جستجو برای هسته پان فرهنگی اختلال شخصیت جامعه ستیز

Abstract Previous research has reported that psychopathy ratings of offenders in Canada and the United States made using the Psychopathy Checklist-Revised were metrically equivalent with each other, but not with ratings of Scottish offenders or with ratings of offenders from the rest of the UK. In this study, we further evaluated the cross-cultural validity of psychopathic personality disorder by comparing ratings from a number of different continental European countries to those from North America. Overall, the findings indicated the presence of a significant culture bias in PCL-R ratings. Cross-cultural stability was highest for symptoms related to deficient affective experience, suggesting that they may be the pan-cultural core of the disorder. The findings are consistent with cultural facilitation models of psychopathology. They have important implications for cross-national comparisons of research based on the PCL-R and implications for the clinical and forensic application of the PCL-R in Europe.

. Introduction Personality disorders are mental abnormalities characterized by chronic disturbances in relating to self, others, and the environment. Because personality is inherently relational in nature, manifested largely in the interpersonal sphere, culture may have a greater impact on personality disorders than on most other forms of mental disorder (Cross & Markus, 1999). Because complex and manifold social processes encourage interpersonal behaviors consistent with important norms and values (Weisz, Suwanlert, Chaiyasit, & Walter, 1987), personality disorders may tend to be an exaggeration of prevalent patterns of adaptation within a society (Alarcon, Foulks, & Vakkur, 1998). There has been considerable debate concerning the impact of culture on the expression of psychopathic personality disorder. Historical and anthropological evidence suggests that psychopathic personality disorder is found across time and across cultures (Cooke, 1996), however, it has been speculated that the disorder may be more prevalent in highly individualistic cultures (Lykken, 1995; Paris, 1998). Individualistic cultures promote the development of self-identity that is independent of relationships with others; factors that at their extreme characterize psychopathy are highly valued in individualistic cultures (Cross & Markus, 1999). To date, there has been little quantitative research on the cross-cultural validity of psychopathic personality disorder (Cooke, 1996). Some studies have examined the distribution of test scores or prevalence of diagnoses across different cultural groups (e.g., Compton et al., 1991; Loranger, 1999; Robins, Tipp, & Przybeck, 1991). Research of this sort is sometimes referred to as the “transport-and-test” approach (Cross & Markus, 1999): researchers take an existing measure, translate it into different languages, and then compare the distribution or reliability of scores across cultures. There are several problems with this strategy (Cooke & Michie, 2002; Cross & Markus, 1999; Lopez & Gaurnaccia, 2000; Rogler, 1999). First, although it is relatively straight-forward to translate a test into another language, it is not at all easy to ensure the cultural equivalence of literal translations. Second, tests may lack cultural relevance, even when they have been translated adequately. It is always possible to administer an old test in a new culture and obtain scores, but it is not at all certain that the validity of test scores is the same across cultures. Third, even assuming cultural equivalence, and cultural relevance of a test, most commonly-used statistics do not provide an adequate means of evaluating potential cultural bias. Methods and indexes commonly used to describe the distribution and structure of scores—such as the mean and standard deviation of total scores and the internal consistency—do not provide a strong test of whether individual items and total scores operate equivalently across cultures. Specialized techniques including Confirmatory Factor Analysis (CFA) and Item Response Theory (IRT) methods are preferable ( Cooke & Michie, 2002; Embretson & Reise, 2000; Lopez & Gaurnaccia, 2000). The PCL-R (Hare, 2003) is a 20-item symptom rating scale of psychopathic personality disorder intended for use in forensic settings. The test manual provides a definition of each item, and evaluators rate the lifetime presence of symptoms on a three-point scale (0–2) on the basis of an interview with the subject and a review of case history information. The PCL-R is North American by birth: it was developed originally in Canada, written in English, and validated primarily in Canada and the United States. Psychometric evaluations conducted within the framework of Classical Test Theory (CTT) indicate that the test has high interrater, internal consistency, and test–retest reliabilities in North America (Hare, 2003); similar evaluations within the framework of IRT indicate that many test items and the total score provide useful information across the full range of the latent trait (Cooke & Michie, 1997). There is also a large body of research supporting the test’s validity in North America (Cooke, Forth, & Hare, 1998). Although born and raised in North America, the PCL-R has since traveled overseas—mostly to Europe (e.g., Af Klinteberg, Humble, & Schalling, 1992; Pham, 1998; Tengstrom, Grann, Langstrom, & Kullgren, 2000; but also to Asia, Australasia, and South America (e.g., Darke, Kaye, Finlay-Jones, & Hall, 1998; Howard, Payamal, & Neo, 1997). In cross-cultural research, interest must be focused not on manifest variables such as test scores, but rather on the latent variables that underlie them (Meredith & Millsap, 1992; Waller, Thompson, & Wenk, 2000). Cross-cultural equivalence requires that the structure of the latent trait and the metric used to measure the latent trait are both invariant across cultures. Thus, two broad questions have to be addressed. First, is the factor structure invariant across cultures? This can be addressed through the application of CFA (Tanzer, 1995). Second, is the association between PCL-R ratings and the latent trait of psychopathy the same across cultures? This question can be addressed through the application of IRT methods (Embretson & Reise, 2000; Lopez & Gaurnaccia, 2000). These techniques are complementary and provide different information about the structure of data. We have conducted a small number of studies using CFA and IRT methods to examine cross-cultural differences (Cooke and Michie, 1997, Cooke and Michie, 1999 and Cooke and Michie, 2001; Cooke, Kosson, & Michie, 2001; Cooke, Michie, Hart, & Clark, 2004; Cooke, Michie, Hart, & Clark, submitted). Three broad conclusions can be reached. First, the disorder has a clear factor structure. Cooke & Michie (2001) used CFA methods to analyze 13 of the 20 PCL-R items that were conceptually distinct and, according to previous IRT findings, psychometrically non-redundant. A hierarchical structure in which the superordinate trait, psychopathy, overarched three highly correlated symptom facets: Arrogant and Deceitful Interpersonal Style, Deficient Affective Experience, and Impulsive and Irresponsible Behavioral Style fitted the data best. This factor structure replicated well in data from Scotland, as well as in ratings based on the Screening Version of the PCL-R (Hart, Cox, & Hare, 1995) and the various criterion sets used in the DSM-IV Antisocial Personality Disorder Field Trial (Widiger et al., 1996). This structure has been replicated by independent investigators (e.g., Skeem, Mulvey, & Grisso, 2003). These finding suggest that the syndromal structure of psychopathic symptoms may be invariant across cultures and measures. Second, within North America there is little evidence of cross-cultural variation across ethno-cultural groups (Cooke et al., 2001; Cooke & Michie, 1997). Cooke et al. (2001) examined the differences between samples of African-American and White offenders in the United States. The three factor structure replicated and was completely invariant across groups. Differential item functioning across the groups on 5 of 20 PCL-R items was observed; however, the differences were small in magnitude and inconsistent in direction, so that they canceled out each other when items were summed to form total scores. Third, there appears to be structural invariance between Europe and North America but not metric invariance (Cooke & Michie, 1999; Cooke et al., submitted). In both these studies there was evidence of metric inequivalence with participants in the UK, on average, obtaining lower total scores for the same level of the underlying trait compared with North American participants: the interpersonal features of the disorder showed greatest differences and the affective features the least. In this paper we examine the generalizability of the PCL-R from North America to continental Europe. We examine four issues. First, is the syndrome of psychopathy specified by the same symptoms in both settings? Second, do PCL-R scores have the same meaning in both settings? Third, where are any differences in the disorder located? And fourth, which symptoms are most diagnostic of psychopathy?

3. Results First, we evaluated the extent to which the three-factor hierarchical model fits ratings from Europe. Following convention, adequate fit was determined by values of the NNFI and CFI greater than 0.90 and RMSEA less than 0.08: models that do not achieve these values can be regarded as mis-specified (Byrne, 1994). Using a combination of indices provides a more conservative and reliable evaluation of fit. The model was parameterized as outlined by Cooke & Michie (2001), using 13 of the 20 PCL-R items. The fit for this model for the European sample was good, χ2 (56, N = 729) = 201.3, p < 0.001, NNFI = 0.94, CFI = 0.96, RMSEA = 0.06. 1 Also, neither the traditional two factor solution for the PCL-R nor the four factor model recently proposed in the PCL-R manual (see Hare, 2003, Fig. 7.3) fitted χ2 (117, N = 586) = 874.8, p < 0.001, NNFI = 0.78, CFI = 0.81, RMSEA = 0.10 and χ2 (131, N = 585) = 864.3, p < 0.001, NNFI = 0.80, CFI = 0.83, RMSEA = 0.10. Second, as a more rigorous test of cross-sample structural invariance, we fitted the three-factor hierarchical model simultaneously to data from Europe versus Canada and the United States. The fit of the baseline (i.e., unconstrained) model was good, χ2 (112, N = 2723) = 558.6, p < 0.001, NNFI = 0.95, CFI = 0.96, RMSEA = 0.04. The fit obtained when the loadings were constrained to be equal across cultures was good, χ2 (125, N = 2723) = 662.4, p < 0.001, NNFI = 0.94, CFI = 0.96, RMSEA = 0.04, although significantly worse than the fit of the unconstrained model, Δχ2 (13, N = 2723) = 103.8, p < 0.001. We conducted IRT analyses of the 13 PCL-R items incorporated in the three-factor hierarchical model. Initially, an unconstrained baseline model was generated in which the mean level of the latent trait and all item parameters were allowed to vary across the two groups (see Table 1). Constraining the a parameters (slopes) to be equal resulted in a significant increase in χ2, Δχ2(13) = 167.9, p < 0.001, indicating that the discriminating power of items was significantly different across cultures. Table 1. IRT parameters for Europe versus Canada and the United States, 13 and 20 item models after anchoring Item 13 items 20 items Europe Canada and the US Europe Canada and the US a b1 b2 a b1 b2 a b1 b2 a b1 b2 1 Glibness/Superficial charma 1.1 −0.3 1.5 1.5 −0.4 1.3 1.1 −0.4 1.4 1.4 −0.4 1.2 2 Grandiose sense of self−wortha 1.2 −0.5 1.1 1.7 −0.6 0.9 1.1 −0.6 1.1 1.6 −0.6 0.9 3 Need for stimulationc 1.3 −1.0 0.4 1.5 −1.6 −0.1 1.5 −1.0 0.2 1.7 −1.5 −0.1 4 Pathological lyinga 2.0 −0.2 0.9 1.5 −0.9 0.8 2.0 −0.3 0.8 1.5 −0.9 0.8 5 Conning/Manipulativea 1.6 −0.7 0.7 1.6 −0.7 0.7 1.7 −0.7 0.6 1.7 −0.7 0.6 6 Lack of remorse or guiltb 2.0 −1.5 −0.3 2.0 −1.5 −0.3 1.9 −1.6 −0.3 1.9 −1.6 −0.3 7 Shallow affectb 1.8 −1.2 0.4 1.8 −1.1 0.4 1.6 −1.4 0.3 1.8 −1.1 0.4 8 Callous/lack of empathyb 1.5 −1.3 0.4 2.1 −1.3 0.3 1.5 −1.4 0.3 2.1 −1.3 0.2 9 Parasitic lifestylec 1.1 −1.3 1.0 1.1 −1.3 1.0 1.2 −1.2 0.8 1.2 −1.2 0.8 10 Poor behavioral controls 1.5 −1.5 −0.2 1.1 −1.4 0.3 11 Promiscuous sexual behavior 0.9 −0.6 0.7 0.8 −1.1 0.4 12 Early behavioral problems 1.0 −0.7 0.3 1.0 −0.6 0.6 13 Lack of long term goalsc 1.7 −1.1 0.1 1.2 −1.6 0.2 1.8 −1.1 0.0 1.3 −1.6 0.2 14 Impulsivityc 1.6 −1.5 −0.1 1.4 −2.2 −0.4 1.9 −1.5 −0.2 1.5 −2.0 −0.4 15 Irresponsibilityc 1.8 −1.2 0.3 1.4 −2.1 −0.2 2.0 −1.2 0.2 1.5 −2.1 −0.2 16 Failure to accept responsibilityb 1.8 −1.1 0.0 1.2 −1.5 0.2 1.7 −1.2 0.0 1.0 −1.6 0.2 17 Many short term marriages 0.6 0.8 2.5 0.7 0.5 1.9 18 Juvenile delinquency 1.1 −0.3 0.5 0.8 −1.0 0.2 19 Revocation of release 1.3 −0.5 −0.1 0.8 −1.7 −0.4 20 Criminal versatility 1.6 −0.3 0.8 0.9 −0.6 1.2 Note: PCL-R = Psychopathy Checklist-Revised ( Hare, 2003). a Items that load on Arrogant and Deceptive Interpersonal Style. b Items that load on Deficient Affective Experience. c Items that load on Impulsive and Irresponsible Behavioral Style. Table options The existence of DIF necessitated anchoring of the latent trait; to ensure continuity with other studies we chose one item from each of the three lower-order factors in the hierarchical model, consistent with previous analysis for the UK (Cooke et al., submitted). The anchors were Items 5 (Conning/manipulative), 6 (Lack of Remorse or Guilt), and 9 (Parasitic lifestyle). Overall, the model fits the data well, with predicted responses for each item falling within 1% of the observed values. The IRT parameters for the constrained model are shown in Table 1. Given equivalent standing on the latent trait, participants from Europe had lower ratings on most of the 13 PCL-R items than did participants from North America. Finally, we replicated the previous analysis for all 20 PCL-R items across cultures using the same three anchors, Items 5 (Conning/manipulative), 6 (Lack of Remorse or Guilt) and 9 (Parasitic lifestyle). The results were unchanged. Bias at the item level (DIF) does not necessarily result in equivalent bias at the level of total scores (DTF), as summing items may either cancel out or amplify their bias. We plotted Test Characteristic Curves (TCCs) for ratings from Europe versus North America (Fig. 2). The TCCs indicated that the association between the latent trait and PCL-R scores varied across cultures. Participants from Europe obtained lower PCL-R total scores than did those from North America given the same level of θ, but only at lower levels of the trait (i.e., θ < 0.5). Test characteristic curves for PCL-R total scores. Fig. 2. Test characteristic curves for PCL-R total scores. Figure options We calculated the root differential test function (rDTF; Raju, Linden, & Fleer, 1995); this indexes the average difference between TCCs in the metric of the test (i.e., raw score units). For the 13 items included in the three-factor hierarchical model, rDTF was 1.1 points (p < 0.001) mean score = 11.9 (SD = 6.8). For the 20-item PCL-R total scores, rDTF was 1.5 points (p < 0.001), mean score = 17.5 (SD = 9.8). Although statistically significant, the magnitude of the rDTF was relatively small in practical terms for individual scores. To help identify the source of DTF observed, we conducted analyses comparing the test characteristic curves of the three lower-order factors of Cooke and Michie’s hierarchical model. The TCCs for Factors 1, 2, and 3 are presented in Fig. 3. The TCC for Factor 2 (Deficient Affective Experience) indicates it was more discriminating than the other factors, with a steeper slope at the point of inflexion. It also discriminated over a wide range of scores around average values of the latent trait. In contrast, Factor 1 (Arrogant and Deceitful Interpersonal Style) discriminated well at high levels of the latent trait, but not at low levels; it also failed to reach its maximum score even at high levels of the trait (i.e., θ = 3.0). By way of contrast, Factor 3 (Impulsive and Irresponsible Behavioral Style) discriminated best at low levels of the trait. It is noteworthy that DIF is present for high scores for Factor 1 but for low scores on Factor 3. (a) Test characteristic curves for Factor 1: Arrogant and Deceitful ... Fig. 3. (a) Test characteristic curves for Factor 1: Arrogant and Deceitful Interpersonal Style. (b) Test characteristic curves for Factor 2: Deficient Affective Experience. (c) Test characteristic curves for Factor 3: Impulsive and Irresponsible Behavioral Style. Figure options Next, we equated Factor scores across the samples using Items 5 (Conning/manipulative), 6 (Lack of remorse or guilt) and 9 (Parasitic lifestyle) as anchors, one anchor per factor. We then calculated rDTF with the following results: Factor 1, 0.29 (p < 0.001) mean = 2.6 (SD = 2.4); Factor 2, 0.07 (p < 0.001) mean = 4.2 (SD = 2.6); and Factor 3, 0.86 (p < 0.001) mean = 4.9 (SD = 3.1). Thus, cross-cultural bias was much smaller for Factor 2 than it was for Factors 1 and 3. Finally, we estimated factor information functions, which provide an estimate of the precision of measurement. In contrast to traditional reliability indexes, information indicates measurement reliability, or precision, across the full range of the latent trait (Fig. 4). Factor 2 provided more information than the other factors across most of the range of the latent trait; only at high trait levels (i.e., θ > 1.0) did Factor 1 provide more information. Factor 3 did not provide the most information at any point of the trait, despite the fact that it is made up of 5 items, versus 4 items for the other factors. Information functions for three-factor scores. Fig. 4. Information functions for three-factor scores. Figure options In summary, there was very little cross-cultural bias in ratings of affective symptoms of psychopathy, but distinct bias in ratings of the interpersonal and behavioral symptoms. In addition, the ratings of affective symptoms were, on average, more discriminating and more precisely measured than were ratings of the other symptoms. Taken together, these findings suggest that Deficient Affective Experience may be the pan-cultural core of psychopathy.