Applying Third Party Information 81 any of these questions, so our comments are largely limited to our perceptions regarding the advantages and disadvantages of these different approaches. One recent study suggests that telephone interviews are comparable to face-to-face interviews in the quality of the information obtained regarding diagnostic information (Rohde, Lewinsohn, & Seeley, 1997). A total of 60 adults were interviewed in person and by telephone concerning Axis I disorders, and another 60 adults were interviewed twice regarding Axis II disorders. Agreement between tele- phone and in-person interviews was contrasted with inter- rater agreement, obtained through a second rating of the original interview. The following kappa (chance-corrected agreement) values were obtained between face-to-face and telephone interviews: anxiety disorders (.84), substance use disorders (.73), alcohol use disorders (.70), and major de- pressive disorder (.67). A lower kappa was observed for adjustment disorder with depressed mood (.31). Judging from these very limited data, it is possible to achieve compa- rable results when asking structured questions by telephone or in person. The disadvantage of using telephone interview- ing involves losing access to cues obtained from observing the individual who responded, although auditory cues would still be available. Our view is that whatever slight advantage may be lost in the telephone interview would be outweighed by the facilitation of ease of access and greater mutual con- venience for both interviewee and forensic clinician. It is always possible to schedule an in-person interview for a longer or less structured interview, at the discretion of the forensic clinician, or to schedule such an interview at the preference of the collateral individual being interviewed. Conducting a third party interview in written form, whether through mail or e-mail, presents a different context. The advantages of structured interviewing are retained— speciﬁc, preplanned questions are asked—but all prospects for follow-up questioning based on substance of response or visual or auditory cues provided by the interviewee are sacri- ﬁced. This format might remain useful when trying to conﬁrm or disconﬁrm previously developed material, but would be less helpful when exploring or trying to develop newer material. Speciﬁc Measures Using Third Party Information There are two kinds of measures that deserve mention for their incorporation of TPI. Some established psychological tests, such as the CBCL (Achenbach, 1991) and the PCL-R (Hare, 1991), have been designed and validated using the ob- servations of collaterals such as parents and teachers (CBCL) or existing records in the form of a prison or hospital ﬁle (PCL-R). The application of TPI with these respective mea- sures is guided by the questions that are asked and the ad- ministration instructions contained in their manuals. These are examples of how structured use of TPI can not only be in- cluded as a valuable source of input, but can also ﬁll a partic- ular niche. In the case of the PCL-R, record-based TPI must be incorporated to ensure that deception in self-report does not unduly inﬂuence the ratings that are assigned. Teachers and parents who serve as informants on the CBCL have the advantage of presumably greater accuracy in some cases than children, for whom developmental immaturity might inter- fere with accurate self-reporting. A second approach to collecting TPI involves providing structure that is tailored to a speciﬁc case. In one such case (see Heilbrun, 1990), a defendant in an inpatient forensic setting whom we suspected of malingering was entirely un- cooperative with any attempts to evaluate him. Because he consistently refused to meet with an evaluator, we tried to obtain extensive behavioral observation data on him by de- veloping a “checklist” consisting of every symptom he had ever reported to a hospital staff member or attributed to him in evaluation reports written prior to his hospitalization. We attempted to translate each symptom into the observable behavior that would be expected from someone who gen- uinely experienced such a symptom; for example, an individ- ual who was hearing voices might appear distracted or talk to himself. When we had reduced this list to a total of 20 items, we asked a ward staff member from day shift (7:00 to 3:00) and another from evening shift (3:00 to 11:00) to indicate yes or no on each item, reﬂecting whether the behavior had been observed at any time during the eight-hour shift. Over the course of 400 ratings (each over an eight-hour shift, for a total of 3,200 hours of observation time), “no symptom was observed in more than 2% of the rating periods and many were not reported at all” (p. 194). Although it would have been useful to incorporate self-report and testing data into this evaluation, this approach demonstrates how collateral observers can be used in a speciﬁc way, performing observa- tions that have been carefully structured, to yield data that were useful in considering the question of whether his re- ported symptoms were genuine. COLLECTING AND APPLYING THIRD PARTY INFORMATION In some important respects, the forensic clinician is like an investigative journalist. The use of multiple sources, the as- sessment of consistency across sources, and the attribution of information to source are all shared methods of gathering and 82 Third Party Information in Forensic Assessment interpreting relevant information (Levine, 1980; Melton et al., 1997). In this section, we offer some speciﬁc comments about how TPI can be collected and applied in FMHA, focusing particularly on collateral interviewing. We address two considerations in collecting and applying TPI from collateral interviews. The ﬁrst concerns individuals who are reluctant to participate in such an interview. The sec- ond broad area involves inﬂuences that can limit the accuracy of information obtained through collateral interviews: bias, lack of speciﬁc expertise, suggestibility, and memory loss (see Table 5.4). Individuals who are asked to participate in FMHA by pro- viding collateral information may be reluctant or simply un- willing to do so. Such unwillingness to participate should be respected. However, there are instances in which an individ- ual’s reluctance to participate may change when further infor- mation is provided. This information should be provided in the form of a notiﬁcation of purpose, which should address the following: (a) the names of relevant individuals, including the forensic clinician, the individual(s) being evaluated, and the attorney representing that individual; (b) the volun- tary nature of participation in the collateral interview; © a description of the legal question(s) that triggered the evalua- tion; (d) who requested the forensic clinician’s involvement (typically, the prosecution or defense attorney in criminal cases, defense or plaintiff’s attorney in civil cases, or the court in either); (e) the purpose(s) for which the evaluation could be used; (f) how the information collected in this infor- mation will be used, including citation of the name of the interviewee in the report and testimony, and attribution of the information obtained in the interview speciﬁcally to its source; and (g) an offer to answer any questions that the indi- vidual may have before he or she decides whether to partici- pate. This notiﬁcation should allow the individual to make an informed choice about participation, and may facilitate in- volvement when reluctance is based on general apprehension about the legal process. More speciﬁc concerns may not be overcome, however. For individuals who are concerned TABLE 5.4 Problems Limiting Collateral Interview Accuracy and Suggested Strategies for Problem Management Problem Problem Description Suggested Strategy Reluctance to participate • Apprehensive about process. • Notiﬁcation of purpose and limits of conﬁdentiality. • Concerned about personal consequences • Informed about voluntary nature of participation. of participating. • Informed that unattributed information cannot be used. • Unwilling to have information attributed. Bias • Lack of impartiality. • Consider potential bias from the beginning. • Strong positive or negative feelings about • May be assessed near the end of the interview with the litigant. question such as “What do you think should happen • Preference for outcome. with ______?” • Third party information should be obtained from multiple sources, particularly when bias is suspected. • Conclusions should be developed from trends rather than single-source observations. Lack of speciﬁc expertise • Interviewee is without training or experience • Initial questions should elicit broad observations in speciﬁc area (e.g., psychopathology, (What did the defendant say? do? act like?). substance abuse). • Later questions should focus on speciﬁc, preselected • May not detect subtle indicators of disorder observations of symptoms and behavior (Did the or capacity being assessed. individual show X? act like Y?). • No questions should elicit conclusions (Was she psychotic?). Suggestibility • May be prone to inﬂuence from leading questions. • Initial questions should elicit broad observations (What did the defendant say? do? act like?). • Later questions should focus on speciﬁc, preselected observations of symptoms and behavior (Did the individual show X? act like Y?). • Allows comparison between uncontaminated description (given with little guidance from the interviewer) and speciﬁc but possibly less impartial version given when asked about speciﬁc relevant areas. Memory loss • May have difﬁculty remembering relevant details • Beginning with general questions and moving to more if saw individual only once. speciﬁc areas. • Inﬂuences such as stress, different race of observer • Providing nonsensitive memory aids, such as date and individual, gun focus, and others factors interfering and location. with eyewitness identiﬁcation may operate. Communicating Third Party Information 83 about the perceived consequences to them or those close to them, this notiﬁcation may be less than reassuring. Some individuals may express a willingness to provide information on “background,” with the assumption that they would not be identiﬁed as the source. This is not possible, however, as the forensic clinician is ethically obligated to identify the respec- tive sources of data used in FMHA. The second broad area concerns attributes of those inter- viewed that might yield inaccurate information. Many indi- viduals who might be interviewed as part of FMHA are not impartial; they may have strong feelings about the individual being evaluated and an associated wish for a certain kind of outcome to the litigation. Potential bias must be considered a possibility for every collateral observer interviewed. Part of our recommended approach to managing the inﬂuence of bias would be carried out in the course of FMHA for a number of reasons: multiple sources should be used and con- clusions should be developed based on trends across sources rather than from a single observation. Professional expertise is usually not present in those who are interviewed as collaterals in FMHA. Because the forensic clinician should be seeking observations, not conclusions, from collateral interviewees, this is less of a problem than it might appear. The questioning should begin by eliciting broad observations and subsequently move to more speciﬁc areas when the general observations have been completed. More speciﬁc areas can be preselected by the forensic clini- cian for relevance and importance and the questions asked in a way that calls for behavioral observations and does not presuppose expertise. The same approach to questioning (initially broad, subse- quently more speciﬁc) is useful to prevent the interview itself from giving the interviewee suggestions about what is being sought. Caution should be used with information provided by a collateral interviewee who does not describe noteworthy aspects of, for example, mental health symptomatology during a broad description of the litigant, but responds afﬁr- matively to questions about whether a number of speciﬁc symptoms have been observed. The greater this discrepancy, we suggest, the more the information should be scrutinized for consistency with that provided by other sources. Finally, there is the very real problem of difﬁculty remem- bering what occurred at a speciﬁc time (for collateral inter- viewees who see the litigant frequently) or recalling what occurred in cases in which the interviewee was the victim or witness of an alleged offense. Inﬂuences such as extreme stress and weapon focus, for example, can further limit the accuracy of an account that may have already been based on fairly brief observation (Tooley, Brigham, Maass, & Bothwell, 1987). We recommend providing nonsensitive but relevant details (e.g., date, location), particularly for intervie- wees who often see the litigant, to facilitate a more focused account. However, we must emphasize the extreme impor- tance of not providing details that could affect the intervie- wee’s account of legally relevant behavior or capacities. COMMUNICATING THIRD PARTY INFORMATION There are two primary ways to communicate TPI in FMHA: in reports and in testimony. We offer comments on each. Each FMHA report should contain a speciﬁc, comprehensive listing of the sources of information used in the evaluation. Such a listing is particularly important because of the recom- mended approach to writing an FMHA report, with all factual information attributed to its source(s). Some form of organi- zation of the source listing can be very useful, particularly in cases that have a large number of documents to be reviewed. Because each source citation should include the name of the source, its author, and its date, the sources could be organized alphabetically, by date, or by broader section, with subse- quent organization within the section. When reviewing third party documents, it is useful to identify the source and content by date. If this is done during the review process, it greatly facilitates writing the report in terms of the sequence of events as documented by third party records. A summary of each event can be recorded in the report and will automatically be placed in the order in which it has occurred. It is a straightforward task to trans- fer material to speciﬁc sections of the report once this is accomplished. Many times, information from different third party sources is inconsistent. Such inconsistent material should be cited fully in the text, perhaps with language pointing out the inconsistency (e.g., “James and his mother both indicated that he has been arrested once for trespassing; by contrast, his juvenile arrest history reﬂected two arrests: one for tres- passing and the second for possession with intent to distrib- ute”). The meaning of all material, including that which is inconsistent with other sources, should be reﬂected in a formulation of ﬁndings (whether this is a separate section or integrated with other sections), but this meaning should not be addressed while describing the results of each source of TPI. Some of the organizational aspects of TPI communication in reports are useful for testimony. When multiple sources are listed in a way that allows quick location of a speciﬁc source, and when all information is attributed by source, it facili- tates providing testimony that is precise and efﬁcient. The at- tribution of “truth” or “validity” to a given source can be 84 Third Party Information in Forensic Assessment problematic in forensic contexts, for two reasons. First, it is ultimately the job of the trier of fact to determine what is true in a legal case. Second, it is typically not feasible to system- atically assess the accuracy of one’s ﬁndings in a given case. Thus, we prefer to speak about sources being “inconsistent” or “consistent” rather than indicating that one source “veri- ﬁes” or “conﬁrms” what another has indicated. Finally, being comprehensive and using multiple sources of information to support ﬁndings are very important in both testimony and report writing. Critical thinking, which should be reﬂected in the writing of the report, may be demonstrated in other ways (such as in response to hypothetical questions) during testimony. Finally, it is important to document all attempts to obtain TPI, whether successful or not. Whatever format is used to identify the respondents and sources of information that were received, a similar format should be used to reference information or collateral interviews that an unsuccessful at- tempt was made to obtain. For example, if an evaluator at- tempts to obtain a police report but this document is not provided, this should be noted and referenced by time and date in the sources of information. If a particular collateral respondent declines to participate in an interview, this should be similarly noted. This type of record encourages the evaluator to contact all relevant sources without pre- dicting who will and will not participate. By noting these failed efforts or contacts in the report, the forensic clinician demonstrates the effort that was made to obtain compre- hensive, relevant TPI. CONCLUSION There have been some important advances in the conceptual consideration of using TPI in FMHA during the past decade. Unfortunately, research in this area has lagged behind prac- tice. In some respects, the application of TPI may remain something of an art, similar to that seen in investigative journalism. In other ways, however, the behavioral and medical sciences have important contributions to make in documenting the use, structuring the applications, and vali- dating the approaches used in collecting and applying TPI. We hope this chapter both promotes needed research and contributes to better practice in this area. We also expect that the appropriate use of TPI in forensic assessment will improve the actual and perceived quality of the evaluations and testimony provided to the courts. On that basis, we strongly encourage the use of TPI as forensic clinicians ad- dress the diverse aspects of human nature that are seen in this area of practice. REFERENCES Achenbach, T. (1991). Manual for the Child Behavior Check- list/4–18 and 1991 proﬁle. Burlington: University of Vermont Department of Psychiatry. Achenbach, T. M., & Edelbrock, C. (1983). Manual for the Child Behavior Checklist and Revised Child Behavior Proﬁle. Burling- ton: University of Vermont, Department of Psychiatry. Ackerman, M. (1999). Essentials of forensic psychological assess- ment. New York: Wiley. Ackerman, M. J., & Ackerman, M. C. (1997). Custody evalua- tion practices: A survey of experienced professionals (revis- ited). Professional Psychology: Research and Practice, 28, 137–145. American Academy of Psychiatry and the Law. (1995). Ethical guidelines for the practice of forensic psychiatry. Bloomﬁeld, CT: Author. American Psychiatric Association. (1998). The principles of med- ical ethics with annotations especially applicable to psychiatry. Washington, DC: Author. American Psychological Association. (1992). Ethical principles of psychologists and code of conduct. American Psychologist, 47, 1597–1611. Appelbaum, K. (1990). Criminal defendants who desire punish- ment. Bulletin of the American Academy of Psychiatry and the Law, 18, 385–391. Appelbaum, P., & Gutheil, T. (1991). Clinical handbook of psychia- try and the law (2nd ed.). Baltimore: Williams & Wilkins. Borum, R., & Grisso, T. (1996). Establishing standards for criminal forensic reports: An empirical analysis. Bulletin of the American Academy of Psychiatry and the Law, 24, 297–317. Borum, R., Otto, R., & Golding, S. (1993). Improving clinical judg- ment and decision making in forensic evaluation. Journal of Psychiatry and Law, 21, 35–76. Burke, W., Roccaforte, W., Wengel, S., McArthur-Miller, D., Folks, D., & Potter, J. (1998). Disagreement in the reporting of depres- sive symptoms between patients with dementia of the Alzheimer type and their collateral sources. American Journal of Geriatric Psychiatry, 6, 308–319. Chang, G., Goetz, M., Wilkins-Haug, L., & Berman, S. (1999). Pre- natal alcohol consumption: Self versus collateral report. Journal of Substance Abuse Treatment, 17, 85–89. Chermack, S., Singer, K., & Beresford, T. (1998). Screening for alcoholism among medical inpatients: How important is corrob- oration of patient self-report? Alcoholism: Clinical and Experi- mental Research, 22, 1393–1398. Committee on Ethical Guidelines for Forensic Psychologists. (1991). Specialty guidelines for forensic psychologists. Law and Human Behavior, 15, 655–665. Cooke, D., Michie, C., Hart, S., & Hare, R. (1999). Evaluating the screening version of the Hare Psychopathy Checklist–Revised References 85 (PCL:SV): An item response theory analysis. Psychological Assessment, 11, 3–13. Coons, P. (1989). Iatrogenic factors in the misdiagnosis of multiple personality disorder. Dissociation: Progress in the dissociative disorders, 2, 70–76. Costa, P., Williams, T., Somerﬁeld, M., et al. (1996, November). Recognition and initial assessment of Alzheimer’s disease and related dementias (Clinical Practice Guideline No. 19, AHCPR Publication No. 97–0702). Rockville, MD: U.S. Department of Health and Human Services. Daubert v. Merrell Dow Pharmaceuticals, 113 S. Ct. 2786 (1993). Ekman, P., & O’Sullivan, M. (1991). Who can catch a liar? American Psychologist, 46, 913–920. Faust, D. (1995). The detection of deception. Neurological Clinics, 13, 255–265. Frazen, M., Iverson, G., & McCracken, L. (1990). The detection of malingering in neuropsychological assessment. Neuropsychol- ogy Review, 1, 247–279. Frederick, R. (1997). Validity Indicator Proﬁle manual. Minneapo- lis, MN: National Computer Systems. Frye v. United States, 293 F. 1013 (D.C. Cir 1923). Goodman-Delahunty, J., & Foote, W. (1995). Compensation for pain, suffering, and other psychological injuries: The impact of Daubert on employment discrimination claims. Behavioral Sciences and the Law, 13, 183–206. Greenberg, S., & Brodsky, S. (in press). The civil practice of foren- sic psychology: Torts of emotional distress. Washington, DC: American Psychological Association. Grisso, T. (1986). Evaluating competencies: Forensic assessments and instruments. New York: Plenum Press. Grisso, T. (1998). Forensic evaluation of juveniles . Sarasota, FL: Professional Resource Press. Hare, R. D. (1991). Manual for the Revised Psychopathy Checklist. Toronto, Ontario, Canada: Multi-Health Systems. Hart, S., Cox, D., & Hare, R. (1995). The Hare Psychopathy Check- list: Screening version. Toronto, Ontario, Canada: Multi-Health Systems. Hasselblad, V., & Hedges, L. (1995). Meta-analysis of screening and diagnostic tests. Psychological Bulletin, 177, 167–178. Heilbrun, K. (1988, March). Third party information in forensic assessment: Much needed, sometimes collected, poorly guided. Presented at the American Psychology-Law Society/Division 41 mid-year conference, Miami, FL. Heilbrun, K. (1990). Response style, situation, third party informa- tion, and competency to stand trial: Scientiﬁc issues in practice. Law and Human Behavior, 14, 193–196. Heilbrun, K. (1992). The role of psychological testing in forensic assessment. Law and Human Behavior, 16, 257–272. Heilbrun, K. (1996, March). Daubert and forensic mental health assessment: Use and implications. Presented at the biennial conference of the American Psychology-Law Society/Division 41, Hilton Head, SC. Heilbrun, K. (in press). Principles of forensic mental health assess- ment. New York: Plenum Press. Heilbrun, K., & Collins, S. (1995). Evaluations of trial compe- tency and mental state at the time of the offense: Report characteristics. Professional Psychology: Research and Prac- tice, 26, 61–67. Heilbrun, K., Rosenfeld, B., Warren, J., & Collins, S. (1994). The use of third party information in forensic assessments. Bulletin of the American Academy of Psychiatry and the Law, 22, 399–406. Hoberman, H. (1999). The forensic evaluation of sex offenders in civil commitment proceedings. In A. Schlank & F. Cohen (Eds.), The sexual predator (pp. 7-1–7-41). Kingston, NJ: Civic Research Institute. Keilin, W., & Bloom, L. (1986). Child custody evaluation practices: A survey of experienced professionals. Professional Psychol- ogy: Research and Practice, 17, 338–346. Kumho Tire Company, Ltd. v. Carmichael, 526 U.S. 137 (1999). LaFortune, K. (1997). An investigation of mental health and legal professionals’ activities, beliefs, and experiences in domestic court: An interdisciplinary survey. Unpublished doctoral disser- tation, University of Tulsa, OK. Levine, M. (1980). Investigative reporting as a research method: An analysis of Bernstein and Woodward’s All the President’s Men. American Psychologist, 35, 626–638. Lidz, C. W., Mulvey, E. P., & Gardner, W. (1993). The accuracy of predictions of violence to others. Journal of the American Med- ical Association, 269, 1007–1011. Matarazzo, J. D. (1990). Psychological assessment versus psycho- logical testing: Validation from Binet to the school, clinic, and courtroom. American Psychologist, 45, 999–1017. Mayer v. Baiser, 497 N. E.2d 827 (1986). McCann, J. (1998). Malingering and deception in adolescents. Washington, DC: American Psychological Association. Melton, G., Petrila, J., Poythress, N., & Slobogin, C. (1997). Psycho- logical evaluations for the courts: A handbook for mental health professionals and lawyers (2nd ed.). New York: Guilford Press. Mulvey, E., & Lidz, C. (1993). Measuring patient violence in dan- gerousness research. Law and Human Behavior, 17, 277–288. Newhill, C., Mulvey, E., & Lidz, C. (1995). Characteristics of vio- lence in the community by female patients seen in a psychiatric emergency service. Psychiatric Services, 46, 785–795. Nicholson, R., LaFortune, K., Norwood, S., & Roach, R. (1995, August). Quality of pretrial competency evaluations in Okla- homa: Report content and consumer satisfaction. Presented at the annual convention of the American Psychological Associa- tion, New York. Nicholson, R., & Norwood, S. (2000). The quality of forensic psy- chological assessments, reports, and testimony: Acknowledging 86 Third Party Information in Forensic Assessment the gap between promise and practice. Law and Human Behavior, 24, 9–44. Otto, R., Barnes, G., & Jacobson, K. (1996, March). The content and quality of criminal forensic evaluations in Florida. Pre- sented at the biennial meeting of the American Psychology-Law Society, Hilton Head, SC. Otto, R., Edens, J., & Barcus, E. (2000). The use of psychological testing in child custody evaluations. Family and Conciliation Courts Review, 38, 312–340. Pfeffer, R. (1995). A social function measure in the staging and study of dementia. In M. Bergener, J. Brocklehurst, & S. Finkel (Eds.), Aging, health, and healing (pp. 459–475). New York: Springer. Resnick, P. (1995). Malingering. Mental and emotional injuries in employment litigation. Washington, DC: Bureau of National Affairs. Robbins, E., Waters, J., & Herbert, P. (1997). Competency to stand trial evaluations: A study of actual practice in two states. Journal of the American Academy of Psychiatry and Law, 25, 469–483. Roesch, R., Hart, S., & Ogloff, J. (Eds.). (1999). Psychology and law: The state of the discipline. New York: Kluwer Academic/ Plenum. Rogers, R. (1984). Towards an empirical model of malingering and deception. Behavioral Sciences and the Law, 2, 93–112. Rogers, R. (Ed.). (1997). Clinical assessment of malingering and deception (2nd ed.). New York: Guilford Press. Rohde, P., Lewinsohn, P., & Seeley, J. (1997). Comparability of telephone and face-to-face interviews in assessing Axis I and II Disorders. American Journal of Psychiatry, 154, 1593–1598. Schacter, D. (1986). Amnesia and crime: How much do we really know? American Psychologist, 41, 287–295. Shapiro, D. (1984). Psychological evaluation and expert testimony: A practical guide for forensic work. New York: Van Nostrand- Reinhold. Shapiro, D. (1991). Forensic psychological assessment: An integra- tive approach. Boston: Allyn & Bacon. Skeem, J., Golding, S., Cohn, N., & Berge, G. (1998). The logic and reliability of evaluations of competence to stand trial. Law and Human Behavior, 22, 519–547. Somerﬁeld, M., & Costa, P. (1999). Toward more evidence-based guidelines for psychology. American Psychologist, 54, 1131–1132. Steadman, H. J., Mulvey, E., Monahan, J., Robbins, P., Appelbaum, P., Grisso, T., et al. (1998). Violence by people discharged from acute psychiatric inpatient facilities and by others in the same neighborhoods. Archives of General Psychiatry, 55, 1–9. Swanson, J., Borum, R., Swartz, M., & Hiday, V. (1999). Violent be- havior preceding hospitalization among persons with severe mental illness. Law and Human Behavior, 23, 185–204. Tooley, V., Brigham, J., Maass, A., & Bothwell, R. (1987). Facial recognition: Weapon effect and attentional focus. Journal of Applied Social Psychology, 17, 845–859. Udoﬁa, O., Etuk, L., & John, M. (1996). Personality assessment through third party interview: How valid? Journal of Psychology in Africa, South of the Sahara, the Caribbean and Afro Latin American, 1, 1–11. Warren, J., Hazelwood, R., & Dietz, P. (1996). The sexually sadistic serial killer. Journal of Forensic Sciences, 41, 970–974. Weiss, R., Greenﬁeld, S., Grifﬁn, M., Najavits, L., & Fucito, L. (2000). The use of collateral reports for patients with Bipolar and substance use disorders. American Journal of Drug and Alcohol Abuse, 26, 369–378. Wong, S. (1988). Is Hare’s Psychopathy Checklist reliable without the interview? Psychological Reports, 62, 931–934. CHAPTER 6 Forensic and Clinical Issues in the Assessment of Psychopathy JAMES F. HEMPHILL AND STEPHEN D. HART 87 THE NATURE OF PSYCHOPATHY 88 Clinical Features 88 Diagnostic Issues 88 Assessment Issues 89 REVIEW OF ASSESSMENT PROCEDURES 90 Structured Diagnostic Interviews 90 Self-Report Questionnaires and Inventories 91 Expert Rating Scales 93 IMPORTANT ISSUES 95 Psychopathy as a Legal Concept 95 Psychopathy in Childhood and Adolescence 95 Psychopathy and Violence Risk 96 Precision of Measurement 96 Association among Assessment Procedures 97 Psychopathy among Various Cultural Groups 97 RECOMMENDATIONS FOR PRACTICE 98 Failure to Use Accepted Assessment Procedures 98 Improper Reliance on Scientiﬁc Literature 98 Training 98 Consider Other Assessment Findings 99 Categorical versus Dimensional Models of Personality Disorder 99 The High Prevalence of Personality Disorder 100 Complexity of Personality Disorder Symptomology 100 Comorbidity with Acute Mental Disorder 100 Causal Role of Psychopathy 100 The Diagnostic Signiﬁcance of Antisocial Behavior 101 Recommendations 101 AREAS FOR FUTURE RESEARCH 101 Examine Stability of PCL Scores 101 Examine Incremental Validity 102 Study Clinical Settings 102 Evaluate Treatment Efﬁcacy 103 SUMMARY 103 REFERENCES 104 Psychopathy—also known as psychopathic, antisocial, or dissocial personality disorder—has been the focus of inten- sive research investigations for the past two decades (Hare, 1996). A large body of research has examined the assessment of the disorder, evaluated different etiological models, de- scribed its patterns of comorbidity with other mental disor- ders, and investigated its association with antisocial behavior (Cooke, Forth, & Hare, 1998; Hare, Cooke, & Hart, 1999). An important factor in the growth of interest concerning psy- chopathy was the development of the original and revised versions of the Psychopathy Checklist (PCL and PCL-R; Hare, 1980, 1991). Unless otherwise stated, the term PCL will be used to refer to both psychological instruments because ﬁndings obtained from the original PCL and the PCL-R are generalizable to the other version (see Hare, 1991, p. 4). Research using these tests has revealed clear associations between psychopathy and criminal behavior, especially speciﬁc forms of interpersonal violence, in a variety of populations (Hart & Hare, 1997). Psychopathy now is recog- nized as a critical factor in risk assessment (Hart, 1998) and can affect decisions involving civil commitment, parole from prison, access to treatment, detention under dangerous of- fender legislation, and even capital sentencing (Hart, 2001; Lyon & Ogloff, 2000; Zinger & Forth, 1998).Accordingly, the assessment of psychopathy is a fundamental skill for clinical- forensic psychologists. This chapter begins with a discussion of the nature of psy- chopathy, focusing on current clinical conceptualizations of the disorder. The second section reviews the most commonly used methods for assessing psychopathy, focusing on the PCL-R. The third section identiﬁes important professional and clinical issues that practitioners should keep in mind when assessing psychopathy. The fourth section examines practice recommendations concerning the assessment of psychopathy in clinical-forensic settings. The chapter concludes with a dis- cussion of issues that are priorities for future research. 88 Forensic and Clinical Issues in the Assessment of Psychopathy THE NATURE OF PSYCHOPATHY Clinical Features Psychopathy is a speciﬁc form of personality disorder. Like all personality disorders, it is characterized by a disturbance in relating to one’s self, others, and the environment. It is chronic in nature, typically is evident in childhood or adoles- cence, and persists into middle or late adulthood (American Psychiatric Association, 1980, 1987, 1994; World Health Organization, 1992). Symptoms of personality disorders are rigid, inﬂexible, and maladaptive personality traits: tendencies to act, think, perceive, and feel in certain ways that are stable across time, across situations, and in interactions with different people. What distinguishes psychopathy from other person- ality disorders is the speciﬁc symptom pattern, detailed in now classic works by Arieti (1963), Cleckley (1941), Karpman (1961), and McCord and McCord (1964). Inter- personally, psychopathic individuals are arrogant, superﬁ- cial, deceitful, and manipulative. Affectively, their emotions are shallow and labile; they are unable to form strong emotional bonds with others and are lacking in empathy, anxiety, and guilt. Behaviorally, they are irresponsible, im- pulsive, sensation seeking, and prone to delinquency and criminality. Diagnostic Issues Laypeople sometimes conclude that psychopathy does not exist, confused by the fact that it is not listed in the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders ( DSM-IV; American Psychiatric Association, 1994) or the 10th edition of the International Classiﬁcation of Diseases ( ICD-10; World Health Organization, 1992). This is, of course, incorrect. As noted previously, at a con- ceptual or linguistic level, psychopathic personality disorder is synonymous with antisocial, dissocial, and sociopathic personality disorder; they are simply different terms for the same disorder (this is explicitly recognized in the DSM-IV; see American Psychiatric Association, 1994; p. 646). Numer- ous other terms have been used to refer to the same disorder, and Werlinder (1978, Appendix) has identiﬁed more than 175 of them. So, psychopathy is listed in the DSM-IV, where it is referred to as antisocial personality disorder, and in the ICD- 10, where it is referred to as dissocial personality disorder. At an operational level, it must be emphasized that various diagnostic criteria sets for psychopathic, antisocial, dissocial, and sociopathic personality disorder deﬁnitely are not equiva- lent. Perhaps the biggest difference is that diagnostic criteria for psychopathic or dissocial personality disorder typically include a broad range of interpersonal, affective, and behav- ioral symptoms (e.g., Cleckley, 1941; Hare, 1980, 1991; Hart, Cox, & Hare, 1995; World Health Organization, 1992). As an example, Table 6.1 summarizes the ICD-10 diagnostic criteria for dissocial personality disorder. In contrast, diagnostic crite- ria for antisocial or sociopathic personality disorder tend to focus more narrowly on overt delinquent and criminal behav- ior (e.g., American PsychiatricAssociation, 1980, 1987, 1994; Feighner et al., 1972; Robins, 1966). As an example, Table 6.2 summarizes the DSM-IV criteria for antisocial personality dis- order. The differences between these two diagnostic traditions are discussed at length elsewhere (Cunningham & Reidy, 1998; Hare, Hart, & Harpur, 1991; Hart & Hare, 1997; Lilienfeld, 1994; Widiger & Corbitt, 1995). Perhaps the most TABLE 6.1 ICD-10 Criteria for Dissocial Personality Disorder A. Callous unconcern for the feelings of others and lack of the capacity for empathy. B. Gross and persistent attitude of irresponsibility and disregard for social norms, rules, and obligations. C. Incapacity to maintain enduring relationships. D. Very low tolerance to frustration and a low threshold for discharge of aggression, including violence. E. Incapacity to experience guilt and to proﬁt from experience, particularly punishment. F. Marked proneness to blame others or to offer plausible rationalizations for the behavior bringing the subject into conﬂict with society. G. Persistent irritability. Source: Adapted from World Health Organization (1992). International Classiﬁcation of Diseases (10th ed.). Geneva, Switzerland: Author. TABLE 6.2 DSM-IV Criteria for Antisocial Personality Disorder A. Antisocial behavior since age 15, as indicated by three or more of the following: 1. Repeated criminal acts. 5. Recklessness. 2. Deceitfulness. 6. Irresponsibility. 3. Impulsivity. 7. Lacks remorse. 4. Irritability and aggressiveness. B. Current age at least 18. C. Conduct disorder before age 15, as indicated by clinically signiﬁcant impairment in social, academic, or occupational functioning resulting from three or more of the following: 1. Bullied. 9. Destroyed property. 2. Fought. 10. Break and enter. 3. Used weapons. 11. Lied. 4. Cruel to people. 12. Stole. 5. Cruel to animals. 13. Stayed out late (before 6. Robbed. age 13). 7. Forced sex on others. 14. Ran away from home. 8. Set ﬁres. 15. Truant. D. Occurrence of antisocial behavior not exclusively during the course of Schizophrenia or manic episodes. Source: Adapted from American Psychiatric Association (1994). Diagnos- tic and Statistical Manual of Mental Disorders (4th ed.). Washington, DC: Author. The Nature of Psychopathy 89 important consequence of the focus on delinquent and crimi- nal behavior in diagnostic criteria sets for antisocial or socio- pathic personality disorder is that they lack speciﬁcity (i.e., misconduct can be a manifestation of other forms of disorders as well), and this can lead to overdiagnosis in forensic settings and underdiagnosis in other settings. (This point is discussed explicitly in the DSM-IV; see American Psychiatric Associa- tion, 1994, p. 647; see also Hare, 1983, 1985; Hare et al., 1991; Widiger & Corbitt, 1995.) Assessment Issues The nature of assessment procedures should reﬂect the decision-making purpose for which the assessments will be used and the nature of the disorder being assessed. Such “goodness-of-ﬁt” has been referred to as method-function match and method-mode match, respectively (Haynes, Richard, & Kubany, 1995). Clinical or expert ratings of psychopathy have a better goodness-of-ﬁt than other assessment methods when used in forensic decision making, as they permit the integration of diverse sources of information. The use of self-report meth- ods (e.g., questionnaires, structured diagnostic interviews) or projective methods to assess psychopathy is potentially prob- lematic, unless ﬁndings are subsequently conﬁrmed through a review of information from other sources, such as collateral informants and ofﬁcial records. Method-Function Match Assessment procedures for psychopathy should take into account the special needs and requirements of forensic deci- sion making. These have been discussed at length by others, both generally (e.g., Committee on Ethical Guidelines for Forensic Psychologists, 1991; Heilbrun, 1992; Melton, Petrila, Poythress, & Slobogin, 1997) and with respect to the assessment of personality disorders (e.g., Hart, 2001). One important legal issue is that assessment procedures should not rely unduly on uncorroborated statements made by the person being evaluated. Three practical issues should also be kept in mind. First, assessment procedures should require minimal levels of co- operation. Contextual pressures encourage response distor- tion, particularly minimization and denial of psychopathic symptomatology. Acute and chronic mental disorders are common in forensic settings and it is not always possible to obtain informed consent from the individuals being assessed. Of course, in cases where informed consent cannot be obtained from individuals to be assessed because they lack the mental capacity, clinicians may be legally and/or ethically required to obtain informed consent from substitute decision makers. In other cases, individuals who have the mental ca- pacity to consent may refuse for a variety of reasons to par- ticipate in the clinical assessments. Individuals who refuse to participate in the assessments but who will nonetheless be as- sessed from collateral information should be told of this so that they can be informed of the assessment procedures be- fore they refuse to participate. Informed consent requires that potential participants be informed of the nature and purpose of the assessment, the risks and beneﬁts associated with par- ticipating and not participating in the assessment, the alterna- tives available to them, and who has access to the assessment ﬁndings (Ogloff, 1995). Although clinicians typically should obtain informed con- sent from the persons being assessed, informed consent is not always legally required (e.g., in some court-ordered assess- ments or reviews of correctional ﬁles; see Ogloff, 1995; Schuller & Ogloff, 2001, pp. 19–20). Second, assessment procedures for psychopathy should require minimal levels of insight. Almost by deﬁnition, people suffering from personal- ity disorders do not have sufﬁcient insight into the impact of their behavior on others. This is particularly true for psycho- pathic individuals, whose symptoms may include affective deﬁcits, including a severe lack of empathy. Third, assess- ment procedures for psychopathy should require minimal literacy skills. Forensic populations are characterized by low levels of educational achievement and a high prevalence of deﬁcits in intellectual and neuropsychological functioning (e.g., Wilson & Herrnstein, 1985). Assessment procedures that rely on reading ability or require sustained attention are problematic for this reason. Method-Mode Match There are at least four important features of psychopathy that should be taken into account when assessing the disorder (Hart et al., 1995). First, psychopathy is associated with symptoms that fall into three distinct domains: interpersonal, affective, and behavioral (Cooke & Michie, 2001; Hare, 1991). A corollary of this is that assessment procedures sam- ple systematically and comprehensively from these symptom domains, ideally providing separate measures of each. Sec- ond, psychopathy as a personality disorder is assumed to be reasonably stable throughout adulthood. One corollary of this assumption is that assessment procedures for psychopa- thy should have moderate to high temporal stability (i.e., test-retest reliability), even over lengthy periods of time. Another corollary is that assessment procedures for psy- chopathy should not be sensitive to the affective state of persons being evaluated (i.e., their mood at the time of 90 Forensic and Clinical Issues in the Assessment of Psychopathy assessment). Third, an important symptom of psychopathy is deceitfulness. A corollary of this is that assessment proce- dures for psychopathy should evaluate the extent to which a person characteristically lies and manipulates. Another is that procedures should attempt to minimize the extent to which deceitfulness interferes with the assessment of other psycho- pathic symptomatology. Fourth, psychopathy is associated with delinquency and criminality. There is, however, lack of consensus regarding the nature of this association. Accord- ing to some, delinquency and criminality are a primary symptom of psychopathy; to others, they are an important secondary symptom or associated feature—perhaps even a consequence—of the disorder. Regardless, a corollary is that assessment procedures should be useful for making distinc- tions among offenders or patients in forensic settings; another is that assessment procedures should be related systemati- cally to, but be distinct from, measures of criminality and delinquency. REVIEW OF ASSESSMENT PROCEDURES In this section, we review some commonly used procedures for the clinical-forensic assessment of psychopathy in adults. The procedures we discuss fall into three general categories: structured diagnostic interviews, self-report questionnaires and inventories, and expert rating scales. A comprehensive re- view of these procedures is beyond the scope of this chapter. These and other assessment methods are elaborated in detail in Volume 10 ( Assessment Psychology )ofthis Handbook. The goal in the present discussion is to highlight their impor- tant strengths and weaknesses in light of the assessment issues discussed previously. Structured Diagnostic Interviews These procedures use interview schedules to gather informa- tion from the person being evaluated to make a diagnosis according to ﬁxed and explicit criteria (e.g., Rogers, 1995). Commonly used structured diagnostic interviews for the as- sessment of psychopathy include the Structured Clinical Interview for DSM-IV, Axis II (SCID-II; First et al., 1995) and the International Personality Disorder Examination (IPDE; Loranger et al., 1994). Structured Clinical Interview for DSM-IV As its name implies, this interview was intended to assist in the diagnosis of DSM-IV personality disorders, including antisocial personality disorder. The SCID-II is intended to be administered by trained and experienced clinicians. The in- terview schedule contains a series of questions designed to tap each symptom of the various personality disorders. The questions are phrased so that they encourage respondents to acknowledge relatively minor adjustment problems; accord- ingly, clinicians ask the standard questions and, if the person admits to problems, they are free to probe or ask follow-up questions to conﬁrm the presence and severity of symptoms. Consistent with this approach, evaluators can administer a self-report questionnaire to the person before the interview and then probe only those areas in which the person admits problems. Clinicians are expected to be familiar with the per- son’s psychiatric history in advance, which assists in the dif- ferential diagnosis of DSM-IV Axis I and II disorders. It is possible, although not a requirement, to incorporate collateral information in a SCID-II assessment. The SCID-II does not yield scores per se. Severity ratings for individual symptoms are used to diagnose the presence or absence of each personality disorder and can also be used to create symptom counts for each disorder. Including time spent taking a psychosocial history, overviewing mental dis- order, and administering the self-report screening question- naire, a SCID-II assessment requires approximately two to three hours to complete. International Personality Disorder Examination The IPDE was designed to permit the diagnosis of both DSM-IV and ICD-10 personality disorders, including DSM-IV antisocial and ICD-10 dissocial personality disorder. The IPDE is intended to be administered by trained and experi- enced clinicians. The interview schedule contains a series of general questions, organized thematically, that are designed to tap symptoms of the various personality disorders. Clinicians ask the standard questions and must follow up with a series of probes to conﬁrm the presence and severity of symptoms. Each question is posed to every respondent. Prior to the inter- view proper, clinicians obtain an overview of the respondent’s psychosocial history. The format of the IPDE encourages clin- icians to incorporate collateral information in their symptom ratings. The IPDE severity ratings for individual symptoms can be used to diagnose the presence or absence of each per- sonality disorder and to create symptom counts and dimen- sional ratings for each disorder. Commentary With respect to the assessment issues discussed previously, it is obvious that structured diagnostic interviews rely heavily on statements made by the respondent. This is particularly Review of Assessment Procedures 91 true for SCID-II assessments that use the self-report ques- tionnaire as a screen. It is possible, though, to corroborate the respondent’s statements by incorporating a review of collat- eral information in the assessment. Structured diagnostic interviews require cooperation by the respondent. Most respondents who consent to undergo as- sessments are willing and able to answer questions about their psychosocial history. It is impossible to complete such interviews when the person refuses consent. Administration of a structured diagnostic interview requires relatively little insight on the part of the respondent. The ability of inter- viewers to ask extensive probe or follow-up questions to de- termine the severity of symptoms minimizes the chances that clinicians will overidentify individuals who satisfy the crite- ria. A greater concern is the possibility of failing to correctly identify individuals who satisfy the criteria due to simple de- nial of symptomatology, especially when the SCID-II self- report questionnaire is used as a screen. Review of collateral information can help to avoid this problem. Administration of a structured diagnostic interview does not require much in the way of literacy or intellectual ability on the part of the re- spondent (except for the self-report questionnaire of the SCID-II). A strength of the interviews is that administration can be spread across several sessions in cases where the respondent’s attention or concentration is impaired without affecting the validity of the assessment results. The content of structured diagnostic interviews is limited in the same way as the diagnostic criteria sets they are trying to evaluate. As measures of the DSM-IV antisocial personality disorder criteria and ICD-10 dissocial personality disorder, for example, the SCID-II and IPDE fail to comprehensively as- sess many of the characteristics of psychopathy that clinicians and laypersons ﬁnd central to the disorder (e.g., Davies & Feldman, 1981; Hare et al., 1991; Rogers, Dion, & Lynett, 1992; Rogers, Duncan, Lynett, & Sewell, 1994; Tennent, Tennent, Prins, & Bedford, 1990; Widiger & Corbitt, 1993). In particular, as measures of DSM-IV criteria, the SCID-II and IPDE underemphasize interpersonal characteristics such as manipulativeness and egocentricity and affective characteris- tics such as callousness and lack of empathy. As a measure of ICD-10 criteria, the IPDE neglects characteristics of self- absorption, grandiosity, and smooth interpersonal style that is characterized by deceit, manipulation, and pathological lying. Neither the SCID-II nor the IPDE yields separate scores or in- dices for the individual symptom clusters, although both yield some kind of dimensional score related to global psycho- pathic symptomatology. An important strength of diagnoses made using the SCID-II and IPDE is that they have adequate reliability, including test- retest reliability, and there is no indication that they are unduly inﬂuenced by mood at the time of assessment (e.g., First et al., 1995; Loranger et al., 1994). The SCID-II/ DSM-IV criteria in- clude an item related to deceitfulness, but the IPDE/ ICD-10 criteria do not. SCID-II and IPDE assessments may be suscep- tible to response distortions on the part of the person being evaluated, especially when the self-report questionnaire is used as a preinterview screen in the case of the SCID-II. This susceptibility can be minimized, however, through the system- atic integration of collateral information in the assessment process. Finally, there is relatively little information concerning the association between criminality and SCID-II/ DSM-IV or IPDE/ ICD-10 diagnoses. As noted previously, the DSM-IV criteria for antisocial personality disorder have been criticized for their lack of speciﬁcity in forensic settings. Epidemiologi- cal research in correctional and forensic psychiatric facilities using criteria on which the DSM-IV criteria were based indi- cates that a very high proportion of offenders and patients, typically between 50% and 80%, fulﬁll the criteria for antiso- cial personality disorder (e.g., Hare, 1983; Robins, Tipp, & Przybeck, 1991). Consequently, it is not possible to differen- tiate meaningfully among offenders or patients with respect to psychopathy in forensic settings using the SCID-II or IPDE. There is no systematic evidence that either diagnosis has prognostic signiﬁcance with respect to future criminality or violence. Self-Report Questionnaires and Inventories These procedures require the person being evaluated to re- spond to a series of speciﬁc questions using a ﬁxed response format. Usually, they are administered in written form, al- though it is possible in many cases to administer them orally or by means of audiocassettes. Commonly used questionnaires and inventories for the assessment of psychopathy include the second edition of the Minnesota Multiphasic Personality In- ventory (MMPI-2; Butcher, Dahlstrom, Graham, Tellgen, & Kaemmer, 1989), the third edition of the Millon Clinical Mul- tiaxial Inventory (MCMI-III; Millon, Davis, & Millon, 1997), and the Personality Assessment Inventory (PAI; Morey, 1991). Several promising questionnaires speciﬁcally de- signed to assess psychopathy have been developed (e.g., Blackburn & Fawcett, 1999; Gustaffson & Ritzer, 1995; Hare, 1985; Levenson, Kiehl, & Fitzpatrick, 1995; Lilienfeld & An- drews, 1996), but they are not reviewed here because they are not extensively used in clinical-forensic contexts. Some evi- dence suggests that, among forensic samples, self-report mea- sures of psychopathy are not related to measures of physical violence (e.g., Edens, Poythress, & Lilienfeld, 1999) and crime severity (e.g., Rogers, Gillis, & Dickens, 1989). 92 Forensic and Clinical Issues in the Assessment of Psychopathy Minnesota Multiphasic Personality Inventory The MMPI-2 is a multiscale self-report inventory intended to be a broad-band measure of personality and psychopathology. All 567 items on the MMPI-2 are declarative

bundet

statements phrased in the ﬁrst person singular. Respondents are asked to indicate whether the statements are true or false, or mostly true or false, as applied to them. The MMPI-2 takes approximately 1 to 1.5 hours to complete (see Pope, Butcher, & Seelen, 1993, p. 14), and according to the MMPI-2 manual, requires “an eighth grade reading level to comprehend the content of all the MMPI-2 items and to respond to them appropriately” (Butcher et al., 1989, p. 14; see also p. 1). The MMPI-2 has been translated into a variety of languages, and norms are available for large, representative samples of community residents. Two clinical scales from the MMPI-2—the Psychopathic Deviate ( Pd ) scale and the Hypomania ( Ma ) scale—have been used singly and in combination to assess characteristics of psychopathy. The MMPI-2 has a number of validity scales, in addition to the clinical scales, that are relevant for con- ducting clinical-forensic assessments. Scores on the Variable Response Inconsistency Scale (VRIN) and True Response Inconsistency Scale (TRIN) validity scales, for example, be- come elevated when many pairs of items similar in content are answered inconsistently. The L , F , and K validity scales are also useful for assessing protocol credibility and response bias in forensic contexts (Pope et al., 1993). Items were se- lected for most MMPI-2 clinical scales by statistically con- trasting for each item the response rate from a clinical group of interest with the response rate from a comparison group or groups (Hathaway & McKinley, 1940). The clinical group that was used to construct the original MMPI Pd scale was composed of adolescents, most of whom were females with a long history of minor delinquency, diagnosed as “psycho- pathic personality, asocial and amoral type.” McKinley and Hathaway (1944) acknowledge that “no major criminal types” (p. 167) were involved in the construction of the MMPI Pd scale. It should be recognized that, in addition to characteristics of clinical interest, this empirical approach to scale construction selects items that reﬂect sample character- istics such as socioeconomic background and education (Wiggins, 1973). Millon Clinical Multiaxial Inventory The MCMI-III is a multiscale self-report inventory intended “to provide information to clinicians . . . who must make as- sessments and treatment decisions about individuals with emotional and interpersonal difﬁculties” (Millon et al., 1997, p. 5). It was constructed using a combination of rational/ theoretical and empirical approaches. The MCMI-III con- tains 175 items, all declarative statements phrased in the ﬁrst person singular. Respondents are asked to rate the degree to which they agree with the statements using a true/false response format. Administration of the MCMI-III takes ap- proximately 30 minutes, and self-administration requires at least an eighth-grade reading ability. The items form a number of overlapping scales and in- dices. Four scales are used to assess response styles that may potentially invalidate MCMI-III proﬁles: the Validity Index (Scale V), which measures “bizarre or highly improbable” (p. 118) responses; the Disclosure Index (Scale X), which measures the tendency to provide self-revealing or secretive responses; the Desirability Index (Scale Y), which measures the tendency to provide overly favorable responses; and the Debasement Index (Scale Z), which measures the tendency to overreport personal difﬁculties. Scales 6A and 6B were de- signed to assess, respectively, antisocial personality disorder and sadistic (or aggressive) personality disorder. Norms for the MCMI-III were derived from a large sample of people as- sessed or treated in a wide range of inpatient and outpatient mental health settings. Norms for community residents are not available. Personality Assessment Inventory The PAI is a multiscale self-report inventory intended to mea- sure “critical clinical variables” (Morey, 1991, p. 1). It com- prises 344 items, all declarative statements phrased in the ﬁrst person singular. Respondents are asked to rate the degree to which the statements are true of them on a 4-point scale (1  very true, 2  mainly true, 3  slightly true, 4  false). Administration of the PAI takes approximately one hour. Self-administration requires approximately grade 4 reading ability; a Spanish translation is available. The items form a number of nonoverlapping scales, including 4 to assess response bias, 11 to assess clinical syndromes, 5 to assess treatment-related characteristics, and 2 to assess interper- sonal style. Norms for the PAI were based on a large, repre- sentative sample of community residents and supplemented with norms from clinical settings. One scale, Antisocial Features (ANT), was designed to assess “personality and behavioral features relevant to the constructs of antisocial personality and psychopathy” (Morey, 1991, p. 18). Three subscales measure distinct facets of psychopathic symptomatology. Antisocial Behaviors (ANT-A) taps a history of conduct problems and criminality. Egocentricity (ANT-E) measures self-centered, callous, and remorseless behavior, or “the pathological egocentricity Review of Assessment Procedures 93 and narcissism often thought to lie at the core of this disor- der” (p. 72). Stimulus Seeking (ANT-S) reﬂects “a tendency to seek thrills and excitement and low boredom tolerance” (p. 72). Commentary Self-reports, by deﬁnition, rely only on statements made by the respondent. There is no opportunity to use collateral in- formation to corroborate the respondent’s statements when scoring self-reports. Self-reports require considerable cooperation. Respon- dents who consent must be willing and able to answer a large number of speciﬁc questions, some of which may strike them as odd or irrelevant to the assessment. Administration of self-reports requires some, albeit limited, insight on the part of the respondent. This is particularly true for self-report tests that include items that tap interpersonal and affective symptoms, which are less concrete and speciﬁc than items that tap behavioral symptoms. Administration (especially self-administration) of self-report measures requires substan- tially intact literacy or intellectual ability on the part of the respondent. Self-report measures vary according to recom- mended minimum level of reading ability (e.g., fourth grade for the PAI, eighth grade for the MMPI-2) and to the degree of sustained attention (e.g., 175 items on the MCMI-III, 567 items on the MMPI-2) necessary to complete them. The content of self-reports typically is restricted, focus- ing primarily on behavioral features of psychopathy. The exception is the PAI ANT scale, which contains multiple subscales to assess various symptom domains. Some self- reports, in particular, the MMPI-2 Pd scale, contain items whose content seems either irrelevant to or negatively asso- ciated with psychopathy. Self-report scales have temporal stability that ranges from adequate to impressive. From data presented in the test manuals, however, it appears that scores on psychopathy-related scales often are moderately or moderately-to-highly correlated with scales of negative af- fect on the same inventory. This raises the possibility that ob- served temporal unreliability on the psychopathy scales is the result of contamination by mood state at the time of assessment rather than true ﬂuctuations in psychopathic symptomatology. Most self-reports contain questions related to deceitful- ness, although they may be quite simplistic in nature (e.g., “As a teenager, did you lie a lot?”). Many self-reports, including all those reviewed here, contain scales or indices to evaluate response distortion. Such scales evaluate only the most com- mon forms of response distortion, such as malingering of gen- eral psychopathology or unduly positive self-presentation; they do not evaluate more speciﬁc or sophisticated distortion, such as malingering of speciﬁc mental disorder or minimiza- tion of responsibility for antisocial behavior. Furthermore, self-reports may be unable to control for the impact of response distortion on the assessment of psychopathy. As a consequence, evaluators may be able to determine that respondents were engaging in response distortion, but are unable to use this information to assist in their assessment of psychopathy. The MMPI-2, MCMI-III, and PAI were not designed for use in forensic settings, but correctional norms of some type either exist or are in development for all three inventories. There is not a large and systematic literature involving self- report measures that has consistently found associations with antisocial, criminal, and violent behaviors among offenders or patients in forensic settings. Despite this, some research concerning the validity of self-report measures has accumu- lated in forensic samples (e.g., Bayer, Bonta, & Motiuk, 1985; Hart, Forth, & Hare, 1991; Salekin, Rogers, & Sewell, 1997). For example, Edens and colleagues (Edens, Hart, John- son, Johnson, & Olver, 2000) examined the correlation between ANT total scores on the PAI and total scores on the PCL-R and the PCL-SV in two different forensic samples. Even though the correlations were among the highest found in a clinical setting between a self-report measure of psy- chopathy and the PCL (i.e., r  .54 with the PCL-SV; r  .40 with the PCL-R), diagnostic agreement was only low to moderate. Similarly, Hart and colleagues (1991) examined the correlation between total scores on Scale 6A of the MCMI-II (Millon, 1987) and total scores on the PCL-R in a large sample of offenders. Again, even though the correlation was high ( r  .45), diagnostic agreement between the mea- sures was low ( κ  .25). Expert Rating Scales These procedures are multi-item rating scales. Trained ob- servers rate the severity of symptoms based on all available clinical data (e.g., interview with the respondent, review of case history information, interviews with collateral infor- mants). The PCL and PCL-R fall into this category, as does the Screening Version of the PCL-R (PCL-SV; Hart et al., 1995). Revised Psychopathy Checklist The original PCL (Hare, 1980) was a 22-item rating scale, later revised and shortened to 20 items (PCL-R; Hare, 1991). The PCL-R was designed for use in adult male forensic pop- ulations, with some items being scored entirely or primarily on the basis of criminal records. Items are scored on a 3-point 94 Forensic and Clinical Issues in the Assessment of Psychopathy scale (0  item doesn’t apply; 1  item applies somewhat; 2  item deﬁnitely applies). Table 6.3 lists the PCL-R items, which are deﬁned in detail in the test manual. Total scores can range from 0 to 40; scores of 30 or higher are considered di- agnostic of psychopathy. Earlier analyses identiﬁed two fac- tors underlying the PCL-R items, one reﬂecting interpersonal and affective features, and the other reﬂecting impulsive and antisocial behavior (Hare et al., 1990). More recent research using conﬁrmatory factor analysis has identiﬁed distinct in- terpersonal, affective, and behavioral factors whose mea- surement is uncontaminated by items reﬂecting antisocial behavior (Cooke & Michie, 2001). There are now hundreds of published articles reporting research using the PCL-R, ranging from basic research on etiology to applied research examining the use of the test in violence risk assessment (for a summary, see Cooke et al., 1998). Psychometric analyses based on classical test theory and item response theory indi- cate that the PCL-R has excellent psychometric properties (Cooke & Michie, 1997; Hare et al., 1990). Normative data presented in the PCL-R manual (Hare, 1991) comprise ratings from seven samples of offenders ( N  1,192) and four samples of forensic patients (N  440), all adult men (age 16 or older) from institutions in Canada, the United States, and the United Kingdom. Translations of the PCL-R into more than a dozen languages are completed or in progress, and research supports its cross-cultural validity (Cooke & Michie, 1999; Hare, Clarke, Grann, & Thornton, 2000). Conducting a psychosocial history interview and re- viewing case history information to facilitate scoring of the PCL-R typically requires at least 90 to 120 minutes; however, if the PCL-R is added to a standard assessment battery, which typically includes an interview and review of case history, completion may require 10 or 15 minutes. Although it is stan- dard clinical practice to complete the PCL-R from both inter- view and collateral ﬁle information, it is possible to complete without an interview if extensive collateral information of high quality is available (see Hare, 1991, p. 6). File-only ratings are sometimes conducted if the person refuses or is unable to consent and if all appropriate ethical and legal requirements have been satisﬁed. When individuals refuse to be interviewed for court-mandated assessments but nonethe- less are assessed exclusively from collateral ﬁle information, they should be told of this in advance of the assessments. This procedure allows individuals to be fully informed when they refuse to participate in clinical interviews. Psychopathy Checklist-Screening Version The PCL-SV is a 12-item scale derived from the PCL-R. It was designed for use in adult populations, regardless of gender, psychiatric status, or criminal history. Table 6.4 lists the PCL- SV items, which are deﬁned in detail in the test manual. Scor- ing of the PCL-SV requires less information, and less detailed information, than does the PCL-R; further, the PCL-SV can be scored even when the person does not have a criminal record or when the complete record is not available. Items are scored on the same 3-point scale used for the PCL-R. Total scores can range from 0 to 24; scores of 12 or higher indicate “possible psychopathy,” and scores of 18 or higher indicate “deﬁnite psychopathy.” Psychometric analyses indicate that the PCL- SV has excellent structural properties and is strongly related to the PCL-R (Cooke, Michie, Hart, & Hare, 1999; Hart et al., 1995). Also, the PCL-SV has a factor structure strongly paral- lel to that of the PCL-R (Cooke & Michie, 2001). Normative data presented in the PCL-SV manual com- prise ratings from numerous samples of male and female TABLE 6.3 Items and Factors in the Hare Psychopathy Checklist-Revised Factor Solutions Item Description Two a Three b 1. Glibness/superﬁcial charm. 1 Interpersonal. 2. Grandiose sense of self worth. 1 Interpersonal. 3. Need for stimulation/proneness to 2 Behavioral. boredom. 4. Pathological lying. 1 Interpersonal. 5. Conning/manipulative. 1 Interpersonal. 6. Lack of remorse or guilt. 1 Affective. 7. Shallow affect. 1 Affective. 8. Callous/lack of empathy. 1 Affective. 9. Parasitic lifestyle. 2 Behavioral. 10. Poor behavioral controls. 2 — 11. Promiscuous sexual behavior. — — 12. Early behavioral problems. 2 — 13. Lack of realistic, long-term goals. 2 Behavioral. 14. Impulsivity. 2 Behavioral. 15. Irresponsibility. 2 Behavioral. 16. Failure to accept responsibility for 1 Affective. own actions. 17. Many short-term marital relationships. — — 18. Juvenile delinquency. 2 — 19. Revocation of conditional release. 2 — 20. Criminal versatility. — — Note: —  item does not load on any factor. a See Hare et al. (1990). b See Cooke & Michie (2001). Source: Adapted from Hare (1991). TABLE 6.4 PCL-SV Criteria for Psychopathy Part 1 Part 2 1. Superﬁcial. 7. Impulsive. 2. Grandiose. 8. Poor behavioral controls. 3. Deceitful. 9. Lacks goals. 4. Lacks remorse. 10. Irresponsible. 5. Lacks empathy. 11. Adolescent antisocial behavior. 6. Doesn’t accept responsibility. 12. Adult antisocial behavior. Source: Adapted from Hart et al. (1995). Important Issues 95 offenders, forensic psychiatric patients, civil psychiatric pa- tients, and university students. Translations of the PCL-SV into several languages are completed or in progress. Con- ducting a psychosocial history interview and reviewing case history information to facilitate scoring of the PCL-SV typi- cally takes at least 60 to 90 minutes; however, if it is added to a standard assessment battery, which typically includes an interview and review of case history, completion may require 5 or 10 minutes. As with the PCL-R, it is possible to complete the PCL-SV without an interview, provided that all ethical and legal requirements have been satisﬁed. Commentary Expert rating scales do not rely heavily on uncorroborated statements made by the respondent. Indeed, under some con- ditions, it is possible to score the PCL-R and PCL-SV with- out conducting an interview. Expert rating scales require relatively little cooperation. Most respondents who consent to undergo assessment are willing and able to answer questions about their psychosocial history. Administration of expert rating scales requires rela- tively little insight on the part of the respondent. Heavy re- liance on collateral information and the ability of interviewers to ask extensive probe or follow-up questions to determine the severity of symptoms minimize the chances that clinicians will over- or underidentify individuals who satisfy the crite- ria. Administration of expert rating scales does not require much in the way of literacy or intellectual ability on the part of the respondent. A strength of the interviews is that, in cases where the respondent’s attention or concentration is impaired, administration can be spread across several sessions without affecting the validity of the assessment results. Expert rating scales have good coverage of all symptom domains of psychopathy. They can be used to obtain separate scores or indices for symptom clusters, as well as dimen- sional and categorical scores related to global psychopathic symptomatology. Expert rating scales have reliability that is adequate or better, including test-retest reliability, and there has been no indication that they are unduly inﬂuenced by mood at the time of assessment (e.g., Cooke & Michie, 1997; Cooke et al., 1999; Hare, 1991; Hare et al., 1990; Hart et al., 1995). The PCL-R and PCL-SV contain items directly related to deceitfulness. Their susceptibility to response distortion is minimal as a result of the systematic integration of collateral information in the assessment process. Finally, there is good information concerning the association between criminality and expert ratings scales. The PCL-R and PCL-SV can be used to make meaningful distinctions among people, even in samples of serious and persistent offenders. In samples in which the prevalence of antisocial personality disorder according to DSM-III or DSM-III-R criteria is between 50% and 80%, the prevalence of psychopathy according to PCL-R criteria is approximately 20% to 25% (Hare, 1991). Also, there is a large body of research indicating that psychopathy is a robust risk factor for criminality and violence (Hart, 1998; Hart & Hare, 1997; Hemphill, Hare, & Wong, 1998; Salekin, Rogers, & Sewell, 1996). IMPORTANT ISSUES Psychopathy as a Legal Concept The term psychopathy has been used throughout this chapter to refer to a psychological, not a legal, concept (Lyon & Ogloff, 2000; Ogloff & Lyon, 1998). This distinction is im- portant because, although the term psychopathy may be used in a variety of legal statutes (e.g., in “sexual psychopath” leg- islation), these statutes often deﬁne and use the term in a manner that bears little relationship to the concept discussed here. Simply diagnosing someone as a “psychopath” does not necessarily mean that the person will satisfy the legal criteria for psychopathy or that the diagnosis will be relevant for the purposes of the assessment (Hart, 2001). It is therefore im- portant that clinicians ﬁrst identify the purpose of the assess- ment; that they be familiar with the relevant law, legal issues, and legal standards for the task at hand; and that they deter- mine whether—and if so, how—an assessment of psychopa- thy is relevant to the legal issue or issues. Ogloff and Lyon have stated: “In many cases the precise ‘label’ given to a de- fendant is irrelevant because it is the person’s behavior and cognitive processes and their implications for the speciﬁc legal issues in question that is critical for the law” (p. 411). Put another way, diagnoses of psychopathy typically are not relevant to the law, but a consideration of the cognitive and behavioral processes of psychopaths that bear on the legal issues are. Psychopathy in Childhood and Adolescence For most people, the major features of personality, normal or abnormal, are evident in childhood or adolescence. This is as true for traits related to psychopathy as it is for those related to other personality disorders. Indeed, there has been some re- search on psychopathy-related traits in childhood and adoles- cence (e.g., Barry et al., 2000; Forth, Hart, & Hare, 1990; Lynam, 1997), sometimes using measures derived from or inspired by the PCL. It is important to recognize that there is no clear consensus among developmental psychopathologists that personality disorder in general, or psychopathy in partic- ular, exists in childhood or adolescence (see Edens, Skeem, 96 Forensic and Clinical Issues in the Assessment of Psychopathy Cruise, & Cauffman, 2001; Vincent & Hart, in press). First, it has been argued that one’s “true” personality does not crystal- lize or stabilize for some years after the maturational changes (both biological and social) that follow puberty. Second, even if personality disorder does exist in childhood or adolescence, it will not be manifested as it is in adulthood. For example, it is not until late adolescence or early adulthood that people enter into important social roles and obligations, such as employ- ment, marital relationships, and parenthood, and have the opportunity to succeed or fail in them. Similarly, how would one assess a symptom such as “glibness and superﬁcial charm” among children? Third, it is difﬁcult to determine the extent to which a personality feature is traitlike—that is, stable across time and contexts—in people who are still young. Research to date has conﬁrmed that it is possible to assess psychopathy-related traits in childhood and adolescence with adequate interrater reliability, and that the associations among these traits may have important parallels to those observed in adults; the psychopathy-related traits are also as- sociated with antisocial behavior in ways parallel to that found in adults (for a review, see Forth & Burke, 1998). So, there are reasonable grounds to suspect that we can assess something in childhood or adolescence that looks, at least su- perﬁcially, similar to psychopathy in adulthood. We cannot, however, conﬁrm this suspicion absent a clear demonstration from longitudinal research that the traits persist into adult- hood. It may be that psychopathy-related traits disappear by adulthood as a result of maturation or other factors, and it is also possible that these traits emerge in early adulthood for some individuals. There is simply no good evidence that we are able to identify “psychopathic children” or “ﬂedgling psychopaths” (see Lynam, 1996). It is critical to continue research in this area. If it turns out that we are able to identify children or adolescents on a devel- opmental trajectory toward adult psychopathy, then perhaps it will be possible to develop early intervention programs that prevent or reduce symptomology (e.g., Frick & Ellis, 1999; Gresham, Lane, & Lambros, 2000). Investigators should keep in mind, however, potential ethical problems (e.g., Edens et al., 2001; Ogloff & Lyon, 1998). The procedures for assessing psychopathy among children have received little validation among independent investigators. Psychopathy and Violence Risk The association between psychopathy and criminal behavior, as well as the appropriate use of psychopathy in violence risk assessments, has been discussed at length elsewhere (Hart, 1998; Hart & Hare, 1997; Hemphill, Hare, & Wong, 1998; Salekin et al., 1996). Here, we remind readers that psychopathy may be sufﬁcient in some cases to conclude that an individual is at high risk for future violence, but it is never a necessary factor. That is, there are many ways that someone can be at high risk for violence that are unrelated to psychopa- thy (Hart, 1998). This is especially true when examining risk for speciﬁc forms of violence, such as spousal assault, stalking, and sexual violence, where violence may be related more to disturbances of normal attachment processes rather than the pathological lack of attachment associated with psychopathy. It is also important to note that there is no good scientiﬁc evidence (contrary to some claims; e.g., Harris, Rice, & Quinsey, 1993) that diagnoses or traits of psychopathy, including scores on the PCL-R, can be used either on their own or in combination with other variables to estimate the absolute likelihood of future violence for a given individual with any reasonable degree of scientiﬁc or professional certainty. This is particularly important given the practice of some professionals to use diagnoses of psychopathy or antisocial personality dis- order to support the conclusion that an individual is “more likely than not” (i.e., more than 50% likely) to commit acts of future violence or sexual violence. In some jurisdictions, such a conclusion can be used to justify indeterminate civil commit- ment as a sexual predator (e.g., Janus, 2000) or even capital punishment (Cunningham & Reidy, 1998, 1999). Such a prac- tice is simply unfounded and unethical at the present time. Precision of Measurement All diagnoses and test scores are imprecise, that is, associated with measurement error. For example, with respect to the PCL-R, the standard error of measurement (SEM) is a statis- tical index of the extent to which raters would be expected to disagree concerning a particular individual’s score. The SEM on the PCL-R is approximately 3.25 points (see Hare, 1991, p. 36). This means that when two reasonably competent raters conduct independent assessments of the same people at around the same time, we expect that in approximately 68% of cases their scores will be within 3 points of each other (i.e., 1 SEM), and in approximately 95% of cases their scores will be within 6 points (i.e., 1.96 SEM). Factors such as the lack of an interview, inadequate collateral information, and even poor training of evaluators may increase measurement error. The important point here is that psychologists should qualify their conclusions in light of measurement error. For example, the cutoff for diagnosing psychopathy on the PCL-R is 30 and higher (Hare, 1991, p. 17). When an individual’s total score on the PCL-R is, say, 31 or 28, then the evaluator should be careful in any report to admit that there is some possibility that other competent evaluators might disagree about the individual’s diagnosis (Salekin et al., 1996). Important Issues 97 Because of the uncertainty associated with categorical diag- noses, evaluators should consider interpreting PCL-R scores dimensionally, that is, by characterizing the individual’s trait strength relative to some comparison group. Association among Assessment Procedures Even though we have emphasized throughout this chapter the conceptual differences among criteria sets for assessing psy- chopathy, readers should keep in mind that the empirical as- sociations between them are nonetheless quite strong. The correlations between PCL-R Total scores and antisocial per- sonality disorder diagnoses or symptom counts typically are large in magnitude (approximately r  .55 to .65), and diag- nostic agreement between the procedures typically is fair to good, even in forensic settings (e.g., Hare, 1980, 1985; Widiger et al., 1996). However, the disorders have different prevalence rates. According to DSM criteria, anywhere be- tween 50% and 80% of offenders and forensic patients are diagnosed with antisocial personality disorder, whereas only approximately 15% to 30% of the same people meet the PCL-R criteria for psychopathy (Cunningham & Reidy, 1998; Hare, 1983, 1985; Hare et al., 1990; Robins et al., 1991). An- other important ﬁnding is that the link between psychopathy and antisocial personality disorder is asymmetric. Most peo- ple (approximately 90%) diagnosed as psychopaths by PCL criteria meet the criteria for antisocial personality disorder, whereas a minority (approximately 30%) of those with anti- social personality disorder meet PCL criteria for psychopathy (e.g., Hart & Hare, 1989). Several studies have found low to moderate correlations (typically between r  .30 and r  .45) between PCL diag- noses and popular self-report measures of psychopathy (e.g., Cooney, Kadden, & Litt, 1990; Hare, 1985, 1991; Hart et al., 1991). These results are not simply the result of method vari- ance, as the correlations among self-reports are as low as the correlations between self-reports and clinical diagnoses. Fur- ther, self-report scales of psychopathy tend to be biased in their assessment of psychopathy, correlating more highly with social deviance aspects of psychopathy (as measured by Factor 2 of the PCL) than with the interpersonal and affective features (as measured by Factor 1; e.g., Harpur, Hare, & Hakstian, 1989; Hart et al., 1991). This may reﬂect a bias in the content of self-reports, as suggested above, but may also represent a tendency for psychopaths to be poor observers or reporters of their interpersonal and emotional styles. Psychopathy among Various Cultural Groups Until recently, there has been little systematic and sustained research examining the inﬂuence of race and culture on the reliability, validity, and psychometric properties of the PCL (see also Cunningham & Reidy, 1998). Kosson, Smith, and Newman (1990) conducted one of the earliest studies exam- ining the inﬂuence of race on PCL scores. These authors, who used the original 22-item version and not the 20-item version of the PCL currently in use, concluded that “the overall pat- tern of results [among African American and White offend- ers] contains more parallels than disparities” (p. 257). There were some differences between African Americans and Whites, however. Compared with Whites, African Americans obtained PCL scores that were on average 2.3 points higher and displayed smaller corrected item-to-total correlations for 2 of the 22 items (Previous diagnosis as a psychopath [or sim- ilar], Pathological lying and deception), and the congruence coefﬁcient between African Americans and Whites was low for PCL Factor 1, suggesting the factor structure found among samples of White male offenders (Harpur, Hakstian, & Hare, 1988) did not parallel those found in their sample. Despite these differences, readers should recognize that Kosson et al. could not rule out the inﬂuence of rater bias on their results because all of their raters were White. Further, the authors had a reasonably small sample (i.e., n  124) of AfricanAmerican offenders with which to make psychometric comparisons. More recently, researchers have been using item response theory (IRT) analyses to investigate the psychometric proper- ties of the PCL (e.g., Cooke & Michie, 1997). IRT is a statis- tical procedure for examining psychometric properties of test items that theoretically results in analyses independent of the particular items administered and samples studied (Henard, 2000). Cooke et al. (1999) outline a number of important ad- vantages of IRT analyses, and Cooke (1996) argues that IRT approaches are particularly well suited for conducting cross- cultural research with the PCL. Cooke, Kosson, and Michie (2001) applied IRT analyses to a sample of White and African American adult male inmates. They concluded that, although 5 of the 20 PCL-R items had signiﬁcant differences in item performance between the African American and the White offenders, these differences were small in magnitude and tended to cancel each other out when PCL items were summed together to form total scores. Cooke et al. also con- ducted conﬁrmatory factor analyses and failed to ﬁnd the dif- ference in factor structure between African Americans and Whites reported earlier by Kosson et al. (1990). Taken to- gether, these authors concluded that there are few differences between African American and White offenders in terms of item functioning and that the PCL-R has similar psychomet- ric properties among both African American offenders and White offenders. Of course, in addition to these psychometric analyses of PCL items, validation studies need to be done to establish the clinical utility of the PCL among a variety of 98 Forensic and Clinical Issues in the Assessment of Psychopathy cultural groups. Recent research concerning recidivism among African American offenders yields ﬁndings that are similar to those among White offenders. That is, inmates with high PCL scores are convicted at higher and faster rates than are inmates with low PCL scores, and these results are partic- ularly marked for violent offences (Hemphill, Newman, & Hare, 2001). RECOMMENDATIONS FOR PRACTICE In this section, we identify a number of issues that arise as part of the clinical-forensic assessment of psychopathy and make recommendations for dealing with them. Some of the points raised are relevant to the clinical-forensic assessment of all personality disorders (e.g., Hart, 2001); others are unique to psychopathy (e.g., Hare, 1998). Note that some sec- tions below have been adapted or excerpted from Hart (2001). Failure to Use Accepted Assessment Procedures Because forensic mental health testimony can have signiﬁ- cant impact on individual and collective freedoms, the stan- dards of practice in forensic psychology must be higher than in regular clinical practice. One common mistake in clinical- forensic practice is the failure to use, or the misuse of, ac- cepted assessment procedures. Forensic psychologists who testify about the assessment of psychopathy should expect to be confronted with opinions from other experts or with au- thoritative treatises regarding recommended practice. For ex- ample, it would be easy for a competent lawyer to attack the credibility of an expert who assessed psychopathy in a crim- inal defendant relying solely on self-report inventories. There are at least three concerns here. One is that a clinical inter- view is the basic method for assessing any form of mental disorder, and triers of fact may be justiﬁably concerned by di- agnoses that are not based on standard procedures. The second is that, arguably, self-report inventories constitute a series of uncorroborated statements made by the accused. Some investigators argue, because of the way the MMPI/MMPI-2 was constructed (i.e., items were selected if they statistically differentiated clinical from comparison groups), that independent corroboration of responses is irrelevant to the interpretive signiﬁcance of MMPI/MMPI-2 scale elevations. It should be emphasized, however, that it is impossible to know whether items included in each clinical scale were statistically selected because they reﬂect charac- teristics of clinical interest or instead reﬂect sample character- istics largely irrelevant to clinical interpretation (e.g., see Wiggins, 1973). Further, items that are purportedly “subtle” in content (i.e., that reliably differentiate clinical from com- parison groups but that do not clearly reﬂect characteristics of the clinical group of interest) may be less clinically discrimi- nating, and hence clinically useful, than items that are “obvi- ous” in content (i.e., that clearly reﬂect characteristics of the clinical group of interest; for a discussion, see Graham, 1999, pp. 186–187). Taken together, these ﬁndings suggest that in- dependent corroboration of responses may be clinically im- portant when interpreting the meaning of MMPI/MMPI-2 scales, particularly in forensic settings. The scales designed to detect response distortion incorpo- rated in most self-report inventories do not obviate this fact. Finally, there is no body of research supporting the concur- rent validity of self-report inventories with respect to clinical diagnoses of psychopathy in forensic settings. The little evi- dence that does exist suggests their concurrent validity is moderate at best (e.g., Edens et al., 2000; Hare, 1991; Hart et al., 1991). Improper Reliance on Scientiﬁc Literature Forensic psychologists should make clear when their testi- mony is based on established scientiﬁc principles and ﬁndings and when it is based on professional experience. Unfortu- nately, it is common for psychologists to fail to cite, or to cite improperly, relevant scientiﬁc literature when forming their opinions. For example, the consistent body of literature that supports the use of psychopathy assessments as a reliable in- dicator of a variety of antisocial, criminal, and violent behav- iors is based on research conducted using the PCL (Hart & Hare, 1997; Hemphill, Hare, et al., 1998; Salekin et al., 1996). It is therefore inappropriate to cite research based on the PCL to support a professional opinion in which the patient was as- sessed using some other measure or set of diagnostic criteria (Hare, 1998). Findings generated from PCL assessments may not generalize to other assessment procedures, and a lack of generalizability from the PCL to other procedures seems likely given the low to moderate correspondence among dif- ferent measures of psychopathy. Training Psychopathy assessments involve considerable clinical judg- ment. To adequately rate most PCL items, clinicians typically must conduct a comprehensive interview, review extensive collateral information, consider behaviors across time and multiple domains, assess the credibility of and differentially weigh many sources of information, reconcile discrepancies, and arrive at a single score. Adequate training and experience Recommendations for Practice 99 concerning the proper use of the PCL is essential for clini- cians who conduct forensic assessments. Although this point seems obvious, particularly in foren- sic contexts, where important clinical decisions are made and lives may be greatly affected, Hare (1998) has amply documented a number of egregious examples concerning the misuse of the PCL. It is important to recognize that clin- icians who wish to refer to the large body of empirical liter- ature concerning the PCL to support their decision to use this instrument must complete the PCL in a manner consis- tent with the way in which the reliability and validity infor- mation was obtained. Clinicians who fail to adhere to the scoring procedures outlined in the respective manuals or who routinely obtain scores that are markedly inconsistent with those obtained by experienced raters may be subject to ethi- cal complaints and professional liability. Given that the PCL is a psychological test, users should be careful to use and interpret the instrument for the purposes for which it was in- tended and validated (American Educational Research As- sociation, American Psychological Association, & National Council on Measurement in Education, 1999). This means that test users constantly need to keep apprised of recent developments, research studies, and the appropriate uses of the PCL. Consider Other Assessment Findings Psychopathy is only one factor, albeit an important one, that is often considered when conducting a comprehensive foren- sic assessment. In addition to interpreting the meaning of PCL ratings, decision-makers routinely should consider other psychological test scores and collateral information from a broad range of sources. Inmates, correctional employees, and parole board members sometimes comment on the heavy weight that is attached to PCL scores when clinical decisions are made. The practice of giving excessive weight in clinical decision-making contexts to PCL scores may be undesirable and concerning if it is widespread. It is true that the PCL is among the most robust measures currently available in the area of risk assessment of violence and that it consistently emerges among the strongest risk variables in recidivism studies conducted in a variety of forensic (e.g., Harris et al., 1993) and civil psychiatric (Steadman et al., 2000) settings. Hart (1998) has even argued that “psychopathy is such a robust and important risk factor for violence that failure to consider it may constitute profes- sional negligence” (p. 133). Nonetheless, to make decisions based solely on PCL scores sometimes can lead to misleading conclusions because, although high scores on the PCL are as- sociated with high risk to violently reoffend, low scores on the PCL are not necessarily associated with low risk to vio- lently reoffend. It is not uncommon for some groups of sex- ual offenders who might be at high risk to reoffend to receive PCL scores and prevalence rates that are substantially lower than those typically found among normative samples of adult male offenders. Porter et al. (2000), for example, found that 6.3% of extrafamilial child molesters received PCL-R scores  30. This percentage contrasts with 22.9% of a nor- mative sample of male prison inmates (Hare, 1991), 35.9% of rapists, and 64% of mixed rapists and child molesters (Porter et al., 2000). Despite having PCL scores that are low on aver- age, many child molesters still pose a signiﬁcant risk of sex- ual recidivism decades after release (Rice & Harris, 1997). To summarize, we argue that clinicians who conduct risk as- sessments and other types of forensic assessments should routinely administer the PCL but should not uncritically rely solely on PCL scores to guide their decision making. Categorical versus Dimensional Models of Personality Disorder There is considerable debate in the scientiﬁc literature con- cerning the appropriateness of categorical versus dimen- sional models of personality disorder (Widiger & Sanderson, 1995), including psychopathy (e.g., Harris, Rice, & Quinsey, 1994; Lilienfeld, 1994; Rogers & Dion, 1991). To summa- rize, the categorical model assumes that personality disorder symptomatology can be deﬁned in terms of a small number of “types” that are more or less independent of each other. Each type is characterized by a speciﬁc set of symptoms, and people with a given type of personality disorder are assumed to be a relatively homogeneous group. Both the DSM-IV and the ICD-10 rely on a categorical model for the diagnosis of personality disorder. In contrast, the dimensional model as- sumes that personality disorder symptomatology can be well described in terms of relative standing on a small number of global traits. The PCL-R and related tests are based on the dimensional model. Forensic psychologists should be prepared to acknowl- edge both the strengths and the limitations of the measure- ment models on which their assessments of psychopathy are based and the consequent impact on their opinions. The cate- gorical model is commonly used in clinical practice and has been a focus of considerable research. This widespread ac- ceptance is compelling to laypeople when they attempt to judge the credibility of a professional opinion, even if it is considered weak evidence of credibility in the scientiﬁc com- munity. As a consequence, forensic psychologists whose opinions regarding psychopathy are based solely on dimen- sional models should be prepared to defend their “unusual” 100 Forensic and Clinical Issues in the Assessment of Psychopathy practice by outlining the clear advantages of the dimensional approach. The High Prevalence of Personality Disorder Regardless of whether forensic psychologists adopt a categori- cal or a dimensional model, their assessments are complicated by the high prevalence of personality disorders in forensic set- tings. According to epidemiological research, between 50% and 80% of all incarcerated adult offenders meet the diagnos- tic criteria for antisocial personality disorder (Hare, 1983; Robins et al., 1991); if one considers all the personality disor- ders contained in the DSM-IV or ICD-10, the prevalence rate may be as high as 90% (Neighbors, 1987). Even using the more conservative PCL-R criteria, the prevalence of psy- chopathy averages approximately 15% in forensic psychiatric patients and approximately 25% in offenders (Hare, 1991). Of course, from the dimensional perspective, things are even worse. Every offender has traits of personality disorder; the only question is, How severe are the traits? Triers of fact may be unaware that personality disorder is pandemic in forensic settings and place undue weight on or draw unwarranted conclusions from the diagnosis. Accord- ingly, forensic psychologists should attempt to provide a con- text for diagnoses of psychopathy in three ways. First, they should explicitly acknowledge its high prevalence (e.g., “Mr. X meets the DSM-IV diagnostic criteria for antisocial person- ality disorder, which is found in approximately 50% to 80% of all incarcerated adult offenders”). Second, they should characterize it in terms of relative severity (e.g., “My assess- ment of Mr. X using the PCL-R indicates that he has traits of psychopathic personality disorder much higher than those found in healthy adults, but only average in severity relative to incarcerated adult male offenders”). Third, they should ex- plain what they believe to be its legal relevance in the case at hand (e.g., “In my opinion, Mr. X poses a high risk for future sexual violence relative to other sexual offenders that is due at least in part to a mental disorder, speciﬁcally, a severe anti- social personality disorder characterized by extreme impul- sivity and lack of empathy”). This last point is discussed in more detail later in this chapter. Complexity of Personality Disorder Symptomatology It is difﬁcult to describe in simple terms a person’s function- ing with respect to a domain as broad as personality. Foren- sic psychologists who rely on categorical models are forced to grapple with the issue of comorbidity (Zimmerman, 1994). Research indicates that people who meet the diagnos- tic criteria for a given DSM-IV or ICD-10 personality disor- der also typically meet the criteria for two or three other personality disorders (e.g., Stuart et al., 1998). Even people with the same personality disorder diagnosis vary consider- ably with respect to the number and severity of symptoms they exhibit. Psychologists who rely on dimensional models are no better off, as the same level of trait severity can be manifested at the behavioral level in many different ways. Regardless of which model they use, psychologists must rely on information provided by the patient or from other sources to reach a judgment regarding the presence or absence of symptomatology, a judgment that is inherently subjective. Forensic psychologists should be prepared to admit— without making a personal apology for the limitations of sci- entiﬁc knowledge—that assessing personality can be a messy business; the types or dimensions used in assessment are somewhat fuzzy and imprecise concepts. Of course, this does not necessarily render invalid the inferences psychologists can draw from the assessment of psychopathy. Also, it should be remembered that acknowledging the limitations of one’s opinions might help to establish the credibility of those opin- ions in the eyes of the triers of fact. Comorbidity with Acute Mental Disorder In forensic settings, personality disorder frequently is comor- bid with acute mental disorders such as substance use, mood, and anxiety disorders (more generally, Trestman, 2000; with respect to psychopathy, e.g., Hart & Hare, 1989; Hemphill, Hart, & Hare, 1994). Acute mental disorders can complicate the assessment of personality disorder, leading to uncertain or even incorrect inferences about personality (e.g., poverty of affect in a person with schizophrenia mimicking the shallow emotion often associated with psychopathy). Also, the exis- tence of acute mental disorder can be obscured by comorbid personality disorder. If the acute mental disorder has an im- pact on psychological functioning or behavior that is independent of but mistakenly attributed to personality disor- der, the evaluator may reach inaccurate conclusions regarding the severity and forensic relevance of the personality disorder. Forensic psychologists should conduct comprehensive as- sessments of acute mental disorder before making diagnoses of psychopathy. They should also clearly indicate the existence of any acute mental disorder and discuss the extent to which it may have inﬂuenced any opinions related to psychopathy. Causal Role of Psychopathy An evaluator’s opinion that a person suffers from psychopa- thy is, in itself, not of much interest in forensic decision making. In the law, personality disorder generally is relevant only if the evaluator’s opinion is that it causes, at least in part, some impairment of competency or elevated risk for Areas for Future Research 101 criminality and violence for this individual (i.e., the psychol- ogist establishes a “causal nexus”). The unwarranted as- sumption of causality may render an opinion inadmissible because it is deemed to be irrelevant, not probative, or more prejudicial than probative. Forensic psychologists should make explicit their opin- ions regarding the causal role played by psychopathy with re- spect to the relevant legal issue, whether impairment or risk. They also should acknowledge that such opinions are, ulti- mately, professional rather than scientiﬁc in nature, that is, based on inference and speculation, not on the direct applica- tion of scientiﬁc principles or ﬁndings. The Diagnostic Signiﬁcance of Antisocial Behavior A history of antisocial behavior may be of considerable diag- nostic signiﬁcance in civil psychiatric settings, where only a minority of patients has been charged with or convicted of criminal offenses. In the DSM-IV, the diagnostic criteria for an- tisocial personality disorder are based largelyon such a history. Obviously, antisocial behavior is of little diagnostic signiﬁ- cance in many forensic settings, in which virtually everyone has arrest records (American Psychiatric Association, 1994). Forensic psychologists should be careful not to overem- phasize antisocial behavior, especially isolated criminal acts, when diagnosing psychopathy. By deﬁnition, personality disorders should be manifested across various domains of psychosocial functioning, across time, and across important personal relationships (American Psychiatric Association, 1994; World Health Organization, 1992). A person who en- gages in antisocial behavior only of a speciﬁc type, only against a speciﬁc person, or only at speciﬁc times may not suf- fer from a personality disorder at all. For example, consider a 50-year-old man who suffers from a sexual deviation and ex- poses his genitals to teenage girls in public places several times per year, but who is otherwise well adjusted (i.e., has a relatively stable marriage, holds a steady job, has good peer relationships). In this case, the sexual deviation accounts for the patient’s antisocial behavior; there is no need to infer the presence of psychopathy or even traits of psychopathy. Other mental disorders commonly associated with speciﬁc patterns of antisocial behavior include impulse control disorders such as kleptomania (stealing) and pyromania (ﬁre-setting). Recommendations Following is a list of speciﬁc recommendations for practice regarding the clinical-forensic assessment of psychopathy. The recommendations are intended to improve the usefulness of expert testimony by clarifying the foundation of profes- sional opinions, increasing the richness of information provided to decision makers, and facilitating discussion of the limitations of the testimony. • Psychopathy should be assessed using methods that inte- grate information obtained from collateral sources with (whenever possible) information from direct interviews; methods based solely on oral or written self-report should not be used. • Psychopathy should be assessed using methods that pro- vide dimensional information regarding symptoms and/or symptom dimensions (e.g., severity ratings and symptom counts), either in addition to or instead of categorical diag- noses made according to established or accepted criteria. • When communicating their opinions, psychologists should acknowledge the limitations of the assessment methods they used and the information on which the assessment was based and discuss the likely impact of these limitations on their conclusions. • Psychologists should conduct comprehensive assessments of acute mental disorder before making diagnoses of psychopathy. • Psychologists should provide a context for their assess- ment of psychopathy by discussing its prevalence in forensic settings. • When communicating their opinions, psychologists should outline the (putative) causal connection between psychopathic symptomatology and any legally relevant impairment from which the person suffers or risk the per- son presents. • Psychologists should avoid overestimating the signiﬁcance of antisocial behavior in the assessment of psychopathy. AREAS FOR FUTURE RESEARCH Despite the popularity of the PCL in forensic settings and the large body of rapidly accumulating research supporting its reliability and validity, there are some important areas that have been inadequately studied. Here, we consider some areas that we believe are research priorities. In addition to these areas, researchers should continue to examine the relia- bility and validity of the PCL in a variety of contexts, sam- ples, and cultural groups. Examine Stability of PCL Scores As discussed previously, psychopathy is presumed to be ﬁrst evident early in life and to remain stable across the lifespan. A corollary of this is that PCL scores should demonstrate high test-retest reliability across time. Another is that individ- uals identiﬁed with psychopathic characteristics early in life 102 Forensic and Clinical Issues in the Assessment of Psychopathy should be the same individuals as those identiﬁed with psy- chopathic characteristics later in life. This line of research is important for both conceptual and practical reasons. From a conceptual perspective, the stability of PCL scores supports the view that psychopathy reﬂects a stable constellation of personality and behavioral characteristics. The PCL is ex- pected to display high test-retest reliability because of the emphasis during assessment on lifetime functioning across many domains of functioning. From a practical perspective, the stability of PCL scores allows practitioners to use the PCL as an important clinical construct relevant to a broad range of clinical tasks that require stability of scores. The clinical application of the PCL for conducting risk assess- ment, for example, assumes that PCL scores are reasonably stable across time; if the scores were not stable, the PCL would not be expected to accurately identify individuals at risk for committing future antisocial and violent behaviors. The stability of the PCL is suggested by the ﬁnding that it consistently is among the most powerful risk factors for anti- social and violent behavior (Harris et al., 1993; Steadman et al., 2000), and that the PCL is still a potent predictor of future criminal behavior with follow-up periods that exceed a decade (e.g., Hemphill, Templeman, Wong, & Hare, 1998; Rice, Harris, & Cormier, 1992). With few exceptions, surprisingly little research has been conducted to examine the test-retest stability of PCL scores. Schroeder, Schroeder, and Hare (1983), who conducted the ﬁrst study of this type, obtained a generalizability coefﬁcient of .89. Their sample was composed of 42 inmates who had each been assessed on the original 22-item PCL and then re- assessed on the same instrument approximately 10 months later. Test-retest reliability of PCL-R scores at one month have been r  .85 among male methadone patients (Alterman, Cacciola, & Rutherford, 1993) and r  .79 among female methadone patients (Rutherford, Cacciola, Alterman, & McKay, 1996). Of course, PCL-R scores conceptually should be stable for periods of time, with long intervals between the ﬁrst and second set of assessments. Rutherford, Cacciola, Alterman, McKay, and Cook (1999) conducted a study with a test-retest interval of two years, and these researchers again found that the PCL-R demonstrated reasonably high test- retest reliability among male and female methadone patients. Nonetheless, these studies conducted with methadone patients should be replicated in forensic samples with long test-retest intervals to establish the stability of PCL-R scores over time, because the psychometric properties of PCL-R scores among substance-dependent patients may differ in important ways from the psychometric properties of PCL-R scores typically found among forensic samples (e.g., Darke, Kaye, Finlay- Jones, & Hall, 1998; McDermott et al., 2000). Examine Incremental Validity It is often useful in applied settings to examine the unique and shared contributions that psychopathy and other variables make to the clinical task at hand. Sechrest (1963) has argued that “validity must be claimed for a test in terms of some increment in predictive efﬁciency over the infor- mation otherwise easily and cheaply available” (p. 154; emphasis in original). In the area of risk assessment, for ex- ample, researchers would examine not only predictive va- lidity coefﬁcients between the PCL and recidivism, but also the additional contribution, if any, that the PCL makes to the prediction of recidivism beyond that offered by other variables. Hemphill, Hare, and Wong (1998) reviewed the evidence concerning the incremental predictive validity of the PCL and other sets of variables with respect to recidivism. They conducted a series of statistical analyses across studies to test the incremental predictive validity of the PCL with these other sets of variables, and they concluded that the PCL con- tributed unique information to the prediction of recidivism beyond that offered by key criminal history and demographic variables and by personality disorder diagnoses; the reverse was not true. PCL scores also were as strongly correlated with general recidivism as were actuarial risk scales designed speciﬁcally to predict reoffending, but PCL scores were more strongly correlated with violent recidivism than were these same actuarial risk scales. Researchers might extend this body of research by routinely testing the incremental validity of the PCL with variables that are theoretically relevant or practically related to the task at hand. By amassing a litera- ture that examines the incremental validity of different mea- sures, clinicians will be in a better position to identify the unique and shared contributions of different measures and to select measures that each contribute unique information to the clinical task. Study Clinical Settings Practicing clinicians do not always score and use the PCL and PCL-SV according to the procedures outlined in the test man- ual (Hare, 1991; Hart et al., 1995). For example, despite Hare’s (1998) cautions that “the PCL-R does not provide an appropriate index of change …atleastnotoverperiodsofless than 10 years or so” (p. 116), we have found that some clini- cians consider the PCL to be a dynamic measure whose scores are sensitive to short-term psychotherapeutic interventions. This misuse of the PCL-R reﬂects a poor understanding of a basic scoring rule clearly described in the administration sec- tion of the test manual (e.g., see Hare, 1991, p. 6), namely, that Summary 103 the PCL-R items should be rated on the basis of the person’s lifetime functioning. Given that PCL assessments can have considerable impact on the lives of those assessed, it is important to determine whether clinicians or raters are using the PCL consistent with the manner in which it was validated. Do raters in clinical practice have the requisite training, experience, and education? Do they obtain scores similar to those of experi- enced raters? Audits of clinical ﬁles would be useful for investigating the accuracy of ratings in clinical practice. Absolute (and not simply relative) scores obtained on the PCL are of particular interest in clinical settings, where di- agnoses often form the basis of important clinical deci- sions. In this regard, it would be instructive to determine what speciﬁc cutpoints, if any, are used in clinical practice; how clinicians interpret PCL scores; whether psychopathy is viewed as a mitigating or an exacerbating factor, or as a treatable or an untreatable condition; the exten

bundet

t to which clinicians separately consider and differentially interpret PCL factor scores; and so forth. Another issue concerning the assessment of psychopathy that is important in clinical practice but that has received little research attention is the ability of those being evaluated to intentionally inﬂuence or manipulate their PCL scores. The impetus to present oneself in a particular way would seem to be considerable in forensic contexts. The public has easy ac- cess via popular books (e.g., Hare, 1993) to detailed accounts of the procedures used to assess psychopathy and to descrip- tions of the key symptoms of psychopathy substantively sim- ilar to the criteria outlined in the PCL-R manual. Given that PCL assessments are based on lifetime functioning and rely heavily on collateral sources, it seems unlikely that PCL scores could be markedly distorted. Research might nonethe- less clarify the parameters under which PCL scores could be distorted (e.g., when collateral information is limited) and the PCL items most susceptible to distortion. Evaluate Treatment Efﬁcacy It makes good sense to believe that psychopaths will change little as a consequence of treatment or other interventions (at least, not in the short term). Psychopaths, by deﬁnition, expe- rience little remorse or guilt that might propel them into treatment. They are not motivated to actively participate in treatment once enrolled because they see little wrong with themselves, they lack insight and do not recognize the adverse impact that their behaviors have on others, and they habitually lie and manipulate others. These characteristics are generally the antithesis of those that have been found to be important for effecting positive therapeutic change. Many readers may be surprised, therefore, to learn that virtu- ally no methodologically sound treatment study has been conducted evaluating the treatment efﬁcacy of a contem- porary treatment program for psychopaths. Most of the evi- dence concerning poor treatment outcomes ascribed to criminal psychopaths is based on anecdotal case studies or weak research designs (e.g., see Dolan & Coid, 1993; Hemphill & Hart, in press; Wong & Elek, 1989; Wong & Hare, in press). Perhaps the most methodologically rigorous and oft-cited research study to date concerning the efﬁcacy of treatment for psychopaths was conducted by Rice et al. (1992). These authors concluded that treated psychopaths were more violent than were untreated psychopaths during a 10.5-year follow-up. It is important to recognize that this treatment program, although considered innovative in the late 1960s and 1970s, is a nontraditional treatment program that “would not meet current ethical standards” (Harris, Rice, & Cormier, 1991; p. 628). Research that evaluates the efﬁcacy of treatment among psychopaths and that addresses a number of basic method- ological concerns is clearly a priority. Methodologically su- perior studies would include large groups of clearly deﬁned psychopaths who have received well-established treatments that have been delivered consistently and evaluated systemat- ically across long follow-up periods using several measures of treatment outcome. Although research methodologies have improved greatly across time (e.g., Hare et al., 2000; Hobson, Shine, & Roberts, 2000; Seto & Barbaree, 1999), there is still considerable room for improvement concerning studies that examine the efﬁcacy of treatment among offend- ers in general and among psychopaths in particular. SUMMARY The procedures for assessing psychopathy can be grouped into three broad categories: structured diagnostic inter- views; self-report questionnaires and inventories; and ex- pert rating scales. This chapter critically examined each of these three broad procedures while keeping in mind the unique assessment issues with respect to forensic contexts and psychopathy assessments. Expert rating systems are considered superior to the other two categories for assess- ing psychopathy. A variety of professional and clinical is- sues that clinicians should keep in mind when conducting psychopathy assessments were discussed, as were practical recommendations for dealing with many of these issues. The chapter concluded with an examination of inadequately studied areas concerning psychopathy that should be a focus of future research. 104 Forensic and Clinical Issues in the Assessment of Psychopathy REFERENCES American Educational Research Association, American Psycholog- ical Association, & National Council on Measurement in Educa- tion. (1999). Standards for educational and psychological testing. Washington, DC: American Educational Research Asso- ciation. American Psychiatric Association. (1980). Diagnostic and statisti- cal manual of mental disorders (3rd ed.). Washington, DC: Author. American Psychiatric Association. (1987). Diagnostic and statisti- cal 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. Arieti, S. (1963). Psychopathic personality: Some views on its psy- chopathology and psychodynamics. Comprehensive Psychiatry, 4 , 301–312. Barry, C. T., Frick, P. J., DeShazo, T. M., McCoy, M. G., Ellis, M., & Loney, B. R. (2000). The importance of callous-unemotional traits for extending the concept of psychopathy to children. Jour- nal of Abnormal Psychology, 109 , 335–340. Bayer, B. M., Bonta, J. L., & Motiuk, L. L. (1985). The Pd sub- scales: An empirical evaluation. Journal of Clinical Psychology, 41 , 780–788. Blackburn, R., & Fawcett, D. (1999). The Antisocial Personality Questionnaire: An inventory for assessing personality deviation in offender populations. European Journal of Psychological Assessment, 15 , 14–24. Butcher, J. N., Dahlstrom, L., Graham, J. R., Tellgen,A., & Kaemmer, B. (1989). Minnesota Multiphasic Personality Inventory–2. Minneapolis: University of Minnesota Press. Cleckley, H. (1941). The mask of sanity . St. Louis, MO: Mosby. Committee on Ethical Guidelines for Forensic Psychologists. (1991). Specialty guidelines for forensic psychologists. Law and Human Behavior, 15 , 655–665. Cooke, D. J. (1996). Psychopathic personality in different cultures: What do we know? What do we need to ﬁnd out? Journal of Personality Disorders, 10 , 23–40. Cooke, D. J., Forth, A. E., & Hare, R. D. (Eds.). (1998). Psychopa- thy: Theory, research, and implications for society . Dordrecht, The Netherlands: Kluwer Press. Cooke, D. J., Kosson, D. S., & Michie, C. (2001). Psychopathy and ethnicity: Structural, item and test generalizability of the Psy- chopathy Checklist–Revised in Caucasian and African-American participants . Manuscript in preparation. Cooke, D. J., & Michie, C. (1997). An item response theory analy- sis of the Hare Psychopathy Checklist–Revised. Psychological Assessment, 9 , 3–14. Cooke, D. J., & Michie, C. (1999). Psychopathy across cultures: Scotland and North America compared. Journal of Abnormal Psychology, 108 , 58–68. Cooke, D. J., & Michie, C. (2001). Reﬁning the construct of psychopathy: Toward a hierarchical model. Psychological Assessment, 13 , 171–188. Cooke, D. J., Michie, C., Hart, S. D., & Hare, R. D. (1999). Evalu- ating the screening version of the Hare Psychopathy Checklist–Revised (PCL:SV): An item response theory analysis. Psychological Assessment, 11, 3–13. Cooney, N. L., Kadden, R. M., & Litt, M. D. (1990). A comparison of methods for assessing sociopathy in male and female alco- holics. Journal of Studies on Alcohol, 51, 42–48. Cunningham, M. D., & Reidy, T. J. (1998). Antisocial personality disorder and psychopathy: Diagnostic dilemmas in classifying patterns of antisocial behavior in sentencing evaluations. Behav- ioral Sciences and the Law, 16, 333–351. Cunningham, M. D., & Reidy, T. J. (1999). Don’t confuse me with the facts: Common errors in violence risk assessment at capital sentencing. Criminal Justice and Behavior, 26, 20–43. Darke, S., Kaye, S., Finlay-Jones, R., & Hall, W. (1998). Factor structure of psychopathy among methadone maintenance pa- tients. Journal of Personality Disorders, 12, 162–171. Davies, W., & Feldman, P. (1981). The diagnosis of psychopathy by forensic specialists. British Journal of Psychiatry, 138, 329–331. Dolan, B., & Coid, J. (1993). Psychopathic and antisocial personal- ity disorders: Treatment and research issues. London: Gaskell. Edens, J. F., Hart, S. D., Johnson, D. W., Johnson, J. K., & Olver, M. E. (2000). Use of the Personality Assessment Inventory to assess psychopathy in offender populations. Psychological Assessment, 12, 132–139. Edens, J. F., Poythress, N. G., & Lilienfeld, S. O. (1999). Identify- ing inmates at risk for disciplinary infractions: A comparison of two measures of psychopathy. Behavioral Sciences and the Law, 17, 435–443. Edens, J. F., Skeem, J. L., Cruise, K. R., & Cauffman, E. (2001). Assessment of “juvenile psychopathy” and its association with violence: A critical review. Behavioral Sciences and the Law, 19, 53–80. Feighner, J. P., Robins, E., Guze, S. B., Woodruff, R. A., Winokur, G., & Munoz, R. (1972). Diagnostic criteria for use in psychi- atric research. Archives of General Psychiatry, 26, 57–63. First, M. B., Spitzer, R. L., Gibbon, M., Williams, J. B. W., Davies, M., Borus, J., et al. (1995). The Structured Clinical Interview for DSM-III-R personality disorders (SCID-II) II: Multi-site test- retest reliability study. Journal of Personality Disorders, 9, 92–104. Forth, A. E., & Burke, H. C. (1998). Psychopathy in adolescence: Assessment, violence, and developmental precursors. In D. J. Cooke, A. E. Forth, & R. D. Hare (Eds.), Psychopathy: Theory, research, and implications for society (pp. 205–229). Dordrecht, The Netherlands: Kluwer Press. Forth, A. E., Hart, S. D., & Hare, R. D. (1990). Assessment of psy- chopathy in male young offenders. Psychological Assessment: A Journal of Consulting and Clinical Psychology, 2, 342–344. References 105 Frick, P. J., & Ellis, M. (1999). Callous-unemotional traits and sub- types of conduct disorder. Clinical Child and Family Psychology Review, 2, 149–168. Graham, J. R. (1999). MMPI-2: Assessing personality and psy- chopathology (3rd ed.). New York: Oxford University Press. Gresham, F. M., Lane, K. L., & Lambros, K. M. (2000). Comorbid- ity of conduct problems and ADHD: Identiﬁcation of ﬂedgling psychopaths. Journal of Emotional and Behavioral Disorders, 8, 83–93. Gustaffson, S. B., & Ritzer, D. R. (1995). The dark side of normal: A psychopathy-linked pattern called aberrant self-promotion. European Journal of Personality, 9, 1–37. Hare, R. D. (1980). A research scale for the assessment of psy- chopathy in criminal populations. Personality and Individual Differences, 1, 111–119. Hare, R. D. (1983). Diagnosis of antisocial personality disorder in two prison populations. American Journal of Psychiatry, 140, 887–890. Hare, R. D. (1985). A comparison of procedures for the assessment of psychopathy. Journal of Consulting and Clinical Psychology, 53, 7–16. Hare, R. D. (1991). The Hare Psychopathy Checklist–Revised. Toronto, Ontario, Canada: Multi-Health Systems. Hare, R. D. (1993). Without conscience: The disturbing world of the psychopaths among us . New York: Pocket Books. Hare, R. D. (1996). Psychopathy: A clinical construct whose time has come. Criminal Justice and Behavior, 23, 25–54. Hare, R. D. (1998). The Hare PCL-R: Some issues concerning its use and misuse. Legal and Criminological Psychology, 3, 99–119. Hare, R. D., Clark, D., Grann, M., & Thornton, D. (2000). Psy- chopathy and the predictive validity of the PCL-R: An interna- tional perspective. Behavioral Sciences and the Law, 18, 623–645. Hare, R. D., Cooke, D. J., & Hart, S. D. (1999). Psychopathy and sadistic personality disorder. In T. Millon, P. H. Blaney, & R. D. Davis (Eds.), Oxford textbook of psychopathology (pp. 555–584). Oxford, England: Oxford University Press. Hare, R. D., Harpur, T. J., Hakstian, A. R., Forth, A. E., Hart, S. D., & Newman, J. P. (1990). The Revised Psychopathy Checklist: Reliability and factor structure. Psychological Assessment: A Journal of Consulting and Clinical Psychology, 2, 338–341. Hare, R. D., Hart, S. D., & Harpur, T. J. (1991). Psychopathy and the DSM-IV criteria for antisocial personality disorder. Journal of Abnormal Psychology, 100, 391–398. Harpur, T. J., Hakstian, R. A., & Hare, R. D. (1988). Factor structure of the Psychopathy Checklist. Journal of Consulting and Clinical Psychology, 56, 741–747. Harpur, T. J., Hare, R. D., & Hakstian, R. A. (1989). Two-factor con- ceptualization of psychopathy: Construct validity and assess- ment implications. Psychological Assessment: A Journal of Consulting and Clinical Psychology, 1, 6–17. Harris, G. T., Rice, M. E., & Cormier, C. A. (1991). Psychopathy and violent recidivism. Law and Human Behavior, 15, 625–637. Harris, G. T., Rice, M. E., & Quinsey, V. L. (1993). Violent recidi- vism of mentally disordered offenders: The development of a statistical prediction instrument. Criminal Justice and Behavior, 20, 315–335. Harris, G. T., Rice, M. E., & Quinsey, V. L. (1994). Psychopathy as a taxon: Evidence that psychopaths are a discrete class. Journal of Consulting and Clinical Psychology, 62, 387–397. Hart, S. D. (1998). The role of psychopathy in assessing risk for violence: Conceptual and methodological issues. Legal and Criminological Psychology, 3, 121–137. Hart, S. D. (2001). Forensic issues. In W. J. Livesley (Ed.), Hand- book of personality disorders: Theory, research, and treatment (pp. 555–569). New York: Guilford Press. Hart, S. D., Cox, D. N., & Hare, R. D. (1995). Manual for the Hare Psychopathy Checklist–Revised: Screening version (PCL:SV). Toronto, Ontario, Canada: Multi-Health Systems. Hart, S. D., Forth, A. E., & Hare, R. D. (1991). The MCMI-II as a measure of psychopathy. Journal of Personality Disorders, 5, 318–327. Hart, S. D., & Hare, R. D. (1989). Discriminant validity of the Psychopathy Checklist in a forensic psychiatric population. Psy- chological Assessment: A Journal of Consulting and Clinical Psychology, 1, 211–218. Hart, S. D., & Hare, R. D. (1997). Psychopathy: Assessment and association with criminal conduct. In D. M. Stoff, J. Breiling, & J. D. Maser (Eds.), Handbook of antisocial behavior (pp. 22–35). New York: Wiley. Hathaway, S. R., & McKinley, J. C. (1940). A multiphasic personal- ity schedule (Minnesota). I: Construction of the schedule. Jour- nal of Psychology, 10, 249–254. Haynes, S. N., Richard, D. C. S., & Kubany, E. S. (1995). Content validity in psychological assessment: A functional approach to concepts and methods. Psychological Assessment, 7, 238–247. Heilbrun, K. (1992). The role of psychological testing in forensic as- sessment. Law and Human Behavior, 16, 257–272. Hemphill, J. F., Hare, R. D., & Wong, S. (1998). Psychopathy and recidivism: A review. Legal and Criminological Psychology, 3, 139–170. Hemphill, J. F. & Hart, S. D. (in press). Motivating the unmotivated: Psychopathy, treatment, and change. In M. McMurran (Ed.), Motivating offenders to change: A guide to enhancing engage- ment in therapy. Chichester, England: Wiley. Hemphill, J. F., Hart, S. D., & Hare, R. D. (1994). Psychopathy and substance use. Journal of Personality Disorders, 8, 169–180. Hemphill, J. F., Newman, J. P., & Hare, R. D. (2001, April). Psy- chopathy and recidivism among Black and White adult male of- fenders . Paper presented at the International Association of Forensic Mental Health Services, Vancouver, British Columbia, Canada. 106 Forensic and Clinical Issues in the Assessment of Psychopathy Hemphill, J. F., Templeman, R., Wong, S., & Hare, R. D. (1998). Psychopathy and crime: Recidivism and criminal careers. In D. J. Cooke, A. E. Forth, & R. D. Hare (Eds.), Psychopathy: The- ory, research and implications for society (pp. 375–399). Dor- drecht, The Netherlands: Kluwer Press. Henard, D. H. (2000). Item response theory. In L. G. Grimm & P. R. Yarnold (Eds.), Reading and understanding more multivariate statistics (pp. 67–97). Washington, DC: American Psychological Association. Hobson, J., Shine, J., & Roberts, R. (2000). How do psychopaths behave in a prison therapeutic community? Psychology, Crime and Law, 6, 139–154. Janus, E. S. (2000). Sexual predator commitment laws: Lessons for law and the behavioral sciences. Behavioral Sciences and the Law, 18, 5–21. Karpman, B. (1961). The structure of neuroses: With special differ- entials between neurosis, psychosis, homosexuality, alcoholism, psychopathy and criminality. Archives of Criminal Psychody- namics, 4, 599–646. Kosson, D. S., Smith, S. S., & Newman, J. P. (1990). Evaluation of the construct validity of psychopathy in Black and White male inmates: Three preliminary studies. Journal of Abnormal Psy- chology, 99, 250–259. Levenson, M. R., Kiehl, K. A., & Fitzpatrick, C. M. (1995). Assess- ing psychopathic attributes in a non-institutionalized population. Journal of Personality and Social Psychology, 68, 151–158. Lilienfeld, S. O. (1994). Conceptual problems in the assessment of psychopathy. Clinical Psychology Review, 14, 17–38. Lilienfeld, S. O., & Andrews, B. P. (1996). Development and pre- liminary validation of a self report measure of psychopathic per- sonality traits in noncriminal populations. Journal of Personality Assessment, 66, 488–524. Loranger, A. W., Sartorius, N., Andreoli, A., Berger, P., Buchheim, P., Channabasavanna, S. M., et al. (1994). The International Personality Disorder Examination: The World Health Organiza- tion/Alcohol, Drug Abuse, and Mental Health Administration international pilot study of personality disorders. Archives of General Psychiatry, 51, 215–224. Lynam, D. R. (1996). Early identiﬁcation of chronic offenders: Who is the ﬂedgling psychopath? Psychological Bulletin, 120, 209–234. Lynam, D. R. (1997). Pursuing the psychopath: Capturing the ﬂedg- ling psychopath in a nomological net. Journal of Abnormal Behavior, 106, 425–438. Lyon, D. R., & Ogloff, J. R. P. (2000). Legal and ethical issues in psychopathy assessment. In C. B. Gacono (Ed.), The clinical and forensic assessment of psychopathy: A practitioner’s guide (pp. 139–173). Mahwah, NJ: Erlbaum. McCord, W., & McCord, J. (1964). The psychopath: An essay on the criminal mind . Princeton, NJ: Van Nostrand. McDermott, P. A., Alterman, A. I., Cacciola, J. S., Rutherford, M. J., Newman, J. P., & Mulholland, E. M. (2000). Generality of Psychopathy Checklist–Revised factors over prisoners and substance-dependent patients. Journal of Consulting and Clini- cal Psychology, 68, 181–186. McKinley, J. C., & Hathaway, S. R. (1944). The Minnesota Multi- phasic Personality Inventory V: Hysteria, hypomania, and psy- chopathic deviate. Journal of Applied Psychology, 28, 153–174. Melton, G. B., Petrila, J., Poythress, N. G., & Slobogin, C. (1997). Psychological evaluations for the courts: A handbook for mental health professionals and lawyers (2nd ed.). New York: Guilford Press. Millon, T. (1987). Millon Clinical Multiaxial Inventory. II: Manual . Minneapolis, MN: National Computer Systems. Millon, T., Davis, R., & Millon, C. (1997). Millon Clinical Multiax- ial Inventory–III (MCMI-III) manual (2nd ed.). Minneapolis, MN: National Computer Systems. Morey, L. C. (1991). The Personality Assessment Inventory pro- fessional manual . Odessa, FL: Psychological Assessment Resources. Neighbors, H. (1987). The prevalence of mental disorder in Michigan prisons. DIS Newsletter, 4, 8–11. Ogloff, J. R. P. (1995). Navigating the quagmire: Legal and ethical considerations. In D. G. Martin & A. D. Moore (Eds.), First steps in the art of intervention: A guidebook for trainees in the helping professions (pp. 347–376). Paciﬁc Grove, CA: Brooks/Cole. Ogloff, J. R. P., & Lyon, D. R. (1998). Legal issues associated with the concept of psychopathy. In D. J. Cooke, A. E. Forth, & R. D. Hare (Eds.), Psychopathy: Theory, research, and implications for soci- ety (pp. 401–422). Dordrecht, The Netherlands: Kluwer Press. Pope, K. S., Butcher, J. N., & Seelen, J. (1993). The MMPI, MMPI-2, and MMPI-A in court: A practical guide for expert witnesses and attorneys . Washington, DC: American Psycho- logical Association. Porter, S., Fairweather, D., Drugge, J., Hervé, H., Birt, A., & Boer, D. P. (2000). Proﬁles of psychopathy in incarcerated sexual offenders. Criminal Justice and Behavior, 27, 216–233. Rice, M. E., & Harris, G. T. (1997). Cross-validation and extension of the Violence Risk Appraisal Guide for child molesters and rapists. Law and Human Behavior, 21, 231–241. Rice, M. E., Harris, G. T., & Cormier, C. A. (1992). An evaluation of a maximum security therapeutic community for psychopaths and other mentally disordered offenders. Law and Human Behavior, 16, 399–412. Robins, L. N. (1966). Deviant children grown up: A sociological and psychiatric study of sociopathic personality . Baltimore: Williams & Wilkins. Robins, L. N., Tipp, J., & Przybeck, T. (1991). Antisocial personality. In L. N. Robins & D. Regier (Eds.), Psychiatric dis- orders in America: The Epidemiologic Catchment Area study (pp. 258–290). New York: Free Press. Rogers, R. (1995). Diagnostic and structured interviewing: A hand- book for psychologists . Odessa, FL: Psychological Assessment Resources. References 107 Rogers, R., & Dion, K. (1991). Rethinking the DSM-III-R diagnosis of antisocial personality disorder . Bulletin of the American Academy of Psychiatry and Law, 19, 21–31. Rogers, R., Dion, K. L., & Lynett, E. (1992). Diagnostic validity of antisocial personality disorder: A prototypical analysis. Law and Human Behavior, 16, 677–689. Rogers, R., Duncan, J. C., Lynett, E., & Sewell, K. W. (1994). Pro- totypical analysis of antisocial personality disorder: DSM-IV and beyond. Law and Human Behavior, 18, 471–484. Rogers, R., Gillis, J. R., & Dickens, S. E. (1989). A research note on the MMPI Pd scale and sociopathy. International Journal of Offender Therapy and Comparative Criminology, 33, 21–25. Rutherford, M. J., Cacciola, J. S., Alterman, A. I., & McKay, J. R. (1996). Reliability and validity of the Revised Psychopathy Checklist in women methadone patients. Assessment, 3, 145–156. Rutherford, M., Cacciola, J. S., Alterman, A. I., McKay, J. R., & Cook, T. G. (1999). The 2-year test-retest reliability of the Psy- chopathy Checklist–Revised in methadone patients. Assessment, 6, 285–291. Salekin, R., Rogers, R., & Sewell, K. (1996). A review and meta- analysis of the Psychopathy Checklist and Psychopathy Checklist–Revised: Predictive validity of dangerousness. Clin- ical Psychology: Science and Practice, 3, 203–215. Salekin, R. T., Rogers, R., & Sewell, K. W. (1997). Construct valid- ity of psychopathy in a female offender sample: A multitrait- multimethod evaluation. Journal of Abnormal Psychology, 106, 576–585. Schroeder, M. L., Schroeder, K. G., & Hare, R. D. (1983). General- izability of a checklist for assessment of psychopathy. Journal of Consulting and Clinical Psychology, 51, 511–516. Schuller, R. A., & Ogloff, J. R. P. (2001). An introduction to psychol- ogy and law. In R. A. Schuller & J. R. P. Ogloff (Eds.), Introduc- tion to psychology and law: Canadian perspectives (pp. 3–28). Toronto, Ontario, Canada: University of Toronto Press. Sechrest, L. (1963). Incremental validity: A recommendation. Edu- cational and Psychological Measurement, 23, 153–158. Seto, M. C., & Barbaree, H. E. (1999). Psychopathy, treatment be- havior, and sex offender recidivism. Journal of Interpersonal Violence, 14, 1235–1248. Steadman, H. J., Silver, E., Monahan, J., Appelbaum, P. S., Robbins, P. C., Mulvey, E. P., et al. (2000). A classiﬁcation tree approach to the development of actuarial violence risk assessment tools. Law and Human Behavior, 24, 83–100. Stuart, S., Pfhol, B., Battaglia, M., Bellodi, L., Grove, W., & Cadoret, R. (1998). The co-occurrence of DSM-III-R personality disorders. Journal of Personality Disorders, 12, 302–315. Tennent, G., Tennent, D., Prins, H., & Bedford, A. (1990). Psycho- pathic disorder: A useful clinical concept? Medicine, Science, and the Law, 30, 39–44. Trestman, R. L. (2000). Behind bars: Personality disorders. Journal of the American Academy of Psychiatry and the Law, 28, 232–235. Vincent, G., & Hart, S. D. (in press). Psychopathy in adolescents. In R. R. Corrado, R. Roesch, S. D. Hart, & J. Gierowski, J. (Eds.), Multi-problem violent youth: A foundation for comparative re- search on needs, interventions, and outcomes . Amsterdam: IOS Press. Werlinder, H. (1978). Psychopathy: A history of the concepts. Analysis of the origin and development of a family of concepts in psychopathology . Doctoral dissertation, Institute of Education, Uppsala University, Sweden. Widiger, T. A., Cadoret, R., Hare, R. D., Robins, L., Rutherford, M., Zanarini, M., et al. (1996). DSM-IV antisocial personality disorder ﬁeld trial. Journal of Abnormal Psychology, 105, 3–16. Widiger, T. A., & Corbitt, E. M. (1993). Antisocial personality dis- order: Proposals for DSM-IV. Journal of Personality Disorders, 7, 63–77. Widiger, T. A., & Corbitt, E. M. (1995). Antisocial personality dis- order in DSM-IV . In J. Livesley (Ed.), DSM-IV personality disorders (pp. 127–134). New York: Guilford Press. Widiger, T. A., & Sanderson, C. J. (1995). Toward a dimensional model of personality disorders. In W. J. Livesley (Ed.), The DSM-IV personality disorders (pp. 433–458). New York: Guilford Press. Wiggins, J. S. (1973). Personality and prediction: Principles of personality assessment . Reading, MA: Addison-Wesley. Wilson, J. Q., & Herrnstein, R. J. (1985). Crime and human nature . New York: Touchstone. Wong, S., & Elek, D. (1989). The treatment of psychopathy: A re- view . Unpublished manuscript. Wong, S., & Hare, R. D. (in press). Program guidelines for the in- stitutional treatment of violent psychopaths . Toronto, Ontario, Canada: Multi-Health Systems. World Health Organization. (1992). ICD-10: International statisti- cal classiﬁcation of diseases and related health problems (10th rev.). Geneva, Switzerland: Author. Zimmerman, M. (1994). Diagnosing personality disorders: A review of issues and research methods. Archives of General Psychiatry, 51, 225–245. Zinger, I., & Forth, A. (1998). Psychopathy and Canadian criminal proceedings: The potential for human rights abuses. Canadian Journal of Criminology, 40, 237–276. CHAPTER 7 Evaluation of Malingering and Deception RICHARD ROGERS AND SCOTT D. BENDER 109 CONCEPTUAL ISSUES 109 Deﬁnitions of Response Styles 109 Perspectives of Malingering in the Forensic Context 110 Explanatory Models of Malingering 110 Explanatory Models of Defensiveness 111 Misassumptions about Malingering and Dissimulation 111 Applications to Forensic Practice 112 EMPIRICAL ISSUES 112 Basic Designs 112 Incremental Validity 113 MALINGERING AND MENTAL DISORDERS 114 Detection Strategies 114 Featured Measures 114 DEFENSIVENESS AND MENTAL DISORDERS 118 Overview 118 Minnesota Multiphasic Personality Inventory 118 Paulhus Deception Scales 119 FEIGNED COGNITIVE IMPAIRMENT 119 Detection Strategies 120 Guidelines for the Classiﬁcation 123 Featured Measures 124 SUMMARY 126 REFERENCES 126 CONCEPTUAL ISSUES The validity of most psychological measures is predicated on the cardinal assumption that evaluatees are responding in a forthright manner and putting forth a sincere effort. Is this as- sumption warranted in forensic practice? External inﬂuences on self-reporting and effort may include the adversarial effects of litigation and pressures exerted by interested others, such as attorneys and family members. Internal inﬂuences may in- clude (a) reactions to questioned credibility, (b) stigmatization of mental disorders or disability status, © effects of a genuine disorder, or (d) efforts to obtain undeserved beneﬁts. Forensic psychologists tend to focus on the last as it relates to malinger- ing and de-emphasize other internal and external inﬂuences. Forensic psychologists may wish to address openly internal and external inﬂuences that potentially arise from their evalua- tions.As part of the informed consent process, they may choose to ask evaluatees about their understanding of the purposes of the evaluation and what they have been told about the evalua- tion by others. Disclosures from the forensic psychologist about the purpose of the evaluation and his or her role may allay some concerns about partiality. Especially in civil cases, an un- hurried and respectful discussion of the evaluation, its purpose, and parameters is needed to address strong negative reactions regarding perceived coercion (e.g., “I had to come”) or ques- tioned legitimacy (e.g., “You think I am making this up”). Tests of cognitive abilities and achievement are premised on optimal effort by evaluatees. Less than optimal effort may vitiate the accuracy of test results and lead to concerns about deliberate underperformance. A largely neglected considera- tion is the effect of genuine disorders on test performance. For example, major depression may reduce performance on cognitive tasks that require sustained attention and concen- tration. Forensic psychologists are cautioned against facile and unwarranted assumptions that suboptimal efforts are always equated with malingering. This section provides an overview of response styles with a summary of accepted terminology. Three general perspec- tives of malingering are explicated. Explanatory models are reviewed with a discussion of inferred motivations for why persons engage in malingering and defensiveness. In addi- tion, misassumptions about response styles are examined in the context of forensic evaluations. Deﬁnitions of Response Styles Rogers (1997) summarized the basic terminology used to de- scribe response styles. Basic deﬁnitions are provided with several updated references: • Malingering (American Psychiatric Association, 2000) is the deliberate fabrication or gross exaggeration of 110 Evaluation of Malingering and Deception psychological or physical symptoms for the fulﬁllment of an external goal. • Defensiveness is the polar opposite of malingering; it is the deliberate denial or gross minimization of symptoms in the service of an external goal. • Irrelevant responding is a disengagement from the assess- ment process typically reﬂected in inconsistent responding that is unrelated to the speciﬁc content (e.g., not reading test items). • Feigning is the deliberate fabrication or gross exaggera- tion of psychological or physical symptoms (Rogers & Vitacco, in press) without any assumptions about its goals. Available tests typically assess feigning, because they are unable to evaluate supposed goals required for the classiﬁcation of malingering or the diagnosis of factitious disorders. • Secondary gain is an imprecise clinical term that should be avoided in forensic evaluations (Rogers & Reinhardt, 1998). In nonforensic settings, the term is used to describe the perpetuation and possible augmentation of symptoms based on unintentional responses to internal (i.e., psycho- dynamic models) or external (i.e., behavioral-medicine models) forces. • Suboptimal effort (also called “incomplete effort”) is a de- scriptive inference that maximum performance was not achieved. Suboptimal effort may be the result of internal states (e.g., fatigue or frustration) or comorbidity (e.g., de- pression subsequent to a head injury). Only when subopti- mal effort is extreme in its presentation should feigning be considered, although internal states and comorbidity must still be addressed. • Dissimulation is a general term to describe an inaccurate portrayal of symptoms and associated features. It is typi- cally used when more precise terms (e.g., malingering and defensiveness) are inapplicable. Perspectives of Malingering in the Forensic Context A heuristic typology is proposed to explain differences in how forensic psychologists approach the evaluation of response styles. Three main perspectives are identiﬁed: intuitional, standard, and specialized. These perspectives are considered in the context of malingering. The intuitional perspective presupposes that malingering and other response styles will be recognizable based on clinical acumen without the need for empirically validated strategies, scales, and indicators. Despite its lack of empirical validation, we suspect that the intuitional perspective is wide- spread in forensic practice. A key example is found with competency to stand trial evaluations. Despite nearly three decades of research on competency evaluations (Rogers, 2001), malingering and related response styles have been virtually ignored. Even the most recent and best-funded competency measure, MacArthur Competency Assessment Tool–Criminal Adjudication (Poythress et al., 1999), implic- itly adopted an intuitional perspective for malingering. While acknowledging that response styles may confound compe- tency evaluations, no indices of any kind are provided (see Poythress et al., 1999, p. 5). The standard perspective routinely evaluates malingering and defensiveness on the basis of traditional tests and mea- sures. The advantages of this approach are twofold: (a) highly efﬁcient use of customary measures for dual purposes (e.g., psychopathology and feigning), and (b) application of empirically tested strategies. The major shortcoming of the standard perspective is that traditional testing lacks the diag- nostic utility for making clinical determinations. The most common examples of the standard perspective involve multiscale inventories (e.g., the Minnesota Multiphasic Personality Inventory 2 [MMPI-2; Butcher, Williams, Graham, Tellegen, & Kaemmer, 1989]) and intelligence testing (i.e., predominantly the Wechsler Adult Intelli- gence Scale–Revised [WAIS-R; Weschler, 1981] rather than WAIS-III; Weschler, 1997). The specialized perspective supplements traditional test- ing with measures that are speciﬁcally designed for the as- sessment of response styles. Common forensic examples include the Structured Interview of Reported Symptoms (SIRS; Rogers, Bagby, & Dickens, 1992) for feigned mental disorders and the Portland Digit Recognition Test (PDRT; Binder & Willis, 1991) for feigned cognitive im- pairment. Despite the additional expenditure time, the spe- cialized perspective is generally superior to the standard perspective in its classiﬁcatory accuracy. The specialized perspective is recommended as the necessary model for the determination of feigning in both clinical and forensic practice. Explanatory Models of Malingering When conducting evaluations and rendering conclusions, forensic psychologists are likely to be inﬂuenced by explana- tory models of malingering. Explanatory models attempt to explain why individuals strive to malinger psychological and physical impairment. Rogers (1990a, 1990b) outlined three explanatory models of malingering: pathogenic, crimi- nological, and adaptational. Several prototypical analyses (Rogers, Sewell, & Goldstein, 1994; Rogers, Salekin, Sewell, Goldstein, & Leonard, 1998) provide general support for Conceptual Issues 111 these explanatory models as distinct explanations for malin- gering. A synopsis of the three explanatory models of malin- gering is provided. The pathogenic model assumes that the underlying moti- vation is an ineffective attempt to control the symptoms and clinical presentation of a chronic and progressive men- tal disorder. With increased impairment, intentionally pro- duced symptoms become gradually less deliberate, until they are involuntary and unintended. The pathogenic model predicts that feigning is an ineffectual attempt at adjust- ment that eventually is resolved by the patient’s further deterioration. The criminological model is championed by the Diagnos- tic and Statistical Manual of Mental Disorders ( DSM-IV-TR; American Psychiatric Association, 2000); it assumes that the primary motivation is characterological. Namely, antisocial persons faced with legal difﬁculties will attempt to garner un- warranted advantages either in circumstances (e.g., a hospital rather than a prison) or material gain (e.g., ﬁnancial settle- ment). Antisocial persons are presumed to be generally de- ceptive. With malingering viewed as a variant of deception, the criminological model predicts an intermittent use of malingering based on situational opportunities. The adaptational model assumes that the person perceives the circumstances as adversarial and considers malingering to be a feasible alternative. This model avoids the monistic notions of “mad” (pathogenic) or “bad” (criminological) and views malingering in terms of a cost-beneﬁt analysis. The adaptational model views malingering as a situational re- sponse based on an appraisal of alternatives. Rogers, Salekin, et al. (1998) found that the pathogenic model was low in prototypicality for both males and females in forensic evaluations. In contrast, both the adaptational and criminological models achieved moderately high prototypi- cal ratings for forensic cases. A potential danger of the crim- inological model is that forensic psychologists may attempt to use this explanatory model as a detection model. The DSM-IV-TR indices only raise the suspicion of malin- gering; they do not constitute formal criteria for the classiﬁ- cation of malingering. Even for suspicions of malingering, these indices (i.e., antisocial personality disorder, medicole- gal evaluation, uncooperativeness, and results inconsistent with objective ﬁndings) falter on both conceptual and empir- ical grounds. Rogers (1997) provides a conceptual analysis of their major shortcomings. Even in defending the DSM-IV- TR indices, LoPiccolo, Goodkin, and Baldewicz (1999) con- ceded most of these shortcomings. Empirically, DSM-IV-TR indices fail entirely even for screening purposes. Their use in a criminal forensic setting resulted in a false-positive rate of approximately 80% (Rogers, 1990a). Explanatory Models of Defensiveness Rogers and Dickey (1991) proposed that explanatory models of defensiveness could be extrapolated from the malingering literature, at least in the case of sex offenders. The pathogenic model is the least persuasive; psychodynamic formulations have suggested that loss of ego functions may result in uncon- scious denial. More persuasive explanations were the crimi- nological and adaptational models, suggesting that denial and gross minimization might result from either a general criminal orientation or an attempt to cope with highly adversarial cir- cumstances. As noted by Rogers and Dickey, sex offenders often are placed in an irresolvable bind: Honesty, disclosing the true extent of their paraphilac behavior, is likely to result in negative sanctions based on the extent of criminal activity; defensiveness, grossly minimizing the true extent of their paraphilac behavior, is likely to result in negative sanctions because nondisclosure is viewed as a barrier to treatment. Sewell and Salekin (1997) expanded on Rogers and Dickey’s (1991) framework and proposed a socioevaluative model of defensiveness. For offenders, especially sex offend- ers, evaluations are consistently linked with punishment and ostracism. The socioevaluative model posits that evaluatees react to the likely threat of a negative outcome and attempt to protect themselves. The socioevaluative model is similar to the adaptational model in its appraisal of a highly adversarial context. It is distinguished from the adaptational model in its generalized reaction. Even when “there is nothing to lose,” the socioevaluative model predicts a generalized response of defensiveness based on past learning. Under the rubric of cognitive distortions, the notion of self-deception has been considered, especially with sex offenders. According to Vanhouche and Vertommen (1999), cognitive distortions involve “learned assumptions” and “sets of beliefs and attitudes” (p. 164) that serve in the denial and minimization of criminal behavior. In the course of the eval- uation, denials of responsibility may be inﬂuenced by “self- deceptive” beliefs (e.g., educative goals of incest). However, such denials are unlikely to explain the overall defensiveness expressed by many offenders. The understanding of defensiveness in forensic practice is constrained by the focus on sex offenders. Although extrapo- lations to other forensic populations are possible, explanatory models of defensiveness remain in their initial stages of development and validation. Misassumptions about Malingering and Dissimulation Forensic psychologists are not immune to common misas- sumptions about malingering and other response styles. 112 Evaluation of Malingering and Deception Moreover, forensic psychologists must be prepared to ad- dress erroneous assumptions made by others in the legal sys- tem. Five key misassumptions, common to forensic practice, are outlined: • Malingering is very rare . Equating infrequency with in- consequentiality, some clinicians neglect the evaluation of malingering except in very obvious cases. Estimates (Rogers et al., 1994, 1996) based on more than 500 foren- sic experts suggest that malingering is not rare, but likely occurs in 15% to 17% of forensic cases. • Malingering is very common . Fueled by fears of fraud and injustice, certain attorneys (e.g., defense counsel in civil litigation and prosecutors in criminal matters) suspect that malingering and dissimulation are very prevalent. Despite speculation that the majority of forensic evaluatees may be malingering, the best estimates (Rogers et al., 1994, 1996) indicate this is not the case. • Malingering occurs at a predictable rate . If stable base rates could be achieved, the classiﬁcation of malingering and other response styles could be improved. In a desire to improve classiﬁcation, clinicians often ignore the fact that malingering does not occur at predictable rates. The best available data (Rogers et al., 1996) found highly variable rates ( SD  14.44). Even within the same setting, rates are likely to vary markedly based on referral issues (see Rogers & Salekin, 1998). • Malingering is most likely to occur in persons with antiso- cial personality disorder (APD) . Psychopaths and persons with APD likely engage in deception (Rogers & Cruise, 2000), but no data indicate an increased likelihood for malingering in forensic settings. This unsupported as- sumption likely is based on a methodological artifact: Because most forensic studies are conducted in criminal settings, the facile connection between malingering and APD is understandable. • Malingering and mental disorders are mutually exclusive . Neither malingering nor mental disorders offer any natural immunity to the other. Some individuals with valid psy- chopathology “gild the lily” by adding feigned symp- toms. Most clinicians are willing to acknowledge the co-occurrence of malingering and mental disorders; how- ever, many forensic reports do not address the mental dis- orders after malingering has been determined. Applications to Forensic Practice Determinations of malingering often supersede all other clini- cal issues. When a forensic psychologist concludes that a person is malingering, this opinion is likely to invalidate all claims by that person, destroying his or her credibility. Because of its overshadowing importance, forensic psycholo- gists carry a further responsibility to ensure the accuracy of their conclusions with respect to malingering. We recom- mend that the classiﬁcation of malingering should never rely on a single indicator. In addition to conﬁrmation by multiple sources, forensic psychologists should systematically exclude alternative explanations (e.g., factitious disorders or irrelevant responding) in their determinations of malingering. To avoid misclassiﬁcations based solely on idiosyncratic data, Rogers and Shuman (2000) put forth the following forensic guideline: No determination of malingering should rest solely on tradi- tional interviews . The classiﬁcation of malingering often appears dispositive of the verdict. Given this observation, what are the responsi- bilities of a forensic psychologist who believes that another expert’s conclusions about the presence of malingering were inaccurate? That psychologist bears the onerous responsibil- ity of comprehensively evaluating the issue of malingering. If the data continue to support his or her conclusion (i.e., the absence of malingering), then great care must be taken to marshal this evidence in a manner to convince the trier of fact. In general, forensic psychologists should assume an un- even playing ﬁeld, with a much heavier burden of disproving than proving malingering. In sentencing and postverdict criminal evaluations, defen- siveness is often the preeminent issue. Courts and other adju- dicative bodies are concerned that dangerous persons not be released prematurely based on minimization of their psycho- logical impairment. Forensic psychologists must exercise a rigorous standard in conducting these evaluations, compa- rable to malingering determinations. EMPIRICAL ISSUES The clinical assessment of response styles rests solidly on their validation. As demonstrated in this section, no single re- search design is sufﬁcient to validate measures of response style . With respect to preparing for testimony, Rogers (1997) provided a thorough review of these research designs. The purpose of this section is to provide forensic psychologists with a brief summary of research designs and their relevance to the assessment of response styles. Basic Designs Three designs predominate the validation of clinical mea- sures for the evaluation of malingering and defensiveness. Empirical Issues 113 Simulation Design Simulation studies use an analog design in which participants are randomly assigned to simulator and control conditions. For feigning studies, the addition of a clinical comparison sample is essential; otherwise, researchers cannot ascertain whether differences are attributable to feigning or to gen- uine disorders. With appropriate debrieﬁng, the simulation design excels at internal validity but has limited external validity. Known-Groups Comparison Known-groups studies are conducted with independently classiﬁed malingerers who are compared with genuine pa- tients. The challenge is the identiﬁcation of actual malinger- ers in sufﬁcient numbers for research. The known-groups comparison excels at external validity but has limited internal validity. Differential Prevalence Comparison Differential prevalence studies assume that certain groups will have a higher prevalence of a speciﬁc response style (e.g., forensic patients for feigning and job applicants for defensiveness). Group differences have little practical sig- niﬁcance without knowing what is the proportion of dis- simulation in different groups, or whether deviant scores represent dissimulation. Differential prevalence comparison fails to establish internal validity and has limited external validity. Bootstrapping Comparisons A fourth design, bootstrapping comparisons, recently has been observed in studies of feigned cognitive impairment. Persons identiﬁed by deviant scores on other measures of feigning are compared to those without these deviant scores. The key issue with bootstrapping comparisons is the selec- tion of measures with nearly perfect speciﬁcity, so that the “feigning” group does not contain genuine patients. Experi- mental rigor can be increased through the classiﬁcation based on several measures representing different detection strategies. The best validation for measures of response styles is a combination of studies representing simulation design and known-groups comparisons. This combination maximizes both internal (simulation design) and external (known- groups comparison) validity. Forensic psychologists should take particular care to select measures with known-groups comparisons, because these studies are frequently omitted from the test validation. Incremental Validity Psychologists often believe that a convergence of ﬁndings across different measures contributes to incremental validity. As a counterposition, Sechrest (1963) demonstrated in his seminal article that the single best measure often is not im- proved by adding additional measures. As a forensic exam- ple, Kurtz and Meyer (1994) found that the SIRS was more accurate for the classiﬁcation of feigning than either the MMPI-2 alone or the combination of the SIRS/MMPI-2. Forensic psychologists must decide whether to use the single best measure or a convergence of measures in establishing classiﬁcatory accuracy for response styles. We recommend that forensic psychologists employ multi- ple indices from different measures when malingering is sus- pected. Because the determination of malingering carries such grave consequences, its assessment should be compre- hensive. The results should be analyzed on two parameters: domain and detection strategies. Feigning can be divided into at least three broad domains (i.e., mental disorders, cognitive impairment, and medical illness) that differ substantially in clinical presentation. For each domain, detection strategies can be identiﬁed for the clinical classiﬁcation of malingering; these detection strategies vary in the extent of their validation and accuracy of classiﬁcation. To facilitate this analysis, subsequent sections of this chapter address domains and their respective detection strategies. Clinicians must be ready to grapple with both convergent and divergent results. What about convergent results? With consistent results from well-validated strategies derived from dissimilar mea- sures, forensic psychologists likely will have conﬁdence in their conclusions about response styles. Such conﬁdence should not be confused with increased accuracy (i.e., incre- mental validity); unless empirically demonstrated, psycholo- gists cannot conclude a higher level of accuracy. What about generally consistent results? The most com- mon ﬁnding in forensic evaluations is that most of the indica- tors agree; however, one or more indices of response styles do not ﬁt with the other indicators. One temptation is to ignore or explain away the discrepant ﬁndings. A more pru- dent course is to evaluate the results, taking into account the accuracy of the measures and the validity of the detection strategies. For example, a “nonfeigning” classiﬁcation on the SIRS has an excellent positive predictive power that is likely to outweigh a more nebulous elevation on an MMPI-2 validity scale. In addition, some detection strategies (e.g., symptom validity testing) are much more robust than others 114 Evaluation of Malingering and Deception (e.g., forced choice testing); their comparative validity can be taken into account in making determinations. What about inconsistent ﬁndings? The ﬁrst possibility is that the results are domain-speciﬁc. For example, an evalua- tee with major depression (a mental disorder domain) may feign problems with attention, concentration, and immediate memory (a cognitive impairment domain) in the context of a disability evaluation. Sometimes, these cases can be resolved based on the accuracy of measures and relative validity of detection strategies. In other cases, the only logical decision is that the results are inconclusive. Forensic psychologists should be aware that some clinicians adopt a “fall-through-the-ice” mentality: Any failure (e.g., an indicator of feigning) is viewed as decisive evidence of a per- vasive response style. Like falling through the ice, the results are immediately catastrophic and summarily generalized. This mentality is empirically unwarranted and is probably more illuminating about the clinician than the evaluatee. MALINGERING OF MENTAL DISORDERS Detection Strategies Rogers (1997) and Rogers and Vitacco (in press) provide ex- tensive descriptions of detection strategies for feigned mental disorders. The purpose of this section is to highlight these pri- mary strategies. These strategies are important for under- standing how scales and speciﬁc indicators are utilized in the assessment of malingering. Using detection strategies, a con- ceptually based approach combines theory and empiricism. It offers judges and juries more than simply numbers and cut scores; it supplies the underlying logic and rationale for how the scales were constructed and the classiﬁcation was reached. A distillation of eight detection strategies for feigned psy- chopathology is enumerated: 1. Rare Symptoms . Items in this strategy are very infre- quently endorsed by clinical populations. Malingerers often are unaware that certain symptoms are infrequently experienced. Rare symptoms represent one of the most robust detection strategies. 2. Improbable Symptoms . A minority of malingerers report or endorse symptoms that have a fantastic or preposterous quality. When a pattern of improbable symptoms is en- dorsed, the credibility of the evaluatee’s reporting is brought into question. 3. Symptom Combinations . Many symptoms commonly occur alone but rarely are paired together (e.g., grandios- ity and increased sleep). To foil this strategy, malingerers would need to have a sophisticated understanding of psychopathology. 4. Symptom Severity . Even severely impaired patients expe- rience only a discrete number of symptoms as “unbear- able.” Malingerers often are unable to estimate which symptoms and how many symptoms should have extreme severity. 5. Indiscriminant Symptom Endorsement . When asked about a broad array of psychological symptoms, some malinger- ers do not respond selectively but endorse a large propor- tion of symptoms. 6. Obvious versus Subtle Symptoms . Malingerers tend to en- dorse a high proportion of obvious symptoms (i.e., clearly indicative of a mental disorder). Obvious symptoms are either considered alone or in relation to subtle symptoms (i.e., “everyday” problems, not necessarily indicative of a mental disorder). When compared to genuine patients, malingerers often report a higher proportion of obvious symptoms. 7. Erroneous Stereotypes . Many persons have misconcep- tions about symptoms associated with mental disorders. When displaying erroneous stereotypes, persons feigning mental disorders can sometimes be detected. 8. Reported versus Observed Symptoms . Marked discrepan- cies between the person’s own account and clinical obser- vations appear useful in the detection of malingerers when standardized measures are used. The risk of this approach is that many genuine patients lack insight about their psy- chopathology. These eight detection strategies account for most of the systematic approaches to feigned mental disorders and constitute the framework for the evaluation of malingered symptomatology. Several additional strategies have been ex- plored. Morel (1998) used forced-choice testing (see section on Malingering and Cognitive Impairment) to test for feigned posttraumatic stress disorder; the bogus effects of emotional numbing were evaluated in a two-choice paradigm. Wildman and Wildman (1999) explored whether malingerers might be detected by their overly virtuous self-descriptions. Featured Measures A single chapter cannot comprehensively review the broad array of psychological measures adapted or developed for the assessment of feigned mental disorders. Therefore, this section addresses three featured measures that have been ex- tensively validated. Featured measures include two multi- scale inventories and one structured interview. Malingering of Mental Disorders 115 As a general caution, forensic psychologists should closely inspect test manuals and validation studies prior to using any test for feigned mental disorders. For example, we have observed numerous forensic reports attempting to use the Millon Clinical Multiaxial Inventory III (MCMI-III; Millon, 1994; Millon, Davis, & Millon, 1997) to assess feign- ing. Is this use warranted based on a careful examination of the MCMI-III’s validation? The answer is clearly negative. For example, the debasement index is promoted as a fake-bad scale for detecting persons attempting to appear psychologi- cally impaired. Close inspection reveals the following: (a) both the 1994 and 1997 MCMI-III test manuals neglected the validation of the MCMI-III debasement index; (b) the MCMI-III debasement index appears confounded by psy- chopathology (i.e., 9 clinical scales correlate  .75 in the normative sample); and © extrapolations from MCMI-II research would be inappropriate because only 19 of 46 (41.3%) MCMI-II items were retained on the MCMI-III debasement index. More than ﬁve years after the MCMI-III’s publication, research (Daubert & Metzler, 2000; Thomas- Peter, Jones, Campbell, & Oliver, 2000) is now beginning to emerge on the debasement index and feigning; more exten- sive research is needed before its use in forensic evaluations. Importantly, validational problems are not limited to the MCMI-III; forensic psychologists are urged to scrutinize closely the validation of all response style measures. Minnesota Multiphasic Personality Inventory-2 A large array of validity indices has been developed to evalu- ate whether MMPI-2 protocols have been feigned. Table 7.1 provides a summary of indices for the detection of both feigning and defensiveness. Summary data include the range of cut scores, available data on effect sizes, and a brief de- scription of scale development. Forensic psychologists are likely to be in a quandary about which MMPI-2 indices should be employed for the evalua- tion of malingering. Standard MMPI-2 texts provide con- ﬂicting conclusions. Championing the traditionalist model, Butcher and Williams (1992) advocated the use of the F and Fb scales, virtually ignoring specialized scales for feigning. Graham (2000) also emphasized the use of traditional MMPI-2 indicators. However, he endorsed one specialized indicator (Fp) and discommended the use of other specialized indices. In stark contrast, Greene (1997, 2000) embraces a comprehensive model, with the use of both traditional and specialized indices of malingering. Both models are critically evaluated in subsequent sections. The traditionalist model of malingering, beyond history and convention, has several advantages that must be considered. In an MMPI-2 meta-analysis, Rogers, Sewell, and Salekin (1994) found the F and Fb had several of the largest effect sizes (2.56 and 1.85, respectively) for feigning when compared to clinical TABLE 7.1 Description of MMPI-2 Validity Indices for Feigning and Defensiveness Effect Sizes a Scale Items Cut Scores r a Feigning Defensive Scale Development Feigning Indices b F 60 8–30 1.00 2.56 Infrequency in normative samples. Fb 40 9–25 .86 1.85 Infrequency in normative samples. Fp 27 NA .75 NA Infrequency in inpatient samples. Dsr2 32 13–28 .61 c 1.54 Stereotypes of mental disorders. FBS 43 NA NA NA Rational: personal injury claims. LW 107 40–67 .84 1.38 Rational: urgent clinical issues. O-S NA d 74–190 .81 2.30 Rational: obvious versus subtle symptoms. Defensiveness Indices e L 15 6–9 .43 .94 Rational: borrowed from earlier scales. K 30 17–22 1.00 .90 Empirical:  30% for defensive patients. Mp 34 16–20 .48 f 1.42 Empirical: identify best impression. Wsd 33 21–23 .28 1.60 Empirical: socially desirable items. Esd 39 35–36 .76 .67 Rational: socially desirable items. S 52 NA .88 NA Differential prevalence with pilots. g Note : NA  not available. a Correlations are reported in Greene (2000) for clinical samples between (a) feigning indices and Scale F and (b) defensiveness indices and Scale K. b Effect sizes and range of cut scores reported in MMPI-2 meta-analyses of feigning (Rogers, Sewell, & Salekin, 1994). c Dsr2 is not reported in Greene (2000); this estimate is based on the original 58-item Ds2 from which the Dsr2 was extracted. d Uses T-score transformations of subscales. e Effect sizes and range of cut scores reported by Baer, Wetter, and Berry (1992). Please note that this meta-analysis is based on the original MMPI and should be viewed only as a general benchmark for MMPI-2 performance. f Based on slightly modiﬁed Od scale. g Pilot applicants were assumed to have a high proportion of defensive persons; they were compared to a normative sample. 116 Evaluation of Malingering and Deception populations. Other research (Bagby, Buis, & Nicholson, 1995; Timbrook, Graham, Keiller, & Watts, 1993) has used hierar- chical multiple regression to evaluate whether the use of addi- tional validity indices would add incremental validity (i.e., account for more of the variance). These studies concluded that the F scale alone appeared to be the most predictive of malin- gering. A ﬁnal advantage of the traditionalist model is its sim- plicity; forensic psychologists do not have to explain to the courts potential

bundet

ly conﬂicting MMPI-2 data. The traditionalist model also has signiﬁcant limitations in the evaluation of feigning. Its primary constraints are outlined: • Both F and Fb are based on the same strategy (rare symp- toms); this overreliance on a single strategy is a weakness of the traditionalist model. This shortcoming is accentu- ated by the ﬂawed development of both F and Fb scales. Items on both were selected if they were infrequently en- dorsed by normative (nonclinical) samples. The critical comparison between genuine and bogus disorders was omitted. The fact that patients with genuine disorders often have marked F and Fb elevations is directly attribut- able to its ﬂawed development. • Studies indicating that specialized MMPI-2 indices do not add incremental validity to scale F have serious method- ological constraints. Because of unaddressed issues with multicolinearity (e.g., 25% of items on F also appear on O-S), results likely are skewed toward nonsigniﬁcance. Also, forensic psychologists are primarily interested in whether the use of specialized indices improves accuracy of classiﬁcation for feigning and genuine disorders. This matter was left unaddressed by these multiple regression studies. The comprehensive model provides, in unambiguous cases, an array of empirically validated strategies for the classiﬁca- tion of feigned and genuine disorders. Forensic psychologists can present data to the court based on multiple detection strate- gies: (a) rare symptoms (i.e., F, Fb, and Fp); (b) erroneous stereotypes (i.e., Dsr2); © overendorsement of obvious symp- toms (i.e., O-S); and (d) indiscriminant endorsement of severe symptoms (i.e., LW). Convergent data from multiple strategies are often compelling, especially because they minimize the limitations found with any particular scale, such as multiple interpretations for marked elevations on the F scale. The challenges of the comprehensive MMPI-2 model are how to understand discordant data and how to explain appar- ent discrepancies to the court. Validity indices on forensic protocols sometimes range from low to marginally elevated to extremely elevated. Occasionally, the pattern of scores is clearly understandable in light of other clinical data. In many cases, the range of validity indices presents a conundrum to forensic psychologists, who must explain their uncertainties to the court. The incremental validity of MMPI-2 indices remains unresolved. Forensic psychologists will opt for either the simple traditionalist model or the more complex comprehen- sive model. They must weigh the risks of overlooking valu- able data (traditionalist model) against the possibilities of unexplainable discrepancies (comprehensive model). For either choice, forensic psychologists must have a clearly articulated rationale. For clinicians seeking guidance with this decision, one recommended course of action is a two-phase approach. Con- sistent with the traditionalist model, the ﬁrst phase comprises standard indicators, which are routinely evaluated in all forensic cases. When standard indicators are marginally or markedly elevated, the second phase consists of 3 to 4 spe- cialized indices, which likely include the Fp, Dsr2, LW, and O-S. In marginal cases, a second phase may resolve ambigu- ities. In marked cases, additional data are sought to conﬁrm or disconﬁrm the initial ﬁndings. Forensic psychologists should be aware of common MMPI-2 missteps . An important responsibility of forensic psychologists is to evaluate the conclusions drawn by other clinicians from test data, including the MMPI-2. A careful scrutiny of MMPI-2 reports reveals three common missteps in using the MMPI-2 for the assessment of feigning: • Inconsistent Proﬁles . A random or otherwise inconsistent proﬁle is likely to have extreme elevations of MMPI-2 feigning indices. Although malingerers may deliberately respond inconsistently, psychologists generally cannot rule out other common reasons for inconsistent proﬁles, including a haphazard completion of the answer sheet with- out carefully reading the test items. The very rare exception occurs when the MMPI-2 feigning indices are consistently above chance endorsement (e.g., raw F  40). • Incompatible Proﬁles . Clinicians sometimes observe that an MMPI-2 proﬁle is incompatible with other documented ﬁndings and erroneously conclude that the client is feign- ing. This grave error is based on the misassumption that certain proﬁles or scale elevations are nearly always linked with certain diagnoses or symptoms. The simplest rebuttal of this error is that a within normal limits (WNL) proﬁle with no clinical elevations is the most common proﬁle among inpatients and outpatients (Greene, 2000). • Validity Scale Conﬁgurations . Historically, the relative elevation of scale F in relationship to scales L and K was interpreted as indicative of feigning. The validity of this interpretation has not been established. Interestingly, Greene (2000) suggested that this conﬁguration is desir- able for psychological intervention. Malingering of Mental Disorders 117 Personality Assessment Inventory The PAI now rivals the MMPI-2 as a multiscale inventory for the evaluation of malingering and other response styles. Although more malingering studies have been conducted with the MMPI-2, the PAI has several important advantages: • The PAI validity scales are nonoverlapping. In contrast, speciﬁc MMPI-2 validity scales overlap with each other and with clinical scales, thereby confounding their inter- pretation and classiﬁcatory utility. • The PAI validity scales typically use a standardized cut score for feigning. In contrast, MMPI-2 validity scales uti- lize a broad range of cut scores. This range diminishes the effectiveness of the MMPI-2’s classiﬁcation of feigned and nonfeigned proﬁles. • The PAI validity indices were tested with both simulation and known-groups designs. In contrast, the MMPI-2 validity scales have very limited testing with actual cases of suspected malingerers. Three PAI indices are used to evaluate feigning. The stan- dard indicator, NIM scale (  11), is based on items infre- quently endorsed by normative and patient samples. More recently, Morey (1996) developed the Malingering index ( 5), composed of eight conﬁgural rules using PAI scales and subscales. Finally, Rogers, Sewell, Morey, and Ustad (1996) cross-validated a discriminant function, which was derived from 20 loadings on PAI scales and subscales. Pri- mary references for the feigning on the PAI include a recent review by Morey and the known-groups comparison by Rogers, Sewell, Cruise, Wang, and Ustad (1998). The following guidelines are based on a synthesis of data from simulation research and known-group designs. In foren- sic evaluations, the guidelines are provided: • Rule out Feigning . A NIM score  77T (raw score  8) indicates a low probability that the evaluatee is feigning. • Screen for Feigning . Marked elevations on NIM (77T to 109T) indicate the need to evaluate thoroughly issues of feigning. Forensic psychologists should examine the PAI Malingering index and specialized measures (e.g., the SIRS) for the assessment of feigning. • Likely Feigning . Extreme elevations on NIM ( 110T) or the Malingering index (  5) indicates a strong likelihood of feigning. The PAI should not be used as the primary measure to evaluate feigning, although low scores may be effective at eliminating cases unlikely to be malingering. For “likely feigning,” the strengths of extreme elevations are twofold: (a) a very low proportion of false-positives (NIM  .02; Malingering index  .01), and (b) high (NIM, PPP  .82) to very high (Malingering index, PPP  .92) classiﬁcations when these cut scores are met. The problem is that relatively few feigners achieve such extreme elevations; the sensitivity estimates are .10 and .09, respectively. Therefore, these ex- treme scores are likely to miss 9 out of 10 feigners. The PAI discriminant function is not recommended for forensic evaluations. Although highly effective in clinical evaluations, its accuracy was substantially diminished when applied to forensic patients in a known-groups comparison. Its sensitivity plummets from .84 to .51, and its speciﬁcity declines from .89 to .72. Even in clinical settings, psycholo- gists are cautioned to inspect the PAI clinical proﬁle before using the discriminant function. A case has been identiﬁed in which all the clinical scales were unelevated and the individ- ual was not feigning, despite a positive ﬁnding on the dis- criminant function. Structured Interview of Reported Symptoms The SIRS is a structured interview for the systematic as- sessment of feigned mental disorders. Rogers, Bagby, and Dickens (1992) outline its general validation; forensic psy- chologists may wish to consult Rogers (2001) for the most recent update of SIRS validity studies. Unlike the MMPI-2 and PAI, the SIRS was developed speciﬁcally for the assess- ment of feigning and related response styles. This focus has resulted in extensive research for both the development of strategy-based scales and their clinical implementation. The SIRS’s primary scales employ all detection strategies described previously, with the exception of “erroneous stereo- types.” Persons feigning mental disorders typically are classi- ﬁed based on three or more scales in the probable feigning range. Less frequently, feigners will have extreme elevations (i.e., deﬁnite feigning range) on one or more primary scales. Forensic psychologists classify SIRS proﬁles into one of three general categories: feigning, indeterminate, and nonfeigning. The principal features of the SIRS are summarized: • Validation. The SIRS has been extensively validated not only by its developers but also by independent researchers (see Rogers, 2001; Rogers et al., 1992). Importantly, the SIRS combines both simulation design and known-groups comparisons to optimize its validation. The SIRS has also been validated with clinical, forensic, and correctional populations. • Clinical Applications . A major emphasis on the SIRS is the individual classiﬁcation of evaluatees with respect to response styles. To reduce misclassiﬁcations, an indeter- minate category was implemented for marginal cases. In 118 Evaluation of Malingering and Deception the classiﬁcation of feigners, the positive predictive power is very high to minimize misclassiﬁcations. Classiﬁcation rules are also available for nonfeigning proﬁles. • Coaching. A particular concern of forensic evaluators is whether evaluatees are coached by others or otherwise “educated” about a response style measure and its scales. Coaching participants on the SIRS strategies does reduce elevations; however, most participants still have marked elevations on the SIRS primary scales. • Generalizability . Available research (see Rogers, 2001) indicates that the SIRS appears to function equally well across gender, ethnicities commonly encountered in foren- sic settings, and type of setting. An important caution is that the SIRS has not been vali- dated for repeat administrations, especially across brief inter- vals. We have observed several forensic cases in which an expert, apparently dissatisﬁed with the results from an ear- lier expert, readministered the SIRS. One grave concern is whether the evaluatee had access to the results of the previ- ous report (written or oral) or reasonably inferred this feed- back from general comments made by his or her attorney. This type of speciﬁc feedback on past SIRS performance may invalidate subsequent administrations. In summary, the SIRS is probably the best-validated mea- sure for the assessment of feigned mental disorders. In foren- sic cases in which malingering is suspected, the SIRS should be a standard component of the assessment. Given the accu- racy of its individual classiﬁcations, results of the SIRS should be weighted heavily when discrepancies occur in the assessment of malingering. DEFENSIVENESS AND MENTAL DISORDERS Overview Defensiveness, involving the denial and minimizing of men- tal disorders, is often cast into a secondary role in forensic evaluations. Cases of potential malingering take center stage because of concerns within the criminal justice system that criminal defendants may evade their punishments or that civil litigants may reap undeserved rewards. Less attention is paid to defensive clients who may be deliberately underreporting their symptomatology, possibly motivated by the stigmatiza- tion of mental illness. As an extreme example, some criminal defendants would rather face the death penalty than admit that they are mentally disordered. Methods of assessing defensiveness in forensic evaluations are not nearly as well developed as those for malingering. Three major reasons contribute to our limited knowledge of defensiveness: • Defensiveness is difﬁcult to assess because clients simply deny or minimize their symptomatology. • Defensiveness is often difﬁcult to distinguish from “lack of insight.” Many chronic patients, especially those with psychotic disorders, do not recognize their symptoms and therefore do not report them. • Defensiveness has been largely neglected by recent foren- sic research. This section focuses on two measures that have been used with varying degrees of success in the assessment of defen- siveness. These measures consist of the MMPI-2 and Paulhus Deception Scales (PDS; Paulhus, 1998). Minnesota Multiphasic Personality Inventory The MMPI-2 has two traditional scales and a handful of spe- cialized scales for the evaluation of defensiveness (Baer, Wetter, & Berry, 1992). Beyond the traditional scales (L and K), this review focuses on two highly effective specialized scales (Wsd and Mp) as well as a recently developed and highly touted scale (S). Table 7.1 summarizes the pertinent information about these ﬁve scales. Baer et al. (1992) conducted a meta-analysis on 25 studies with a ﬁrst-rate review of defensiveness on the original MMPI. As an important and unexpected ﬁnding, Baer and her colleagues found that Wiggins’ Social Desirability (Wsd) scale and the Positive Malingering scale (Mp) outperformed the traditional defensiveness scales, L and K. More recent MMPI-2 studies have highlighted the importance of special- ized scales in the determination of defensiveness. Key ﬁnd- ings are summarized: • Baer, Wetter, and Berry (1995) found that traditional scales are vulnerable to coaching; tips on how to avoid de- tection foiled scales L and K (i.e., negligible effect sizes of  .06 and  .04, respectively). In contrast, Wsd produced a moderate effect size with coaching (.86). • Studies have indicated that specialized indices of defen- siveness add incremental validity. Specialized indices in- clude Wsd, S, Edwards Social Desirability (Esd), and Other Deception (Od), which add incremental validity to the traditional scales (Baer, Wetter, Nichols, Greene, & Berry, 1995; Bagby et al., 1997). As a concrete example, Baer et al. (1995) found that a discriminant function based on scales L and K produced a 78% classiﬁcation, while the addition of Wsd and S improved this classiﬁcation to 90%. Feigned Cognitive Impairment 119 A critical issue for forensic psychologists is whether MMPI-2 indices are effective in forensic cases in which defensiveness is likely to occur. Bagby, Nicholson, Buis, Radovanovic, and Fidler (1999) addressed this issue indi- rectly by comparing the clinical proﬁles of defensive and non- defensive parents in custody and access evaluations. Using a variation of standard indicators (i.e., L  K), they found vir- tually no difference between defensive and nondefensive proﬁles ( M effect size  0.00). With specialized indices (Wsd and S), very modest effect sizes were found ( M effect size  .17). The use of single cut scores may have modest util- ity when most parents are engaging in some level of defen- siveness (i.e., overall M elevations for clinical scales  51.2). Alternative explanations are that most parents in child cus- tody litigation do not have psychological impairment, or their psychological impairment is not captured by the MMPI-2. The basic recommendation for forensic practice is that psychologists routinely score Wsd in all cases. In addition to robust effect sizes, the Wsd has two major advantages: it is less vulnerable to coaching than other indices, and it has a narrow range of cut scores. Other specialized indices (S and Mp) are likely to be used selectively in cases where defen- siveness is suspected. Paulhus Deception Scales Paulhus (1998) developed the PDS, composed of two scales for measuring defensiveness. The purpose of each scale is examined in detail. The Impression Management (IM) scale is intended to measure deliberate efforts at social desirability, although the scale correlates moderately with personality traits of con- scientiousness and agreeableness. Under “high-demand” circumstances, scores on the IM scale tend to increase. Com- plicating the interpretation of the IM scale is the ﬁnding that highly religious persons tend to have very high scores (see Paulhus, 1998, p. 9, note 1); the question remains whether re- ligious persons deliberately engaged in social desirability or the IM scale is confounded by devout beliefs. The Self-Deceptive Enhancement scale (SDE) is intended to measure “an unconscious favorability bias closely related to narcissism” (Paulhus, 1998, p. 9). High SDE scores are associ- ated with self-described personal adjustment; observers vary in their descriptions from conﬁdent and well-adjusted to arro- gant and domineering. Perhaps the most controversial part of the SDE scale is its designation of an “unconscious” bias. Some forensic psychologists are likely to be unwilling to adopt the PDS’s explicit psychoanalytic framework. Moreover, the admissibility of expert evidence following Daubert must take into account the falsiﬁability of scale interpretations, a formidable challenge for the unconscious formulation for the SDE scale.Although the test manual reports factor analytic results supporting two dimensions, it does not provide evi- dence that the second dimension was unconscious. In citing his earlier research, this factor was described as a portrayal of “exaggerated mental control or dogmatic overconﬁdence” (Paulhus, 1998, p. 23). This description leaves open the ques- tion of unconscious motivation. Results of Pebles and Moore (1998) further question the validity of the SDE scale. When simply asked to “make a good impression,” participants easily doubled their scores on SDE from 5.5 to 11.6. The ability of uncoached participants to achieve an extreme elevation (T score  85) casts doubt on the SDE as an unconscious measure of self-deception. Salekin (2000) provided a useful summary of the PDS in relation to forensic practice. He observed problems in under- standing the SDE scale in relation to psychopathy (e.g., grandiosity and superﬁcial charm) and narcissism. He also noted the absence of cross-validated cutting scores and the lack of research with clinical-forensic samples. Amplifying on this latter point, an inspection of the test manual suggests that the PDS validation does not include any identiﬁed clini- cal sample; instead, Paulhus (1998) relied on general popula- tion, college students, prison entrants, and military recruits. Without formal comparisons to Axis I and Axis II disorders, forensic psychologists have no way of knowing whether scale elevations signify defensiveness or simply reﬂect a nor- mative pattern in patient populations. FEIGNED COGNITIVE IMPAIRMENT Feigned cognitive impairment shares a similar deﬁnition and concomitant goals with other types of malingering. However, it differs fundamentally from the malingering of mental dis- orders in two crucial ways: tasks required of the malingerer and detection strategies. As observed by Rogers and Vitacco (in press), the principal task for feigned cognitive impairment is “effortful failure.” In other words, would-be malingerers must convince the examiner that their efforts to succeed are sincere and that their ostensible impairments are genuine. Effortful failure is strikingly different from fabrication of symptoms and associated features typically required for feigned mental disorders. Because of these differences, forensic psychologists must use detection strategies that focus speciﬁcally on cognitive feigning. As a concrete exam- ple, strategies such as “rare symptoms” make little concep- tual sense for the detection of purported deﬁcits on the WAIS-III. Therefore, detection strategies speciﬁc to feigned cognitive impairment must be considered. 120 Evaluation of Malingering and Deception The deﬁnition of malingering does not change, despite differences in presentation and detection strategies. The ma- lingering of cognitive impairment must involve the gross exaggeration or fabrication of intellectual and neuropsycho- logical deﬁcits for an external goal. This point must be em- phasized. Many studies have attempted to substitute other terms, such as “incomplete effort,” “suboptimal effort,” and “poor motivation.” These terms cannot be equated with either feigning or malingering. Most clients are required to participate in forensic evalua- tions. The level of perceived coercion is likely to vary widely by circumstances of the evaluation and the individual charac- teristics of the clients. The far-reaching implications of these evaluations are not overlooked. For instance, the client’s ﬁnancial well-being is often at stake in civil proceedings. Although generally adequate, forensic evaluations do not represent the optimal conditions for the assessment of cogni- tive functioning. To expect clients to put forth optimal efforts under suboptimal conditions appears naïve. The concept of “poor motivation” is both imprecise and inferential. What are the standards for judging certain motiva- tion as “poor,” “adequate,” or “good”? The simple designa- tion of poor motivation may have devastating consequences for a client. The process of assessing gradations of motivation is poorly understood and highly inferential. Forensic psychol- ogists will want to avoid this level of imprecision and the po- tential ethical concerns of drawing unwarranted conclusions. An important distinction must be drawn. Forensic psychol- ogists certainly encounter clients who put forth an incomplete or suboptimal effort. The reasons for this suboptimal effort are typically unknown but may include (a) decreased interest and effort as a result of genuine cognitive impairment; (b) de- creased interest and effort as a result of a comorbid condition (e.g., depression secondary to head injury); © expectations of failure based on recent performance; (d) stress and preoc- cupation with the potential consequences of the evaluation (e.g., loss of disability income); (e) reaction to inferences from the examiner’s questions that the impairment is trivial; and (f) attempts to feign cognitive impairment. Psychologists must address these six reasons for suboptimal effort. Two types of conclusions are possible: 1. In a minority of cases, forensic psychologists may feel conﬁdent that they are able to address effectively each of these reasons for suboptimal effort. In very rare cases, they may have sufﬁcient data to conclude that the subopti- mal effort was a result of feigning and systematically rule out other explanations. 2. In most cases, forensic psychologists lack the data to address systematically the various reasons for suboptimal effort. How should forensic psychologists describe suboptimal effort in the great majority of cases in which feigning cannot be isolated as the predominant reason? To avoid any serious misunderstandings, we recommend that forensic psycholo- gists employ two safeguards: address the possible reasons for suboptimal effort, and proactively clarify the lack of known relationship between this diminished effort and feigning. An example of this recommendation is provided for a female client evaluated following a motor vehicle ac- cident: “The client did not appear to put forth her best pos- sible effort during several tests of her cognitive ability. Reasons for this could include cognitive and emotional im- pairment as a result of her car accident, her expectations of failure, stresses related to the evaluation, or deliberate at- tempts to appear more impaired. These test ﬁndings cannot be used to establish feigning or any other reason for subop- timal effort.” Evaluations of feigned cognitive deﬁcits pose several im- portant ethical issues for forensic psychologists. Because many cognitive feigning measures are single-purpose scales (i.e., only intended for dissimulation), what type of informed consent is required ethically? Youngjohn, Lees-Haley, and Binder (1999) argue that informing clients about cognitive feigning measures may reduce their effectiveness; instead, they advocate instructing clients to put forth maximum effort. Although maintaining the effectiveness of cognitive feigning measures is a laudable goal, it should not be achieved via the neglect of informed consent. In describing the nature of psy- chological services (American Psychological Association, 1992, Ethical Standard 1.07a), a basic obligation occurs to describe their broad objectives, including response styles. This obligation can be satisﬁed by a general statement at the onset of the evaluation; this statement may also serve a bene- ﬁcial purpose in diminishing the likelihood of malingering (Johnson & Lesniak-Karpiak, 1997). A second ethical issue is posed by deliberate misrepresentations to the evaluatee. For example, the Rey 15-item test (see Lezak, 1995) is sometimes intentionally misdescribed as a “difﬁcult” memory task, when this is known to be inaccurate. Forensic psychologists should categorically avoid any misrepresentations to persons being evaluated. The next section outlines the detection strategies for feigned cognitive impairment. It summarizes the recent literature on the effectiveness of speciﬁc strategies and presents an over- view of speciﬁc measures. Detection Strategies Rogers, Harrell, and Liff (1993) identiﬁed six basic detection strategies for feigned cognitive impairment. These strategies have been augmented by forced-choice testing and reaction Feigned Cognitive Impairment 121 time (Rogers & Vitacco, in press) and pairwise comparisons of comparable items (Frederick, 1997). In general, detection strategies can be grouped into two domains: detection by ex- cessive impairment and detection by unexpected patterns . Examples of excessive impairment are failures on very easy items (i.e., ﬂoor effect) and failures below chance on forced- choice formats (i.e., symptom validity testing or SVT). Exam- ples of unexpected patterns include similar performance on easy and difﬁcult items (i.e., performance curve) and unex- pected answers on forced-choice formats (i.e., magnitude of error). In general, detection strategies using unexpected pat- terns are less transparent than excessive impairment and likely to be robust indicators of feigning. Three common detection strategies are subsumed within the “excessive impairment” domain, with feigned perfor- mance overreaching the level of impairment typically found in brain-injured patients. These strategies include ﬂoor effect, SVT, and forced-choice testing (FCT). Table 7.2 summarizes these detection strategies and provides representative exam- ples of the sample cognitive measures. TABLE 7.2 Detection Strategies on Feigned Cognitive Impairment: Measures and Validation Strategy Scale Clinical Usefulness Floor effect Rey 15-Item Many studies found good speciﬁcity but modest sensitivity; it is limited by varying cut scores and possible false-positives with speciﬁc conditions. Floor effect TOMM Several studies found high classiﬁcation rates; it is not tested with comorbid mental disorders. Floor effect HDMT Guilmette, Hart, & Giuliano (1993) found that lower than 90% correct yielded high classiﬁcations; it needs cross-validation. Floor effect Digit Span Two studies found cut score  7 had good speciﬁcity but modest sensitivity; research has relied on differential prevalence design. Floor effect LMT Inman et al. (1998) reported 3 studies supporting the use of the LMT as a screen. Perfor. Curve TONI-S Frederick & Foster (1991) found very positive results when restricted to higher scores; it is limited by the small number of memory-impaired patients and needs replication. Perfor. Curve Ravens-S Gudjonsson & Shackleton (1986) found moderately high classiﬁcation rates; it was partially replicated by McKinzey, Podd, Krehbiel, Mensch, & Trombka (1997). Perfor. Curve DCT Several studies yield moderately high classiﬁcations, but studies use different cut scores. Perfor. Curve LNNB-S McKinzey et al. (1997) found high rates on cross-validation; it appears clinically useful for LNNB administrations. Mag. of Error WMS-R-S Martin, Franzen, & Orey (1998) found moderately high classiﬁcation; it needs replication. Atypical WAIS-R-S Mittenberg, Theroux-Fichera, Zielinski, & Heilbronner (1995) found moderate classiﬁcation but with a substantial false-positive rate. Atypical CVLT-S Sweet et al. (2000) found moderately high classiﬁcation but did not report sensitivity or speciﬁcity estimates. Atypical WMS-R-S Mittenberg, Azrin, Millsaps, & Heilbronner (1993) found high classiﬁcation; it has been replicated (Iverson, Slick, & Franzen, 2000). Sequelae NSI Ridenour, McCoy, & Dean (1998) provide promising data on the overall level of reported symptoms to identify simulators; it needs replication with a range of neuropsychological conditions. SVT PDRT Several studies found superb speciﬁcity but poor sensitivity. SVT HDMT Several studies found superb speciﬁcity but poor sensitivity. SVT TONI-S Frederick & Foster (1991) found superb speciﬁcity but poor sensitivity. SVT TOMM Several studies found superb speciﬁcity but poor sensitivity. FCT PDRT Moderate classiﬁcation; research is limited by differential prevalence design and lack of studies on comorbidity. FCT HDMT Guilmette et al. (1993) used performance below 75% correct to achieve a high classiﬁcation; it needs replication with large samples and evaluation of comorbidity. FCT 21-Item Memory Highly variable classiﬁcation rates were found across studies. FCT “b” Test Boone et al. (2000) found promising data; it needs replication. FCT WMT Iverson, Green, & Gervais (1999) summarize past research that shows promise as a screen. Consistency TONI-S Frederick & Foster (1991) found this useful in conjunction with other strategies. Time PDRT-C Rose, Hall, & Szalda-Petree (1995) found shorter response times for simulators than brain-injured patients; it needs replication. Time TOMM Rees, Tombaugh, Gansler, & Moczynski (1998) found longer response times for simulators than brain-injured patients; it needs replication. Note : TOMM  Test of Memory Malingering (Tombaugh, 1996); HDMT  Hiscock Digit Memory Test (Hiscock & Hiscock, 1989); LMT  Learning Mem- ory Test (Inman et al., 1998); Digit Span  sum of raw scores for highest number forward plus highest number backward; Perfor. Curve  Performance Curve; TONI-S  specially scored Test of Nonverbal Intelligence (Frederick & Foster, 1991); Ravens-S  specially scored Ravens Standard Matrices (Raven, 1981); DCT  Dot Counting Test (Lezak, 1995); LNNB-S  specially scored Luria-Nebraska Neuropsychological Battery (Golden, Purisch, & Hammeke, 1985); Mag. of Error  Magnitude of Error; WMS-R-S  specially scored Wechsler Memory Scales–Revised subtests (Wechsler, 1987); Atypical  Atypical Presentation; WAIS-R-S  specially scored Wechsler Adult Intelligence Scale-Revised (Wechsler, 1981); CVLT-S  specially scored California Verbal Learning Test (Delis, Kramer, Kaplan, & Ober, 1987); Sequelae  psychological sequelae; NSI  Neuropsychological Symptom Inventory (Rattan, Dean, & Rattan, 1989); SVT  symptom validity testing; PDRT  Portland Digit Recognition Test (Binder, 1992); consistency  consistency across parallel items; WMT  Word Memory Test (Green, Astner, & Allen, 1996); Time  response time; PDRT-C  computerized version of the PDRT. 122 Evaluation of Malingering and Deception Floor effect strategy is based on the notion that malinger- ers have difﬁculty distinguishing which cognitive abilities are unlikely to be compromised in patients with genuine neuropsychological impairment. This strategy was ﬁrst pro- mulgated by Andre Rey in the 1940s (see Lezak, 1995) in devising a cognitive task (Rey 15-item memory test) that ap- pears moderately complex (recall of 15 separate items) but is actually simple (items are organized into easy-to-remember sequences). As operationalized, the ﬂoor effect strategy typi- cally uses a very simple recall and recognition task that can be successfully completed by most (  90%) cognitively im- paired persons. For example, most patients with genuine cog- nitive impairment are able to achieve a 90% accuracy on the second trial of the Test of Memory Malingering (TOMM; Tombaugh, 1996). The majority of simulators do not recog- nize the simplicity of the memory task, especially when given repeated trials. The ﬂoor effect strategy has become a popular detection method for cognitive feigning (see Table 7.2). Despite its intu- itive appeal, forensic psychologists should be cautious in applying the ﬂoor effect strategy for two reasons. First, the range of genuine cognitive impairments militates against the selection of items that work equally well for all cognitive deﬁcits. For example, the second trial of the TOMM appears to be highly effective with brain injury cases (false-positives  2.2%) but not with dementia (false-positives  27.0%). Sec- ond, evaluatees can be easily coached to foil the ﬂoor effect. Symptom validity testing (SVT) examines an improbable failure rate based on statistical probability. First championed by Brady and Lind (1961), most SVT methods have a two- choice format; even persons with total incapacity should not score signiﬁcantly below chance. The SVT strategy has been used by numerous cognitive measures, typically in combina- tion with other strategies. Because the SVT takes into ac- count total incapacity, this strategy tends to be effective only with extreme forms of malingering. Generally successful in less than one-third of simulating cases, the SVT is unique among detection strategies in ruling out other reasons for poor performance. The only logical reason for below-chance performance is the recognition of the correct response and subsequent selection of the incorrect response. Forensic psy- chologists can be very conﬁdent in their conclusions about cognitive feigning when performance on SVT is signiﬁcantly below chance. Memory complaints in forensic cases are sometimes fo- cused on personal recollections (e.g., amnesia for the offense). Frederick, Carter, and Powel (1995) proposed that SVT could be used to address purported amnesia by constructing two- choice alternatives for the events in question. Care must be taken to develop equally plausible alternatives (Denney, 1996; Frederick & Denney, 1998) and to test these alternatives on naïve persons to ensure that they have an equal likelihood of being selected. For example, a question about the victim’s hair color may elicit “brown” more often than “blond” responses based on reasonable inferences about the prevalence of differ- ent hair colors (see Rogers & Shuman, 2000). Forced-choice testing (FCT) is simply lower-than- expected performance based on normative data. Unlike other detection strategies, FCT does not apply a logical principle (e.g., ﬂoor effect) or mathematical probability (e.g., SVT). It simply evaluates group differences and attempts to establish an optimum cut score. FCT appears to have been introduced because SVT yielded only modest sensitivity rates (Binder & Willis, 1991). Without extensive samples of cognitively impaired individuals, including those with comorbid men- tal disorders (e.g., major depression or substance abuse), the false-positive rates of FCT cannot be established. Forensic psychologists must be careful to distinguish between FCT (questionable speciﬁcity) and SVT (very high speciﬁcity) in drawing their conclusions. The second domain for cognitive feigning is “unexpected patterns” that capitalize on unlikely responses to speciﬁc items or sets of items. Detection strategies include magnitude of error, performance curve, and consistency across parallel items. Methods using these strategies are summarized in Table 7.2. Magnitude of error (MOE) evaluates the degree of inac- curacy for incorrect responses. Especially in multiple-choice formats, incorrect responses can be grouped into “expected” and “unexpected” categories by inspecting patients with gen- uine cognitive impairment. A reasonable assumption is that most malingerers focus on what items to answer incorrectly, rather than how to answer items incorrectly. Extrapolating from case reports, Rogers et al. (1993) formally identify this strategy. Martin, Franzen, and Orey (1998) designed a multiple-choice format for Visual Reproduction and Logical Memory subtests of the WMS-R (Wechsler, 1987). They found MOE was highly effective at identifying simulators who endorsed a high proportion of unexpected errors. Bender (2000) found the MOE to be the most effective strategy for identifying simulators, even when simulators were warned about MOE. Performance curve is based on the thesis that malingerers do not take into account item difﬁculty in choosing which items to fail. First identiﬁed by Goldstein (1945), perfor- mance curve compares the proportion of correct items across different gradations of item difﬁculty. When plotted on a graph, genuine patients and controls typically evidence a neg- ative curve with lower performance on more difﬁcult items. In contrast, some malingerers exhibit ﬂat or even positive Feigned Cognitive Impairment 123 curves. This strategy appears to be moderately effective across different measures, including Raven standard progres- sive matrices (Gudjonsson & Shackleton, 1986; McKinzey, Podd, Krehbiel, & Raven, 1999), the Dot Counting Test (DCT; Binks, Gouvier, & Waters, 1997), and the Luria- Nebraska Neuropsychological Battery (LNNB; McKinzey, et al., 1997). In addition, several versions of the performance curve are central to the Validity Indicator Proﬁle (VIP; Frederick, 1997). In summary, performance curve strategy appears to be robust, with consistent, positive ﬁndings across different measures. Atypical presentation was traditionally considered an un- standardized evaluation of symptoms that did not make “neu- ropsychological sense” (Rogers et al., 1993). However, more recent studies have examined disparate ﬁndings that rarely occur in patients with genuine cognitive impairment. For example, bona ﬁde patients generally score higher on the WMS-R Attention/Concentration index than the General Memory index, whereas simulators tend to manifest the oppo- site pattern (Mittenberg, Azrin, Millsaps, & Heilbronner, 1993; Iverson et al., 2000). Atypical presentation has also been applied to the WAIS-R in the Vocabulary and Digit Span difference. Mittenberg, Theroux-Fichera, Zielinski, and Heilbronner (1995) found that a discriminant function accu- rately identiﬁed 70.5% of the participants, although the false- positive rate was unacceptably high (36.8%) for forensic use. Descriptive data from disability evaluations cast further doubt on Vocabulary-Digit Span difference. Contrary to predictions, Williams and Carlin (1999) found that claimants with atypical presentations had signiﬁcantly higher IQ scores than those with expected presentations. Finally, research on the Califor- nia Verbal Learning Test (CVLT; Delis, Kramer, Kaplan, & Ober, 1987) indicated that simulators evidence atypical per- formance on both recognition and recall (Sweet et al., 2000; Trueblood & Schmidt, 1993). Psychological sequelae is a variation of atypical perfor- mance that extends beyond cognitive abilities. Rogers et al. (1993) noted that simulators sometimes report symptoms of a mental disorder (Miller & Cartlidge, 1972) or physical com- plaints that are not typically found with genuine patients. For example, Heaton, Smith, Lehman, and Vogt (1978) found that simulators of head injury commonly reported elevations on six MMPI clinical scales. One limitation to this strategy is that nonprofessionals appear to have an intuitive understanding of concomitant symptoms for common conditions, such as mild brain injury (Lees-Haley & Dunn, 1994) and postconcussion syndrome (Mittenberg, D’Attilio, Gage, & Bass, 1990). How- ever, promising work by Ridenour, McCoy, and Dean (1998) suggests that evaluatees can be presented with a wide array of neuropsychological symptoms, with simulators potentially identiﬁable by the range and severity of reported symptoms. This strategy requires further evaluation before clinical implementation. To evaluate consistency across comparable items, Freder- ick and Foster (1991) proposed a consistency ratio for exam- ining performance across items of equal difﬁculty. Frederick (1997) elaborated on this approach in his development of the VIP. This strategy is difﬁcult to implement because clinicians need items that have been rigorously tested across diverse clinical samples to ensure comparability in item difﬁculty. Especially for crystallized intelligence, cognitive abilities (e.g., vocabulary) may be highly variable in genuine patients. As an important caveat, consistency across parallel items should be not confused with consistency of test results. Many genuine patients produce anomalous results on neuropsycho- logical testing. By themselves, inconsistent test results are not helpful to the classiﬁcation of malingering. Response time measures the average time to complete test items. Research is mixed on whether simulators take more time (Rees et al., 1998) or less time (Rose et al., 1995) than pa- tients with compromised cognitive functioning. For practical purposes, response time is typically limited to computer ad- ministrations. At present, response time is not recommended as a general detection strategy. Guidelines for the Classiﬁcation Forensic psychologists involved in neuropsychological cases are faced with several daunting tasks. The ﬁrst task is a thor- ough understanding of detection strategies for feigned cogni- tive impairment and the available measures employing these strategies. Although not exhaustive, Table 7.2 summarizes most of the cognitive feigning measures reported in the clin- ical literature. In malingering cases, forensic psychologists bear the onerous responsibility of knowing the range of cog- nitive feigning measures, their detection strategies, and their general utility. Table 7.2 provides a useful starting point in developing this expertise. The second task for forensic psychologists is the selection of detection strategies and cognitive feigning measures for suspected malingering cases. Psychologists will likely be inﬂuenced by the clinical presentation in their selection of strategies and methods. Two issues must be considered: • Purported Deﬁcit . Does the measure address the supposed impairment? Reported problems with analytic thinking are unlikely to be addressed by simple tests of memory recognition. • Detection Strategy . Do the selected measures represent different selection strategies? As a general rule, detection 124 Evaluation of Malingering and Deception strategies should represent both the excessive impairment (ﬂoor effect and SVT) and the unexpected pattern (MOE and performance curve) domains. Slick, Sherman, and Iverson (1999) propounded stringent standards for deﬁnite and probable malingering of cognitive impairment. For deﬁnite malingering, they proposed that only below-chance performance on SVT accompanied by ex- ternal incentives would be sufﬁcient for this determination. For probable malingering, they proposed at least two of the following: (a) indicators of feigning on one or more well- validated measures of feigned cognitive impairment; (b) dis- crepancies between test data and known patterns of brain functioning; © discrepancy between test data and observed behavior within a speciﬁc domain on two or more neuropsy- chological tests; (d) discrepancy between test data and reli- able collateral reports; and (e) discrepancy between test data and documented background history. Alternatively, they pro- posed only one of the above plus discrepancies with self- reported symptoms or history. For possible malingering, proposed criteria include any major discrepancy between self-reported symptoms and other data (history, patterns of brain functioning, behavioral observations, or collateral in- formation) or exaggerated/fabricated responses on tests of psychological impairment, such as the MMPI-2. Slick et al. (1999) should be applauded for their efforts to systematize the classiﬁcation of malingered cognitive impair- ment. However, this model has substantial limitations for forensic practice. Three major constraints are outlined: 1. Deﬁnite malingering is too narrowly construed. Exclusive reliance on SVT would exclude the great majority of ma- lingerers that are not feigning extreme impairment. We propose that deﬁnite malingering include either SVT or multiple indicators of feigning (including detection strate- gies from the unexpected patterns domain), plus marked discrepancies between test performance and collateral data. 2. Probable malingering is too broadly construed. Foren- sic psychologists should be aware that distinctions be- tween probable and deﬁnite malingering may not have any differential effect on the legal outcome. Therefore, great care must be exercised in establishing probable malin- gering in forensic cases. A major difﬁculty with the Slick et al. model is that the determination of probable malinger- ing can be rendered without the objective application of systematic decision rules. Discrepancies in self-reporting and collateral sources can be explained without invoking the concept of malingering. We propose that “probable malingering” be invoked only when multiple indices of feigned cognitive impairment are present in addition to marked discrepancies. 3. Possible malingering should not be used in forensic cases. Most complex forensic cases have some discrepancies in test data and subsequent reports. As an analogue, forensic psychologists often reach different conclusions about complex neuropsychological cases based largely on the same data. Test and collateral ﬁndings might be viewed as “discrepant” based on the propensities of a particular neu- ropsychologist rather than the response style of the evalu- atee. Terms such as “inconsistent presentation” can be used without the pejorative effects intrinsic to the term “malingering.” Featured Measures Three measures of feigned cognitive impairment from a broad array of potential measures are summarized: PDRT, VIP, and TOMM. They were selected based on their avail- ability and substantial validation. Portland Digit Recognition Test Binder and Willis (1991) developed the PDRT as a 72-item digit recognition test of motivation and effort. A ﬁve-digit number is presented and followed by a distractor (i.e., count- ing backwards). Increasing intervals are included to increase the apparent difﬁculty of the task. The client is asked to choose the previously presented string of digits from two choices. The two-choice format allows the assessment of below-chance performance (i.e., SVT). Alternatively, the client’s performance is compared to expected accuracy of cognitively impaired patients (i.e., FCT). Binder (1993) investigated the SVT in a differential prevalence design. He found that none of the nonforensic pa- tients with moderate to severe head injuries scored below chance. In contrast, 17% of the forensic sample with only mild head injuries scored below chance. He concluded that the SVT is an effective detection strategy, and ﬁnancial in- centives explained the differences in performance. In the same research, Binder also used an FCT with a cut score of  39 (no more than 54.2% correct) to distinguish patients with “unambiguous brain dysfunctions” from simulators. This research did not appear to take into account either co- morbid conditions (e.g., depression) or the effects of stress and preoccupation with the potential consequences of the evaluation. Variations of the PDRT include abbreviated and computer- ized versions. Discontinuation rules can be employed when an individual performs well on the ﬁrst 18 or 36 items, thereby Feigned Cognitive Impairment 125 shortening the administration time. A modiﬁcation of the computer administration allows for an examination of un- usual response times. Rose et al. (1995) found higher reaction times in patients than in simulators on the PDRT. They con- cluded that the patients required more time to process the ma- terial due to cognitive slowing associated with head injury. Alternatively, the simulators may have underestimated the impact brain injury has on processing speed and failed to slow their responses accordingly. The PDRT is appropriate for use in forensic contexts when employed to evaluate SVT via below-chance performances. When used appropriately, SVT virtually eliminates false- positives, making below-chance performances highly indica- tive of feigning. However, this strategy has only modest sensitivity, meaning that most feigners are not identiﬁed by SVT on the PDRT. Forensic psychologists are likely to be divided on the usefulness of the FCT with the PDRT, even as a screen for feigning. Without ruling out other explanations (e.g., comorbidity), the relationship between unexpectedly poor performance and potential feigning has not been fully evaluated. Finally, the RT strategy has not been sufﬁ- ciently validated as to warrant its forensic application. Validity Indicator Proﬁle The VIP (Frederick, 1997) employs a two-choice format for the assessment of suboptimal effort on two subtests addressing verbal and nonverbal abilities. The VIP is distinguished from other cognitive measures by its use of multiple strategies fo- cused predominantly on unexpected patterns. The strategies include three estimates of response consistency and ﬁve esti- mates of performance curve. Because of the high inter- correlations for response consistency ( Mr  .81), forensic psychologists may be concerned whether they are discrete or largely redundant scales. Estimates of SVT are also possible, although not employed as a primary strategy. The VIP classiﬁes proﬁles as either “valid” or “invalid” rather than feigning per se. Invalid proﬁles are sorted into three categories (Frederick, 1997, p. 2): (a) “careless” (poor effort but motivated to do well); (b) “irrelevant” (intention to per- form poorly but not a sustained effort); and © “malingered” (intention to perform poorly with a sustained effort). Using the broad categories of valid and invalid, the classiﬁcation rates are moderately high. The VIP nonverbal subtest has a sensitiv- ity rate of 73.5% and a speciﬁcity rate of 85.7%. The VIP verbal test has a sensitivity rate of 67.3% and a speciﬁcity rate of 83.1%. The VIP is best conceptualized as a measure of suboptimal effort rather than feigning. Very few simulators and suspected malingerers are correctly classiﬁed in the “malingering” category. Speciﬁc estimates of malingering classiﬁcations are provided: • Nonverbal subtest. 3 of 52 (5.8%) simulators and 1 of 49 (2.0%) suspected malingerers were classiﬁed correctly in the malingering category (Frederick, 1997, p. 28, Table 8). The combined accuracy is 4/101 or 4.0%. • Verbal subtest. 4 of 52 (7.7%) simulators and 1 of 49 (2.0%) suspected malingerers were classiﬁed correctly in the malingering category (Frederick, 1997, p. 29, Table 9). The combined accuracy is 5/101 or 5.0%. • Combined subtests. The classiﬁcation integrating both tests for malingering is not reported but should not exceed 9.0%. An extrapolation from Table 12 (p. 29), summariz- ing the concordance for invalid subtests, yields 6.2% as an approximate estimate. These estimates derived from the VIP test manual do not support its use for the classiﬁcation of malingering or feign- ing. As a measure of suboptimal effort, should forensic psy- chologists conclude that “invalid” proﬁles are likely the result of feigning? Substantial percentages of brain-injured patients have “invalid” results on the nonverbal (26.2%) and verbal (36.1%) subtests. Depending on the prevalence rate for feigned cognitive impairment, invalid proﬁles may be found at comparable rates between brain-injured patients with no apparent motivation to feign, and simulators and suspected malingerers. The VIP should not be used clinically with two groups manifesting cognitive impairment, namely, those with mental retardation or learning disabilities. As noted by Frederick (1997), the VIP should not be used to evaluate patients with mental retardation (i.e., operationalized as Shipley IQs 75). Almost all (95.0%) of these participants produced invalid proﬁles. Psychologists are cautioned not to use educational attainment as an indirect measure of mental retardation; approximately two-thirds (67.5%) had at least a high school education. In addition, persons with learning disabilities were systematically excluded from the cross- validation phase and are not included in the classiﬁcation tables. In summary, the VIP is an ambitious effort to evaluate response styles through the use of multiple detection strate- gies and the evaluation of both nonverbal and verbal abilities. The most judicious use of the VIP is the assessment of sub- optimal effort. Forensic psychologists should be careful not to equate suboptimal effort with deception or fraud. Depend- ing on base rates, invalid VIPs may be just as likely to represent genuine impairment as any form of dissimula- tion. In rare cases where the VIP designates a protocol as 126 Evaluation of Malingering and Deception “malingering,” it is likely to be the result of feigning (found in seven cases) or possibly random responding (found in two cases). Test of Memory Malingering The TOMM (Tombaugh, 1996, 1997) is a two-alternative memory recognition task composed of 50 line drawings. Presented in two trials, the optimum cut score (45 or 90% correct) occurs in Trial 2. Scores at or above the cut score cor- rectly classiﬁed 95% of nondemented patients; scores below the cut score identiﬁed 100% of the simulators. A small num- ber of patients in a differential prevalence design had average scores substantially below the cut score ( M  32.8). In addition to the ﬂoor effect, the TOMM also uses SVT, which apparently has a low detection rate for feigners (Rees et al., 1998). Several cautions apply to the use of the TOMM in foren- sic practice. First, the TOMM appears to produce much lower results when applied to patients in litigation or those seeking disability (Tombaugh, 1996). Although some litigating pa- tients are malingering, the differential prevalence design leaves questions unanswered about its applicability to foren- sic cases. As noted by Smith (1998), the directions for the cognitively impaired group differed substantially from the standard TOMM instructions. In its validation, cognitively impaired participants were (a) verbally redirected to the task, (b) focused on both alternatives with expanded instructions, and © selectively re-instructed about the task (patients with dementia). This focusing and prompting may have artiﬁcially inﬂated TOMM scores for those with genuine cognitive im- pairment. The real danger is that the standard instructions may substantially increase false-positives, wrongly classify- ing genuine patients as feigners. SUMMARY Forensic psychologists are faced with formidable challenges in the assessment of malingering and defensiveness. As noted in this chapter, many clinicians and attorneys have misunder- standings and misassumptions about response styles. Foren- sic psychologists must be able to address these inaccuracies, including the potential misuse of DSM-IV indices. Clinically, they develop expertise through the knowledge of detec- tion strategies and their application to psychological mea- sures. Although faced with a daunting number of response style measures, they select empirically validated scales that are domain-speciﬁc (e.g., feigned psychopathology versus feigned cognitive impairment) and relevant to the immediate case. Forensic psychologists carefully integrate multiple sources of data, consistent with established detection strate- gies, in rendering their opinions on response styles to the courts. Enduring challenges remain for forensic research on re- sponse styles. The next century should bring additional de- tection strategies that are rigorously tested by both simulation designs and known-groups comparisons. For cognitive as- sessment in particular, detection strategies need to be both expanded to cover diverse neurocognitive abilities and re- ﬁned to improve clinical classiﬁcation. From a forensic- psychological perspective, the standardized assessment of feigned medical conditions remains a vast, uncharted terri- tory that requires both sophisticated conceptualization and sound empiricism. REFERENCES American Psychiatric Association. (2000). Diagnostic and statisti- cal manual of mental disorders (4th ed., text rev.). Washington, DC: Author. American Psychological Association. (1992). Ethical principles of psychologists and code of conduct. American Psychologist, 47, 1597–1611. Baer, R. A., Wetter, M. W., & Berry, D. T. R. (1992). Detection of underreporting of psychopathology on the MMPI: A meta- analysis. Clinical Psychology Review, 12, 509–525. Baer, R. A., Wetter, M. W., & Berry, D. T. R. (1995). Effects of in- formation about validity scales on underreporting of symptoms on the MMPI-2: An analogue investigation. Assessment, 2, 189–200. Baer, R. A., Wetter, M. W., Nichols, D. S., Greene, R., & Berry, D. T. R. (1995). Sensitivity of MMPI-2 validity scales to under- reporting of symptoms. Psychological Assessment, 7, 419–423. Bagby, R. M., Buis, T. E., & Nicholson, R. A. (1995). Relative ef- fectiveness of the standard validity scales in detecting fake-good and fake-bad responding: Replication and extension. Psycholog- ical Assessment, 7, 84–92. Bagby, R. M., Nicholson, R. A., Buis, T., Radovanovic, H., & Fidler, B. J. (1999). Defensive responding on the MMPI-2 in family custody and access evaluations. Psychological Assessment, 11, 24–28. Bagby, R. M., Rogers, R., Nicholson, R. A., Buis, T., Seeman, M. V., & Rector, N. A. (1997). Effectiveness of the MMPI-2 validity indicators in the detection of defensive responding in clinical and nonclinical samples. Psychological Assessment, 9, 406–413. Bender, S. D. (2000). Strategic detection of cognitive feigning: Pre- liminary support for the TOCA. Unpublished doctoral disserta- tion, University of North Texas, Denton. References 127 Binder, L. M. (1992). Portland Digit Recognition Test. Unpub- lished test, Department of Veteran Affairs Medical Center, Port- land, OR. 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. Binder, L. M., & Willis, S. C. (1991). Assessment of motivation after ﬁnancially compensable minor head trauma. Psychological Assessment: A Journal of Consulting and Clinical Psychology, 3, 175–181. Binks, P. G., Gouvier, W. D., & Waters, W. F. (1997). Malingering detection with the Dot Counting Test. Archives of Clinical Neuropsychology, 12, 41–46. Boone, K. B., Lu, P., Sherman, D., Palmer, B., Back, C., Shamieh, E., et al. (2000). Validation of a new technique to detect malin- gering of cognitive symptoms: The b Test. Archives of Clinical Neuropsychology, 15, 227–241. Brady, J., & Lind, D. (1961). Experimental analysis of hysterical blindness. Archives of General Psychiatry, 4, 331–339. Butcher, J. N., Williams, C. L., Graham, J. R., Tellegen, A., & Kaemmer, B. (1989). MMPI-2: Manual for administration and scoring. Minneapolis: University of Minnesota Press. Daubert, S. D., & Metzler, A. E. (2000). The detection of fake-bad and fake-good responding on the Millon Clinical Multiaxial Inventory III. Psychological Assessment, 12, 418–424. Delis, D. C., Kramer, J. H., Kaplan, E., & Ober, B. (1987). Califor- nia Verbal Learning Test–Adult version. San Antonio, TX: Psy- chological Corporation. Denney, R. L. (1996). Symptom validity testing of remote memory in a criminal forensic setting. Archives of Clinical Neuropsy- chology, 11, 589–603. Frederick, R. I. (1997). The Validity Indicator Proﬁle. Minneapolis, MN: National Computer Sy

bundet

stems. Frederick, R. I., Carter, M., & Powel, J. (1995). Adapting symptom validity testing to evaluate suspicious complaints of amnesia in medicolegal evaluations. Bulletin of the American Academy of Psychiatry and the Law, 23, 231–237. Frederick, F. I., & Denney, R. L. (1998). Minding your p s and q s when using forced-choice recognition tests. Clinical Neuropsy- chologist, 12, 193–205. Frederick, R. I., & Foster, H. G. (1991). Multiple measures of malingering on a forced-choice test of cognitive ability. Psycho- logical Assessment: A Journal of Clinical and Consulting Psychology, 3, 596–602. Golden, C. J., Purisch, A. D., & Hammeke, T. A. (1985). The Luria- Nebraska battery manual. Palo Alto, CA: Western Psychologi- cal Services. Goldstein, H. (1945). A malingering key for mental tests. Psycho- logical Bulletin, 42, 104–118. Green, P., Astner, K., & Allen, L. M., III. (1996). The Word Memory Test: A manual for oral and computer-administered forms. Durham, NC: CogniSyst Inc. Greene, R. L. (1997). Assessment of malingering and defensiveness on multiscale inventories. In R. Rogers (Ed.), Clinical assessment of malingering and deception (2nd ed., pp. 169–207). New York: Guilford Press. Greene, R. L. (2000). The MMPI-2/MMPI: An interpretive manual. Boston: Allyn & Bacon. Gudjonsson, G. H., & Shackleton, H. (1986). The pattern of scores on Raven’s Matrices during “faking bad” and “non faking” performance. British Journal of Clinical Psychology, 25, 35–41. Guilmette, T. J., Hart, K. J., & Giuliano, A. J. (1993). Malingering detection: The use of a forced-choice method in identifying or- ganic versus simulated memory impairment. Clinical Neuropsy- chologist, 7, 59–69. 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, 892–900. Hiscock, C. K., & Hiscock, M. (1989). Reﬁning the forced-choice method for the detection of malingering. Journal of Clinical and Experimental Neuropsychology, 11, 967–974. Inman, T. H., Vickery, C. D., Berry, D. T. R., Lamb, D. G., Edwards, C. L., & Smith, G. T. (1998). Development and initial validation of a new procedure for evaluating adequacy of effort given dur- ing a neuropsychological testing: The Letter Memory Test. Psy- chological Assessment, 10, 128–139. Iverson, G. L., Green, P., & Gervais, R. (1999). Using the Word Memory Test to detect biased responding in head injury litiga- tion. Journal of Cognitive Rehabilitation, 17, 4–8. Iverson, G. L., Slick, D. J., & Franzen, M. D. (2000). Evaluation of a WMS-R malingering index in a non-litigating clinical sample. Journal of Clinical and Experimental Neuropsychology, 22, 191–197. Johnson, J. L., & Lesniak-Karpiak, K. (1997). The effect of warning on malingering on memory and motor tasks in college samples. Archives of Clinical Neuropsychology, 12, 231–238. Kurtz, R., & Meyer, R. G. (1994, March). Vulnerability of the MMPI-2, M Test, and SIRS to different strategies of malingering psychosis. Paper presented at the American Psychology-Law Society, Santa Fe, NM. Lees-Haley, P. R., & Dunn, J. T. (1994). The ability of naïve subjects to report symptoms of mild brain injury, posttraumatic stress disorder, major depression, and generalized anxiety disorder. Journal of Clinical Psychology, 50, 252–256. Lezak, M. D. (1995). Neuropsychological assessment (3rd ed.). New York: Oxford. LoPiccolo, C. J., Goodkin, K., & Baldewicz, T. T. (1999). Current issues in the diagnosis and management of malingering. Annals of Medicine, 31, 166–174. 128 Evaluation of Malingering and Deception Martin, R. C., Franzen, M. D., & Orey, S. (1998). Magnitude of error as a strategy to detect feigned memory impairment. Clini- cal Neuropsychologist, 12, 84–91. McKinzey, R. M., Podd, M. H., Krehbiel, M., Mensch, A., & Trombka, C. (1997). Detection of malingering on the Luria- Nebraska Neuropsychological Battery: An initial and cross- validation. Archives of Clinical Neuropsychology, 12, 505–512. McKinzey, R. M., Podd, M. H., Krehbiel, M., & Raven, J. (1999). Detection of malingering on Raven’s standard progressive matri- ces: A cross-validation. British Journal of Clinical Psychology, 38, 435–439. Miller, H., & Cartlidge, N. (1972). Simulation and malingering after injuries to the brain and spinal cord. Lancet, 1, 580–586. 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. Mittenberg, W., Azrin, R., Millsaps, C., & Heilbronner, R. (1993). Identiﬁcation of malingered head injury on the Wechsler Mem- ory Scale–Revised. Psychological Assessment, 5, 34–40. Mittenberg, W., D’Attilio, J., Gage, R., & Bass, A. (1990, February). Malingered symptoms following mild head trauma: The post- concussion syndrome. Paper presented at the 18th meeting of the International Neuropsychological Society, Orlando, FL. Mittenberg, W., Theroux-Fichera, S., Zielinski, R., & Heilbronner, R. (1995). Identiﬁcation of malingered head injury on the Wechsler Adult Intelligence Scale–Revised. Professional Psy- chology: Research and Practice, 26 (5), 491–498. Morel, K. R. (1998). Development and preliminary validation of a forced-choice test of response bias for posttraumatic stress disorder. Journal of Personality Assessment, 70, 299–314. Morey, L. C. (1996). An interpretive guide to the Personality Assessment Inventory (PAI). Tampa, FL: Psychological Assess- ment Resources. Paulhus, D. L. (1998). Paulhus Deception Scales (PDS): The Bal- anced Inventory of Desirable Responding–7. North Tonawanda, NY: Multi-Health Systems. Pebbles, J., & Moore, R. J. (1998). Detecting socially desirable responding with the Personality Assessment Inventory: The pos- itive impression management and the defensiveness index. Journal of Clinical Psychology, 54, 621–628. Poythress, N. G., Nicholson, R., Otto, R. K., Edens, J. F., Bonnie, R. J., Monahan, J., et al. (1999). Professional manual for the MacArthur Competence Assessment Tool-Criminal Adjudica- tion. Odessa, FL: Psychological Assessment Resources. Rattan, G., Dean, R. S., & Rattan, A. I. (1989). Neuropsychological Symptom Inventory. Muncie, IN: Author. Raven, J. C. (1981). Manual for Raven’s Progressive Matrices and Mill Hill Vocabulary Scales. London: H. K. Lewis. Rees, L. M., Tombaugh, T. N., Gansler, D. A., & Moczynski, N. P. (1998). Five validation experiments with the Test of Memory Malingering. Psychological Assessment, 10, 10–20. Ridenour, T. A., McCoy, K. D., & Dean, R. S. (1998). Discriminant function analysis of malingerers’ and neurological headache pa- tients’ self reports of neuropsychological symptoms. Archives of General Neuropsychology, 13, 561–567. Rogers, R. (1990a). Development of a new classiﬁcatory model of malingering. Bulletin of the American Academy of Psychiatry and Law, 18, 323–333. Rogers, R. (1990b). Models of feigned mental illness. Professional Psychology: Research and Practice, 21, 182–188. Rogers, R. (Ed.). (1997). Clinical assessment of malingering and deception (2nd ed.). New York: Guilford Press. Rogers, R. (2001). Handbook of diagnostic and structured inter- viewing. New York: Guilford Press. Rogers, R., Bagby, R. M., & Dickens, S. E. (1992). Structured In- terview of Reported Symptoms (SIRS) and professional manual. Odessa, FL: Psychological Assessment Resources. 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). New York: LEA. Rogers, R., & Dickey, R. (1991). Denial and minimization among sex offenders: A review of competing models of deception. An- nals of Sex Research, 4, 49–63. Rogers, R., Harrell, E. H., & Liff, C. D. (1993). Feigning neuropsy- chological impairment: A critical review of methodological and clinical considerations. Clinical Psychology Review, 13, 255–274. Rogers, R., & Reinhardt, V. (1998). Conceptualization and assess- ment of secondary gain. In G. P. Koocher, J. C. Norcross, & S. S. Hill, III (Eds.), Psychologist’s desk reference (pp. 57–62). New York: Oxford University Press. Rogers, R., & Salekin, R. T. (1998). Beguiled by Bayes:Are-analysis of Mossman and Hart’s estimates of malingering. Behavioral Sciences and the Law, 16, 147–153. Rogers, R., Salekin, R. T., Sewell, K. W., Goldstein, A. M., & 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., Cruise, K. R., Wang, E. W., & Ustad, K. L. (1998). The PAI and feigning: A cautionary note on its use in forensic-correctional settings. Assessment, 5, 399–405. Rogers, R., Sewell, K. W., & Goldstein, A. M. (1994). Explanatory models of malingering: A prototypical analysis. Law and Human Behavior, 18, 543–552. Rogers, R., Sewell, K. W., Morey, L. C., & Ustad, K. L. (1996). Detection of feigned mental disorders on the Personality Assess- ment Inventory: A discriminant analysis. Journal of Personality Assessment, 67, 629–640. References 129 Rogers, R., Sewell, K. W., & Salekin, R. (1994). A meta-analysis of malingering on the MMPI-2. Assessment, 1, 227–237. Rogers, R., & Shuman, D. W. (2000). Conducting insanity evalua- tions. New York: Guilford Press. Rogers, R., & Vitacco, M. J. (in press). Predictors of adolescent psy- chopathy: The role of impulsivity, hyperactivity, and sensation seeking. In B. Van Dorsten (Ed.), Forensic psychology: From classroom to courtroom. Boston: Kluwer Academic/Plenum. Rose, F. E., Hall, S., & Szalda-Petree, A. D. (1995). Portland Digit Recognition Test–computerized: Measuring response latency improves the detection of malingering. Clinical Neuropsycholo- gist, 9, 124–134. Salekin, R. T. (2000). Test review: The Paulhus Deception Scales. American Psychology-Law Society News, 20, 8–11. Sechrest, L. (1963). Incremental validity: A recommendation. Edu- cational and Psychological Measurement, 23, 153–158. Sewell, K. W., & Salekin, R. T. (1997). Understanding and detect- ing dissimulation in sex offenders. In R. Rogers (Ed.), Clinical assessment of malingering and deception (2nd ed., pp. 328–350). New York: Guilford Press. 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 Neuropsy- chologist, 13, 545–561. Smith, G. P. (1998). Test review: The Test of Memory Malingering. American Psychology-Law Society News, 18, 16–18, 29–30. Sweet, J. J., Wolfe, P., Sattlberger, E., Numan, B., Rosenfeld, J. P., Clingerman, S., et al. (2000). Further investigation of traumatic brain injury versus insufﬁcient effort with the California Verbal Learning Test. Archives of Clinical Neuropsychology, 15, 105–113. Thomas-Peter, B. A., Jones, J., Campbell, S., & Oliver, C. (2000). Debasement and faking bad on the Millon Clinical Multiaxial Inventory-III: An examination of characteristics, circumstances and motives of forensic patients. Legal and Criminological Psychology, 5, 71–81. Timbrook, R. E., Graham, J. R., Keiller, S. W., & Watts, D. (1993). Comparison of the Wiener-Harmon subtle-obvious scales and the standard validity scales in detecting valid and invalid MMPI-2 proﬁles. Psychological Assessment, 5, 53–61. Tombaugh, T. N. (1996). TOMM: The Test of Memory Malingering. North Tonawanda, NY: Multi-Health Systems. Tombaugh, T. N. (1997). The Test of Memory Malingering (TOMM): Normative data from cognitively intact and cogni- tively impaired individuals. Psychological Assessment, 9, 260–268. Trueblood, W., & Schmidt, M. (1993). Malingering and other valid- ity considerations in the neuropsychological evaluation of mild head injury. Journal of Clinical and Experimental Neuropsy- chology, 15, 578–590. Vanhouche, W., & Vertommen, H. (1999). Assessing cognitive dis- tortions in sex offenders: A review of commonly used versus recently developed instruments. Psychologica Belgica, 39, 163–187. Wechsler, D. (1981). Weschler Adult Intelligence Scale–Revised . San Antonio: Psychological Corporation. Wechsler, D. (1987). Wechsler Memory Scale–Revised manual. San Antonio, TX: Psychological Corporation. Wechsler, D. (1997). Weschler Adult Intelligence Scale–Third edition. San Antonio, TX: Psychological Corporation. Wildman, R. W., & Wildman, R. W., Jr. (1999). The detection of malingering. Psychological Reports, 84, 386–388. Williams, R. W., & Carlin, M. (1999). Malingering on the WAIS-R among disability claimants and applicants for vocational assis- tance. American Journal of Forensic Psychology, 17, 35–45. Youngjohn, J. R., Lees-Haley, P. R., & Binder, L. M. (1999). Com- ment: Warning malingerers produces more sophisticated malin- gering. Archives of Clinical Neuropsychology, 14, 511–515.