Psychometric Techniques

Instruments that directly index posttraumatic symptoms are successfully faked by naive and coached simulators, partly because they lack validity scales to detect malingering (Fruehetal., 2000; Liljequistetal., 1998; Morel, 1998). Accordingly, a number of psychological tests have been proposed to assist in detecting the feigning of psychological symptoms, including the Morel Emotional Numbing Test (MENT; Morel, 1998), Millon Clinical Multiaxial Inventory (MCMI; Grillo et al., 1994; Lees-Haley, 1992; Zisken, 1995), Minnesota Multiphasic Personality Inventory (MMPI/MMPI-2; Fairbank, McCaffrey and Keane, 1985; Lees-Haley, 1992,1997; Lyons and Wheeler-Cox, 1999; McCaffrey and Bellamy-Campbell, 1989; Perconte and Goreczeny, 1990; Rogers et al., 1993a; Zisken, 1995), and the Personality Assessment Inventory (PAI; Liljequist et al., 1998; Rogers, Ornduff and Sewell, 1993b; Rogers et al., 1996,1998; Rogers, Ustad and Salekin, 1998a; Wang et al., 1997).

The most studied measure to index genuineness of clinical presentation is the MMPI/MMPI-2 (Hathoway and McKinley, 1991). The MMPI-2 has an array of validity scales designed to index motivation underlying responses to items about psychopathology, including the F, Fb, L, K, Gough Dissimulation Index, Fp, S and Mp (for reviews, see Butcher and Miller, 1999; Greene, 1997; Pope, Butcher and Seelen, 2000). The ability of the MMPI-2 to discriminate between genuine and malingered presentations has been studied in a range of populations, including chronic pain, brain injury, and PTSD (Butcher and Miller, 1999). A number of studies have indicated the utility of the MMPI-2 to distinguish genuine from malingered PTSD (Fairbank et al., 1985). Even when malingerers are given information about PTSD, the MMPI-2's F, Fb, Dissimulation Index, and Ds validity scales can distinguish malingerers from genuine presentations (Wetter et al., 1993). In addition, malingerers have also been distinguished by their scores on O-S, OT and FBS (Elhai et al., 2000, 2001).

The PAI (Morey, 1991) is another self-report personality inventory designed to assess response styles, clinical disorders, treatment planning and screen for psychopathology. The PAI was developed to overcome a number of psychometric limitations associated with the MMPI-2 (see Boyle andLennon, 1994; Helmes andReddon, 1993; Liljequist et al., 1998). The PAI also has the advantage of item responses that reflect four gradations of endorsement (i.e. 'notatalltrue', 'slightly true', 'mainlytrue', and'very true'), and using non-overlapping scales to maximize discriminant validity (Rogers et al., 1998a). The PAI contains a number of Validity scales, including the Negative Impression scale (NIM), Positive Impression scale (PIM), Malingering Index, and Critical Items scale. Morey (1991) reported very high scores on the NIM for college students who were instructed to feign mental disorders and also found empirical support for the use of the NIM in classifying simulators and genuine patients. In terms of PTSD, Liljequist et al. (1998) administered the PAI to students instructed to feign PTSD, and substance abuse veterans with or without PTSD. Malingerers produced higher scores on the NIM and Malingering Index. In a study of civilians, Bowen and Bryant (2001) compared treatment-seeking patients with posttraumatic stress, naive simulators who were provided with no information about posttraumatic stress, and sophisticated simulators who were provided with information about posttraumatic stress symptoms. Both naive and sophisticated malingerers produced PAI profiles that over-endorsed the majority of clinical scales relative to genuine respondents, and also endorsed more items on the NIM validity scale, Malingering Index and Critical items list. The initial evidence points to the utility of the PAI as a measure of feigning PTSD.

Although the MMPI-2 and the PAI have significant potential to index attitudinal and motivational factors in the presentation of an individual, the evidence supporting the use of any particular scales with particular claimants can be effectively challenged. Although the MMPI-2 is widely used as a means to detect malingering, the recommended cut-off scores for malingering varies markedly across studies (e.g. Lyons and Wheeler-Cox, 1999; Perconte and Goreczeny, 1990). There is also evidence that using the PAI's NIM scale as an index of malingering can result in a proportion of genuine PTSD cases being misclassified as malingerers (Calhoun et al., 2000). There has also been doubts raised about the cross-validation of MMPI-2 profiles in litigation settings with specific populations, such as adolescents (Archer, 1989). Questions about the accuracy of profiles generated by multiscale inventories are reflected in legal challenges to the admissibility of the MMPI-2. For example, Byrd v. State, 593 NE 2d (Ind. 1992) noted the limitations of the MMPI, stating that its utility is 'not as a primary source of information, but instead as a means of confirming or challenging clinical impressions previously gained through direct contact with the patient' (460). Courts have also challenged the admissibility of computer-scored MMPI-2 profiles on the grounds that there is uncertainty about the expertise of the programmer, recording procedures, and accuracy of computer-scored profiles (Sullivan v. Fairmont Homes, Inc., 543 NE 2d 1130 (Ind. App. 1 Dist. 1989)). Despite these limitations, it has been argued that the best defence of the MMPI-2 is the extensive data attesting to its reliability, which is the cornerstone of many jurisdictions' decision to accept expert testimony (see Pope et al., 2000).

0 0

Post a comment