Blog

Although there is an abundance of research suggesting that the PCL-R is the “gold standard”, it has some limitations. First, the PCL-R is highly labor intensive to administer and score. Requirements for accurate scoring include a trained examiner, a semi-structured clinical interview, and extensive review of all available institutional file data. This often results in a trained mental health professional spending three to four hours to administer and score the test, which can be costly (Edens et al., 2001).

Another limitation noted by these authors is that the ratings are based on institutional records. Too often institutional records are unavailable, incomplete, or omit certain information in which affects reliability of PCL-R scoring. Next, these authors explained that the PCL-R has been developed and validated on correctional and forensic samples, therefore, our understanding of psychopathic personality patterns has been based too heavily on unsuccessful criminal psychopaths. A final limitation is that the PCL and its derivatives is only one measurement of psychopathy and some researchers argue that the antisocial features (found in Factor 2, Facet 4) are not central to the disorder of psychopathy itself, rather, they are behavioral manifestations (e.g. Cooke, Michie, & Hart, 2006; Walters, Knight, Grann, & Dahle, 2008). As a result, criminality may not be a key component of psychopathy, but the predictive abilities of the PCL-R largely rest with the Factor 2 items.

It is important to note that due to several of the above limitations some evaluators and researchers choose to conduct file-based evaluations. Hare (2004) informed that in some situations it may be impossible to conduct an interview. For example, the individual may refuse to be interviewed or cooperate. Hare explained that a generous amount of research indicates that reliable and valid PCL-R ratings can be obtained based solely on high quality collateral information. Some advantages are that a substantial amount of useful information can be gathered in a short time and that it is unnecessary to wait years to find out if a variable predicts a behavior. Hare recommended that ratings based only on file/collateral information should be indicated in the evaluation and/or study.

In contrast, Hare (2004) explained that there are some consequences to only file-based evaluations. Specifically, it may be difficult to score some of the interpersonal and affective items without direct observation of the individual’s interpersonal style. In addition, ratings based on file reviews often result in considerably lower PCL-R scores than otherwise obtained from a standard administration. Presently, there is a lack of research in which standard assessments and file reviews are directly compared. Researchers have found that a file-based cut off score of 26 was optimal for predicting a standard PCL-R score of 30 (Hare, 2004).

To assist in resolving the lengthy administration and scoring time required by using the PCL-R, Hart et al. (2003) developed the Psychopathy Checklist: Screening Version (PCL:SV). The PCL:SV was developed to be a briefer measure useful for screening individuals who may have a low base rate of psychopathy (i.e., community individuals, sex offenders; Guy & Douglas, 2006). The PCL:SV can be used initially as a screening measure for individuals at less risk of being psychopathic, and then the more traditional PCL-R can be employed if someone reaches the cutoff score of 30. These authors explained that the PCL:SV conceptualizes psychopathy in an analogous manner as the PCL-R. Multiple studies suggest that the PCL:SV is strongly related to the PCL-R (Hart et al., 2003), however it requires half of the time to complete. Part 2 scores of the PCL:SV have higher interrater reliability than Part 1 and paralleled those found using PCL-R Factor 1 and Factor 2 scores, with averaged interclass Part 1 correlations of .92 for federal inmates, .90 for forensic outpatients, and .80 for forensic inpatients (Hart et al.).

The PCL:SV is comprised of twelve questions and is divided into part 1 and part 2, intending to be conceptual analogues of the PCL-R’s empirically derived Factor 1 and Factor 2 (Hart et al., 2003). Part 1 scores reflect interpersonal and affective symptoms of psychopathy. Part 2 scores reveal symptoms of social deviance related to psychopathy. The PCL:SV provides three dimensional scores with scores ranging from 0 to 24. Usually scores of 18 and above are only obtained by psychopaths and warrant administration of the full PCL-R. In contrast, scores of 12 and below are typically only obtained by non-psychopaths. Further, individuals who score between 13 and 17 may be psychopathic and should be further evaluated. Hart et al. emphasized that no single cut-off score is ideal for all situations.

Compared to the PCL-R, the PCL:SV is only in its infancy (Hart, 2003). Thus, there is a limited amount of research documenting the validity and reliability of this measure. Douglas et al. (2005) examined the reliability and validity of the PCL:SV in Swedish correctional and forensic psychiatric samples. The PCL:SV and HCR-20 were given to 560 male and female participants. Results suggested that the structural reliability of the PCL:SV was very high for most indices. Overall, the PCL:SV scores were meaningfully related to aggression toward others, supporting the construct validity of the shortened measure.

Hart et al. (2003) assessed the psychometric properties of the PCL:SV. The sample consisted of 586 participants. These included correctional offenders, college students, and civil and forensic psychiatric patients. These authors found that the PCL:SV had a high degree of internal consistency (ranging from .69 to .91), item homogeneity (ranging from .17 to .42), and interrater reliability (ranging from .67 to .92). These authors also found that the standard error of prediction (or SEM3) was satisfactory. Overall, the PCL:SV demonstrated sound psychometric properties.

Hart et al. (2003) also examined the concurrent, convergent, and discriminant validity of the PCL:SV. The PCL:SV Total scores correlated about the same as the PCL-R Factor 1 and Factor 2 scores, with a weighted mean correlation of .67 and .68 respectively. Hart et al. also found support for the PCL:SV subscales. Part 1 scores correlated higher with PCL-R Factor 1 scores than with PCL-R total or Factor 2 scores (.68, .61, and .40, respectively). Similarly, Part 2 PCL:SV scores correlated higher with PCL-R Factor 2 scores (r = .81) than with PCL-R Factor 1 scores (r = .48). Additionally, Hart et al. compared the PCL:SV to Antisocial Personality Disorder ratings and self-report measures. They found analogous results as the PCL-R. The weighted mean correlation between PCL:SV Total scores and APD ratings was .70, while the weighted mean correlation between the PCL:SV Total scores and the Millon Clinical Multiaxial Inventory-II Antisocial Scale was .68. Although there is minimal data regarding the association between psychopathy and age and race, these authors noted that they are independent variables that do not influence PCL:SV scores (weighted mean correlation of -.07 and -.08, respectively).  However, Hart et al. found significant correlations between Total PCL:SV scores and sex with a weighted mean correlation of -.33. These authors concluded that there may be a small significant difference in the prevalence of psychopathic symptoms with males having more symptoms than females.

In addition, Cooke, Michie, Hart, and Hare (1999) evaluated whether the PCL:SV could be regarded as a short form of the PCL-R. These authors analyzed the normative sample from the PCL:SV manual and reported that 8 of the 12 items in the PCL:SV strongly paralleled their equivalent PCL-R items. The other four PCL:SV items were equally useful or better than their corresponding PCL-R items. These authors hypothesized that this may be due to the revision process in which made the items more general. These authors concluded that the PCL:SV is an effective short form of the PCL-R.