Document not found! Please try again

Feigning Schizophrenic Disorders on the MMPI--2

0 downloads 0 Views 725KB Size Report
Jun 10, 2010 - 25-min film on schizophrenia successfully faked that disorder in comparison to. 46% who were naive. In contrast, supplying simulators with ...
This article was downloaded by: [University of Toronto Libraries] On: 26 August 2014, At: 10:46 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Personality Assessment Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/hjpa20

Feigning Schizophrenic Disorders on the MMPI--2: Detection of Coached Simulators Richard Rogers , R. Michael Bagby & Debi Chakraborty Published online: 10 Jun 2010.

To cite this article: Richard Rogers , R. Michael Bagby & Debi Chakraborty (1993) Feigning Schizophrenic Disorders on the MMPI--2: Detection of Coached Simulators, Journal of Personality Assessment, 60:2, 215-226, DOI: 10.1207/s15327752jpa6002_1 To link to this article: http://dx.doi.org/10.1207/s15327752jpa6002_1

PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

JOURNAL OF PERSONALITY ASSESSMENT, 1993, 60(2), 215-226 Copyright o 1993, Lawrence Erlbaum Associates, Inc.

Feigning Schizophrenic Disorders on the MMPI-2: Detection of Coached Simulators Richard Rogers Downloaded by [University of Toronto Libraries] at 10:46 26 August 2014

University of North Texas

R. Michael Bagby Clarke Institute of Psychiatry University of Toronto

Debi Chakraborty Clarke Institute of Psychiatry

The Minnesota Multiphasic Personality Inventory (MMPI) and, more recently, its revised version (the MMPI-2) have represented the "gold standard" in the psychometric assessment of malingering and other response styles. In this study, we provide a stringent test of the MMPI-2 validity indices and their ability to detect feigned schizophrenia in four groups of simulators (n = 72). Simulators were randomly assigned to one of four conditions: (a) coached on symptoms of schizophrenia, (b) coached on strategies for the detection of fakers, (c) coached on both symptoms and strategies, or (d) uncoached. Simulators were compared to subjects responding under an honest condition (n = 13) and a comparison group of schizophrenic inpatients (n = 37). We found knowledge of strategies alone allowed many simulators (i.e., one third or more, depending on the validity indices) to elude detection. In contrast, knowledge of the disorder appeared less useful to simulators in avoiding detection. Coaching on both strategies and symptoms was not as effective as strategies alone. Consistent with previous studies, uncoached simulators were detected with moderately high levels of accuracy.

Psychometric studies of malingering make the crucial yet untested assumption that individuals feigning disorders are naive t o both the purposes of psychological testing a n d t h e nature of diagnosis. Almost without exception, simulation studies of feigning (for reviews, see Greene, 1988; Pankratz, 1988; Schretlen, 1988; Stermac, 1988) have not provided subjects with either sufficient prepara-

Downloaded by [University of Toronto Libraries] at 10:46 26 August 2014

216

ROGERS, BAGBY, CHAKRABORTY

tion time or background materials. Given the importance of malingering in terms of financial incentives and quality of life (Rogers, 1990), the most prudent course would be to assume that at least some malingerers do prepare (e.g., the Bianchi case evidenced elaborate preparation; see Orne, Dinges, & Orne, 1984). However, the range of preparation and sophistication among malingerers remains empirically untested. The few available studies of informed simulators underscore the need to test the fakability and detectability of each psychometric method. For example, in their classic study of feigned schizophrenic disorders on the Rorschach, Albert, Fox, and Kahn (1980) found that 72% of university subjects who watched a 25-min film on schizophrenia successfully faked that disorder in comparison to 46% who were naive. In contrast, supplying simulators with information on schizophrenia from the Diagnostic and Statistical Manual of Mental Disorders (3rd ed. [DSM-III]; American Psychiatric Association, 1980) did not appear to assist them in escaping detection on the M test, a 33 item true-false screening measure of feigned schizophrenia; similar percentages of simulators were correctly identified when given the information (i.e., 78.2%; Beaber, Marston, Michelli, & Mills, 1985) as when they were not (i.e., 79.8%; Gillis, Rogers, & Bagby, 1991). Petersen and Viglione (1991) tested the effects of psychological training (graduate students in clinical psychology vs. professionals outside mental health) on ability to fake specific disorders on the MMPI. They found that simulators feigning nonpsychotic disorders (depression and acute reactions) frequently eluded detection. Available studies of the MMPZ-2 (Bagby, Rogers, Dickens, Nussbaum, & Nohara, 1991; Graham, Watts, & Timbrook, 1991) have not investigated coached simulators. Rogers and his colleagues conducted two related studies of feigning on the Structured Interview of Reported Symptoms (SIRS; Rogers, Gillis, Dickens, & Bagby, 1991)and the effects of coaching on detection. In the first study, Rogers, Gillis, Bagby, and Monteiro (1991) provided university subjects with information about strategies employed by the SIRS to identify simulators and ideas about how to avoid detection. Coached simulators had significantly lower scores than uncoached simulators on all 13 of the SIRS scales that assess malingering. However, coached simulators were still significantly elevated on 7 of the 13 SIRS scales compared to university controls and psychiatric inpatients. In a second study, Rogers, Kropp, and Bagby (1992) examined the ability of correctional residents with prior mental health contact and information about specific disorders (schizophrenia, mood disorders and posttraumatic stress disorders) taken from the Diagnostic and Statistical Manual of Mental Disorders (3rd ed., rev. [DSM-111-R]; American Psychiatric Association, 1987) to fake a believable mental disorder, Despite this detailed information, simulators were easily detectable with highly significant differences from bona fide inpatients on 9 of the 13 SIRS scales ( a h see Rogers, Bagby, & Dickens, 1992). The pivotal question is, "How do malingerers prepare to fake?" Do they

Downloaded by [University of Toronto Libraries] at 10:46 26 August 2014

FEIGNING ON THE MMPI-2

217

educate themselves about specific disorders? Do they study how psychometric measures assess feigning? Without answers to these cluestions, we need to investigate how both knowledge of diagnosis and detection strategies may affixt subjects' ability to fake successfully. In this study, we examined how well subjects could fake a schizophrenic disorder under one of four conditions: no coaching, coaching about schizophrenia, coaching about MMPI-2 validity indices, and combined coaching about schizophrenia and validity indices. We chose to study the MMPI-2 because this measure is likely to become, like the original MMPI, the primary psychometric measure of response styles that include malingering. A second issue to examine is the ability of simulators to achieve clinical elevations on relevant scales. Rogers (1984) suggested that nearly all psychometric methods are fakable. We are particularly interested in whether simulators under specific experimental conditions are able to achieve clinically interpretable elevations o n Scales 6, 7, 8, and 9.

METHOD Subjects A total of 72 community subjects (35 male and 37 female) were recruited through written notices; these were posted community bulletin boards and placed in local mailboxes. Prospective subjects were asked to call one of the authors and were explained the general purpose of the study. Subjects were screened to ensure they met minimal educational requirements ( 2Grade 8), and they gave informed consent. The sample was re dominantly White (61.1%) and Asian (27.7%), with a mean age of 38.12 years (SD = 12.47). Th~esample consisted of professionals (27.8%), students (23.6%), people in business or sales (12.5%), and secretaries (5.5%). Approximately 7% of the sample was unemployed, and 23.6% did not provide information about their occupational status. A clinical sample of 37 schizophrenic inpatients from the Continuing Care and Forensic Divisions of the Clarke Institute of Psychiatry were selected as a comparison group. All inpatients were diagnosed with schizophrenic disorders by a psychiatrist experienced with the DSM-Ill-R who also had diagnostic input from the patients' clinical teams (i.e., psychologists, social workers, and nursing staff).Given the increasing demand for inpatient beds, schizophrenics admitted to these units were actively psychotic. In addition, patients were screened t~ ensure they had minimal educational requirements. This sample differed from the nonclinical sample with respect to gender (31 males and 6 females), X2 (1, N = 109) = 12.66, p < .01; because males and females from adult psychiatric samples manifest nearly identical scale elevations on validity indices (set. Greene, 1988, pp. 138-149), we do not believe that this gender disparity affect5

218

ROGERS, BAGBY, CHAKRABORTY

the interpretation of the study's results. The clinical sample had a mean age of 32.83 years (SD = 11.36) and was composed of Whites (81.1%),Blacks (10.8%), and Native Canadians (8.1%).

Downloaded by [University of Toronto Libraries] at 10:46 26 August 2014

Procedure We used a between-groups design, and four groups were asked to feign schizophrenia on the MMM-2 under one of the following experimental conditions: (a) no coaching, (b) coaching on schizophrenic symptoms, (c) coaching on MMPI validity indices (i.e., strategies), and (d) coaching on both symptoms and strategies. In addition, a nonpatient control group was asked to complete the MMPI-2 carefully and honestly. Subjects were randomly assigned to either the control group or one of the four experimental conditions. Finally, a comparison group of schizophrenic inpatients was administered the MMPI-2 under standard instructions. The MMPI-2 validity indices (F, FB,F - K, DS-r,' Obvious minus Subtle items, and Lachar and Wrobel [LW] critical item list; see Greene, 1988, 1991) served as the dependent variables. Subjects in each of four experimental conditions were asked to feign a schizophrenic disorder of sufficient severity as to warrant hospitalization. As an incentive for a convincing presentation, subjects were told that "successful" simulators in each of the four conditions were eligible for a $50.00 lottery. Because cutting scores for the MMPI-2 validity indices are not well established, all simulators were placed in their respective lottery. Subjects in the control (honest) condition were also given an opportunity for their own $50.00 lottery. All nonpsychiatric subjects were given the MMPI-2 to take home and were asked to complete it during the next week.2 Subjects were provided with background material of symptoms and strategies in accordance with the experimental conditions. Symptoms of schizophrenia consisted of the DSM-111-R (American Psychiatric Association, 1987, pp. 194-195) criteria with explanatory notes on the meaning of technical terms such as delusions, hallucinations, loosening of associations, catatonic behavior, inappropriate affect, and thought broadcasting. Strategies for the detection of faking (Greene, 1988) included those found in the F scale, the Ds-r scale, Obvious minus Subtle items, and LW critical items; descriptions of these strategies are listed in the Appendix. Those 'MMPI-2 retained 34 of the 40 items on the Ds-r; we have designated this revised scale as Ds-rZ 'Although the take-home MMPI represents a departure from standard practice, we believe that it is justified in this study on two grounds. First, the MMPI is frequently administered with minimal supervision by nonprofessional staff, and no allowances are made in the scoring or interpretation despite highly diverse settings (Dahlstrom, Welsh, & Dahlstrom, 1972). Second, a primary purpose of on-site administration with nonpsychotic and well-educated individuals is to minimize any form of cheating. Because our study focuses on the ability of subjects to simulate, we wanted to encourage rather than discourage feigning. Naturally, the comparison sample of schizophrenic inpatients were administered the MMPI under standard conditions.

FEIGNING ON THE W I - 2

219

subjects receiving background information were instructed to discard it prior to beginning the MMPI-2. Upon completion of the MMPI-2, subjects were debriefed, and the five lottery prizes were subsequently awarded.

Downloaded by [University of Toronto Libraries] at 10:46 26 August 2014

RESULTS We examined differences among simulators, nonpatient controls, and the schizophrenic comparison group on MMPI-2 validity indices through a series of analyses of variance (ANOVAs) with Duncan's multiple range tests (alpha = .05). As summarized in Table 1,significant differences were found across groups. As expected, controls and schizophrenic inpatients generally had lower scores than simulators. However, simulators coached on strategies alone had significantly lower scores on F, F - B, and fewer endorsements on LW critical item list than uncoached simulators and simulators coached on symptoms alone. These lower scores on the validity indices resulted in simulators in the strategies-only condition to be relatively indistinguishable from schizophrenic inpatients. The only exception was 0 - S for which mean scores for all simulation conditions were substantially higher (>90 points) than both inpatients and controls. Our primary concern was the effectiveness of MMPI-2 validity indices to correctly identify simulators and bona fide patients. Toward this end, we sought TABLE 1 Differences Among Coached (Symptoms, Strategies, and Both) Simulators, Uncoached Simulators, Controls, and Schizophrenic Patients on MMPI-2 Validity Indices Vdzdity Indices

F FB F-K Ds-r,

0-S

L-W

Couched Simulators Symptomsa

30.47, 20.13, 20.80, 20.07, 180.00, 65.33,

Strategiesb

16.71,, 10.29, 5.43, 15.50, 118.21, 45.00,

Botha

Uncoucheda

Controls'

Schixd

24.60, 15.80ab 15.93, 18.33, 191.53, 62.00,

31.00, 20.80, 19.93, 19.47, 154.93, 63.55,

5.38, 3.54, -7.69, 9.00, 27.77, 39.31,

15.51, 10.73, 1.51, 11.38, 8.4% 37.00,

Note. All validity indices are reported as raw scores with the exception of 0 - S. Mean scores with common subscripts are not significantly different (alpha = .05) on Duncan's multiple-range tests. For example, only Symptoms and Uncoached simulators use subscript a on the F scale, which means that they do not differ from each other but do differ from all other conditions. Symptoms =: simulators coached with DSM-111-R symptoms of schizophrenia; Strategies = simulators coached on the underlying smategiesof MMPI-2 validlty indices; Both = simulators coached with symptoms and strategies; Schiz = schizophrenic mpatients; 0s-r2 = 34 items on the Ds-r chat were retained on the MMPI-2; 0 - S = the overall T-score difference between Obvious and Subtle items on the Weiner (1948) subscales; L-W = the Lachar and Wrobel (1979) critical item lisr. "n = 15, ,n = 14. 'n = 13. dn = 37. *p < .001.

Downloaded by [University of Toronto Libraries] at 10:46 26 August 2014

220

ROGERS, BAGBY, CHAKRABORTY

to establish the most efficient cutting scores in differentiating honest responders (controls and schizophrenics) from different categories of simulators (see Table 2). The commonly used validity indices of F and F - K (see Franzen, Iverson, & McCracken, 1990; Schretlen, 1988) did not fare well. These indices proved inefficient with subjects coached on strategies, either alone or in combination with symptoms. In addition, they misclassified approximately one third of other simulators (uncoached or symptoms alone). In contrast, two validity indices appeared superior to others: Ds-rz and 0 - S. These indices were able to correctly identify nearly two thirds of those coached on strategies alone and approximately 80% of all other simulation conditions. We were interested in what differences simulation and honest groups would manifest on the MMM-2 Scales 6, 7,8, and 9. Simulators who were coached on symptoms (alone or in combination with strategies) evidenced in high scale elevations (see Table 3) that were comparable to those who were uncoached. Substantial numbers of simulators (33.3% of symptoms only, 7.1% on strategies only, 20.0% of strategies and symptoms, and 60% of uncoached) had extreme elevations on Scale 8 (T-scores > 100) and may have endorsed too many symptoms to be considered schizophrenic (Graham, 1987), although they may

TABLE 2 Effectiveness of MMPI-2 Fake Bad Decision Rules for Classifying Coached (Symptoms, Strategies, and Both) Simulators, Uncoached Simulators, Controls, and Schizophrenic Patients on the MMPI-2 ClassificationAccuraq (%) Decision Rules F > 28 FB > 17 F-K>14 Ds-r, > 15 0 - S > 106 L-W > 61

Coached Simulators Symptomsa

Snaregiesb

Both"

Uncoached"

Controlsc

Schird

73.3 73.3 66.7 80.0 87.7 66.7

21.4 28.6 28.6 64.3 64.3 21.4

46.7 40.0 47.7 73.3 93.3 66.7

60.0 73.3 66.7 73.3 80.0 60.0

100.0 100.0 100.0 100.0 92.3 92.3

97.3 81.1 81.1 83.8 81.1 94.1

Note. Profiles that meet decision rules are classified as fake bad. With respect to classification accuracy, the proportion of simulators under each condition that is identified as fake bad is an estimate of positive predictive power; the proportion of honest responders (controls and schizophrenic inpatients) that do not meet the decision rule is an estimate of negative predictive power (for a discussion of classification rules, see Gottesman & Prescott, 1989). Symptoms = simulators coached with DSM-Ill-R symptoms of schizophrenia; Strategies = simulators coached on the underlying strategies of MMPI-2 validity indices; Both = simulators coached with symptoms and strategies; Schiz = schizophrenic inpatients; Ds-r, = 34 items on the Ds-r that were retained on the MMPI-2; 0 - S = the overall T-score difference between Obvious and Subtle items on the Weiner (1948) subscales; L-W = the Lachar and Wrobel (1979) critical irem list. "n = 15. bn = 14. ' n = 13. dn = 37.

FEIGNING O N THE MMPI-2

;!21

TABLE 3 Differences Among Coached (Symptoms, Strategies, and Both) Simulators, Uncoached Simulators, Controls, and Schizophrenic Patients on MMPI-2 Scales 6, 7, 8, and 9

-

MMPI-2 Scales 6 7 8

Downloaded by [University of Toronto Libraries] at 10:46 26 August 2014

9

-

Coached Simulators Symptomsa

Strategtesb 72.57, 68.93,, 78.71, 53.86,

92.87, 80.20, 93.40, 66.67ah

Both" 89.53, 76.40, 92.07, 64.07,,,

Uncoached" 90.53, 79.27, 98.20, 74.13,

Controlsc

Schizd

F*

52.46, 48.62, 52.15, 58.08,,

72.35, 62.49, 70.51, 56.08,

9.24 10.61 16.26 5.63

-

Note. MMPI-2 scales are reported as K-corrected T.scores; those with common subscripts are not significant (alpha = .05) on Duncan's multiple-range tests. Symptoms = simulators coached with DSM-111-R symptoms of schizophrenia; Strategies = simulators coached on the underlying strategies of MMPI-2 validity indices; Both = simulators coached wlth symptoms and strategies; Schiz = schizophrenic inpatients. "n = 15. ,n = 14. 'n = 13. dn = 37. * p I ,001"

TABLE 4 Percentages of Simulators, Controls, and Schizophrenic Inpatients Achieving Clinically Elevated Elevations on MWI-2 Scales 6, 7, 8, and 9 MMPI-2 Scales

Coached Simulators Symptomsa

Strategiesb

Both"

Uncoached"

Controls'

schizd

-

Note. Interpretable elevations refer to K-corrected T-scores r 65. Symptoms = simulators coached with DSM-Ill-R symptoms of schizophrenia; Strategies = simulators coached on the underlying strategies of MMPI-2 validity indices; Both = simulators coached with symptoms and strategies; Schiz = schizophrenic inpatients. "n = 15. n, = 14. 'n = 13. dn = 37.

still have been considered psychotic (Greene, 1991).~With respect t o clinical practice, we were interested in how successful simulators would be in achieving clinically elevated elevations (K-corrected T-scores 2 65) o n Scales 6, 7, 8, artd 9. We predicted that a greater proportion of those subjects with specific information o n schizophrenia would achieve elevations o n Scales 6 and 8. As observed in Table 4, this prediction was inaccurate because most uncoached simulators also achieved clinically significant elevations. Note that a substantial percentage of schizophrenics did not have clinically 31t is interesting to note that this criterion (T-scores < 100 for schizophrenics) worked well, because 94.6% of our sample met this criterion.

222

ROGERS, BAGBY, CHAKRABORTY

significant elevations on Scales 6 (i.e., 40.5%) and 8 (i.e., 37.8%). Moreover, a surprising proportion of community controls had such elevations (Scale 6, 23.1%; Scale 8, 15.4%).

Downloaded by [University of Toronto Libraries] at 10:46 26 August 2014

DISCUSSION Previous research (Bagby et al., 1991; Graham et al., 1991) established cutting scores for F and other validity indices based on uncoached simulators and patient groups. Our results are consistent with these studies in providing moderately high levels of detection for uncoached simulators. However, an enduring problem with the original MMPI was the wide variability in cutting scores (see reviews by Berry, Baer, & Harris, 1991; Franzen et al., 1990; Schretlen, 1988); this problem appears to persist with the MMPI-2. For example, Butcher, Dahlstrom, Graham, Tellegen, and Kaemmer (1989) recommended F greater than 11; Graham et al. (1991) suggested F greater than 23; and we found F greater than 28 yielded the optimum classification. Given our limited knowledge of MMPI-2 validity indices, clinicians must decide for themselves what types of classification errors (i.e., malingerers categorized as patients or patients categorized as malingerers) are acceptable. In many clinical settings, stringent cutting scores (e.g., F > 28) that minimize the misclassification of bona fide patients as malingerers may be more acceptable, given the very negative consequences frequently associated with determinations of malingering. A n important caveat in interpreting our results are the relatively small sample sizes in the specific simulation conditions; replications with larger samples are of our findings. Other caveats apply important in evaluating the generali~abilit~ to our limited knowledge of actual malingerers on the MMPI (Roman, Tuley, Villanueva, & Mitchell, 1990) and application of validity indices to adolescent populations (Herkov, Archer, & Gordon, 1991). Bearing these caveats in mind, we found two validity indices (Ds-r2and 0 S) that appeared to be relatively effective, even with coached simulators. We recommend further investigations be conducted of the Ds-r2 before any decisions are made to eliminate this scale. In addition, clinicians may wish to examine 0 - S in any cases of suspected feigning. We found that 0 - S greater than 106 correctly identified 49 of 59 simulators (83.1%) and 30 of 37 schizophrenic inpatients (81.1%). Our findings are in stark contrast with Greene's (1991) cutting score for overreporting symptoms of 0 - S greater than 230. As he acknowledged, his cutting score assumes that scores in the top 5% of psychiatric populations, recomputed from MMPI normative data of Hedlund and Won Cho (1979), are "presumptive" evidence of o ~ e r r e ~ o r t i n ~ . ~ 4Use of Greene's cutting score for overreporting (0- S > 230) correctly identified 10 of 59 simulators (16.9%); use of his cutting score for acceptable protocols (0- S ranging from 4 to 130) misidentified 15 of 59 simulators (25.4%).

Downloaded by [University of Toronto Libraries] at 10:46 26 August 2014

FEIGNING ON THE MMPI-2

i!23

We found that the majority of coached and uncoached simulators achieved clinically significant elevations on Scales 6 and 8. Some uncoached simulators yielded extreme elevations that were atypical of schizophrenics; however, this finding offers little comfort because these individuals are still likely to be seen as mentally disordered. As noted in the introduction, the implicit assumption that malingerers do not prepare is counterintuitive and devoid of empirical support. We hypothesize that at least some malingerers expend considerable energy in preparation for faking. We assumed that this effort would be divided between knowledge of the purported disorder and efforts to foil the assessment: methods. In this study, we investigated the effects of coaching by symptoms, strategies implicit in MMPI-2 validity indices, and a combination of both symptoms and strategies. We fouind that knowledge of strategies, as embodied in a two-page summary, allowed at least one third of simulators to avoid detection, A n important consideration is whether potential malingerers could access scholarly material on malingering and distill information on MMPI-2 detection strategies. We surmise that most would have difficulty both in access and comprehension. If this supposition is true, then this study is a stringent test of the MMPI-2's ability to detect sophisticated malingerers. Interestingly, knowledge of schizophrenic symptoms was generally not helpful. Indeed, simulators coached on strategies alone consistently outperformed those given both symptoms and strategies in eluding detection by MMM validity indices. We hypothesize that subjects given both are distracted by the symptoms and made less effective use of the strategies. Reviewing the available studies of the MMM-2, we are convinced that clinicians should not rely solely o n the MMPI-2 for making the determination of fake-bad response styles. We recommend that cliniciains review for possible adoption more specialized measures of feigning, such as the SIRS (Rogers, Bagby, & Dickens, 1992; Rogers, Gillis, Dickens, & Bagby, 1991) or the M Test (Beaber et al., 1985; Rogers, Bagby, & Gillis, 1992).

ACKNOWLEDGMENT We thank John Graham for his cogent comments o n an earlier draft of this article.

REFERENCES Albert, S., Fox, H. M., h Kahn, M. W. (1980).Faking psychosis on a Rorschach: Can expert judges detect malingering?Journal of Personality Assessment, 44, 115-1 19. American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disorders (3rd. ed.). Washington, DC: Author.

Downloaded by [University of Toronto Libraries] at 10:46 26 August 2014

224

ROGERS, BAGBY, CHAKRABORTY

Amerlcan Ps~chiatr~c Assoclanon. (1987). Dzagnostzc and statutzcal manual ofmental dzsorders (3rd ed., rev.). Washington, DC: Author. Bagby, R. M., Rogers, R., D~ckens,S. E., Nussbaum, D., & Nohara, M (1991). Dlsnzmznattng mconsutent, malzngenng, and dejenszve response styles on the Mrnnesota Multzphaszc Personalzty Invento~y-2 Manuscr~ptsubm~ttedfor pubhation. J., & M~lls,M. J. (1985). A brlef test for measuring malinger~n~ Beaber, J. R., Marston, A., M~chell~, in schizophrenic tndlv~duals.American Journal of Psychlatry, 142, 1478-1481. Berry, D. T. R., Baer, R. A., & Harris, M. J. (1991). Detection of mal~ngerlngon the MMPI: A metanalyt~crevlew. Clznzcal Psychology Reuzew, 11, 585-598. Butcher, J. N., Dahlstrom, W G., Graham, J. R., Tellegen, A., & Kaemmer, B. (1989). Manual for admznzstratzon and sconng MMPI-2 Mmneapolis: Universtty of Mmnesota Press. Dahlstrom, W. G., Welsh, G. S., & Dahlstrom, L. E. (1972). An MMPI handbook Clznccal interpretation (Vol. l), Mmneapolls: Unrversity of Minnesota Press. Franzen, M. D., Iverson, G. L., & McCracken, L. M. (1990). The detection of malmgering m neuropsychological assessment. Neuropsychology Revzew, 1, 247-279. Gdlis, 1. R., Rogers, R., & Bagby, R. M. (1991).Validity of theM test: Simulation design and natural group approaches. Journal of Personaltty Assessment, 57, 130-140. Gottesman, I. I., & Prescott, C. A. (1989). Abuses of the MacAndrew MMPI alcoholism scale: A critical revlew. Clznzcal Psychology Revzew, 9, 223-242. Graham, J. R. (1987). The MMPP A practzcal guzde (2nd ed.) New York: Oxford Un~vers~ty Press Graham, J. R., Watts, D., & Timbrook, R (1991). Detecnng fake-good and fake-bad MMPI-2 profiles. Journal of Personaltty Assessment, 57, 264-277. Greene, R. L. (1988). Assessment of malinger~ng and defensweness by objectwe personal~ty measures. In R. Rogers (Ed.), Clznzcal assessment of malingerrng and deceptzon (pp. 123-158). New York: Gullford. Greene, R. L. (1991). The MMPI-2/MMPI An mterpretzve manual Boston: Allyn & Bacon. Hedlund, J. L., & Won Cho, D. (1979). [MMPI data research tape for M~ssouriDepartment of Mental Health pat~ents].Unpublished raw data. Herkov, M. J., Archer, R. P., &Gordon, R A. (1991). MMPI response sets among adolescents: An evaluat~onof the limitat~onsof the subtle-obvious subscales. Psychologrcal Assessment A Journal of Clznzcal and Consultzng Psychology, 3,424-426. Lachar, D., & Wrobel, T. A. (1979). Validating chnrcians' hunches: Construct~onof a new MMPI critrcal item set. Journal of Cowultzng and Clznzcal Psychology, 47, 277-284 Orne, M. T., D~nges,D. F., & Orne, E. C. (1984). On the differentlal dlagnos~sof multlple personality d~sorderm the forensic context. Internahonal Journal of Clznlcal and Expenmental Hypnosu, 32, 118-169. Pankratz, L (1988). Mal~ngeringon mtellectual and neuropsychological measures. In R. Rogers (Ed.), Clrntcal assessment of malzngerzng and deceptzon (pp. 169-192). New York: Gullford. Rogers, R. (1984). Towards an empmcal model of malmgering. Behauloral Sczences and the Law, 2, 93-112. Rogers, R. (1990). Models of feigned mental dlness. Professzonal Psychology Research and Practrce, 21, 182-188 Rogers, R., Bagby, R. M , & D~ckens,S. E. (1992). Structured Intervzew of Reported Symptoms (SIRS) and test manual Odessa, FL: Psychological Assessment Resources Rogers, R., Bagby, R. M., & Gill~s,J. R (1992). Improvements tn the M test in assessment of malingering Bulletzn of the Amerzcan Academy of Psychzatry and Law, 20, 101-104. Rogers, R., G~llrs,J. R., Bagby, R. M., & Monteiro, E. (1991) Detection of malmgerlng on the SIRS. A study of coached and uncoached simulators Psychologzal Assessment A Journal of Consultzng and Clzn~calPsychology, 3, 673-677. Rogers, R., Glll~s,J R., D~ckens,S E., & Bagby, R. M. (1991). Standard~zedassessment of

FEIGNING ON THE MMPI-2

225

Downloaded by [University of Toronto Libraries] at 10:46 26 August 2014

malingering: Validation of the SIRS. Psychological Assessment: A Journal of Clinical and Consulting Psychology, 3, 89-96. Rogers, R., Kropp, P. R., & Bagby, R. M. (1992). Faking specific disorders: A study of the Structured Interview of Reported Symptoms (SIRS). Journal of Clinical Psychology, 48, 643-647. Roman, D. D., Tuley, M. R., Villanueva, M. R., & Mitchell, W. E. (1990). Evaluating WdPI validity in a forensic psychiatric population: Disringuishing between malingering and genuine psychopathology. Criminal Justice and Behior, 17, 186-198. Schretlen, D. J. (1988). The use of psychological tests to identify malingered symptoms of a mental disorder. Clinical Psychological Review, 8, 451-476. Stermac, L. (1988). Projective testing and dissimulation. In R. Rogers (Ed.), Clinical assessment of malingering and deception (pp. 159-168). New York: Guilford. Weiner, D. W. (1948). Subtle and obvious keys for the MMPI. Joumal of Consulting Psychology, 12, 164-170.

Richard Rogers Department of Psychology University of North Texas P.O. Box 13587 Denton, TX 76203-3587 Received May 21, 1992

APPENDIX The Minnesota Multiphasic Personality Inventory-2 (MMPI-2) is a 566 true-false item questionnaire used by psychologists to help diagnose cases of mental illness. There are several built-in strategies that let the psychologist know if the person completing the questionnaire is faking the mental illness. When a person fakes a mental illness for some outside goal it is known as malingering. The strategies that are incorporated into the MMPI-2 are based o n the results of research that looks at the scores received by both patients and normal people. By studying both people with and without mental illness, psychologists can begin to see what patterns will give the malingerer away. Please read the following carefully. It is a brief description of the strategies that can be used and may be helpful to you in the second part of this study.

Strategy 1: Obvious Minus Subtle Some of the items on the MMPI-2 reflect symptoms that you would most likely thinkof when you consider a "crazy" person. These are obvious items. Similarly, there are items that ask about common, underlying symptoms of a mental illness. These are the subtle symptoms. In the MMPI-2, there are 110 subtle and 146 obvious symptoms.

226

ROGERS, BAGBY, CHAKRABORTY

Malingerers are often identified because they endorse many more obvious than subtle symptoms. As a result, when the score is obtained, it reflects the fact that the possibility exists that the person is faking their symptoms. T o avoid detection, the malingerer should not endorse too many obvious symptoms.

Downloaded by [University of Toronto Libraries] at 10:46 26 August 2014

Strategy 2: Ds-r This scale was developed by comparing groups and identifying which answers were different when normal people malingered and when they answered honestly. Basically, malingerers rely heavily on their stereotypes of how mentally ill individuals would respond. T o successfully "deceive" the MMPI-2, the person must keep in mind all of the symptoms of a mental illness, not just the stereotypical ones.

Strategy 3: Critical Items In this strategy, a list of critical items has been identified. These include the problems that finally cause people to seek professional help. They also include the specific symptoms used by the psychologist in diagnosing a mental illness. A person who is trying to malinger often responds positively to a large number of these items but ignores other items that, although not essential, are nevertheless present. T o avoid detection, the malingerer should limit the number of severe problems he or she endorses positively.

Strategy 4: F Scale This is the traditional scale used to identify malingering. If a patient responds positively to a large number of these items, they would be showing a pattern that is atypical of normal people. However, the content of the items is such that people with mental illness are not likely to respond positively to them. The items may have a bizarre content so that malingerers would think that a positive response would indicate mental illness. The opposite is true: Mentally ill people do not frequently endorse them. The malingerer in this case should not respond positively to items that are extremely bizarre but should keep in mind the broad range of symptoms that may be seen in mental illness.