Assessment 17(2) 259–268 © The Author(s) 2010 Reprints and permission: http://www. sagepub.com/journalsPermissions.nav DOI: 10.1177/1073191109356685 http://asmnt.sagepub.com Violence Risk Assessment and Facet 4 of the Psychopathy Checklist: Predicting Institutional and Community Aggression in Two Forensic Samples Glenn D. Walters 1 and Kirk Heilbrun 2 Abstract The Psychopathy Checklist and Psychopathy Checklist–Revised (PCL/PCL-R) were used to predict institutional aggression and community violence in two groups of forensic patients. Results showed that Facet 4 (Antisocial) of the PCL/PCL-R or one of its parcels consistently achieved incremental validity relative to the first three facets, whereas the first three facets failed to achieve incremental validity relative to Facet 4. One of the two Facet 4 parcels, Parcel G (General Acting Out), was the only PCL-R measure to consistently achieve success in classifying individual cases using the receiver operating characteristic approach. These findings suggest that Facet 4 and its parcels may play a role in violence risk assessment, although the generalizability of these findings to various forensic settings and contexts requires further study. Keywords Psychopathy Checklist, Facet 4, Parcel G, forensic patients, violence risk assessment Violence risk in forensic patients has been assessed with a variety of different procedures, including the Psychopathy Checklist (PCL; Hare, 1980) and Psychopathy Checklist– Revised (PCL-R; Hare, 2003). Hare (1998) notes that the PCL/PCL-R was not designed for risk assessment even though it is capable of predicting violent recidivism at a level commensurate with actual risk assessment procedures. Studies conducted on institutional (Buffington-Vollum, Edens, Johnson, & Johnson, 2002; Kroner & Mills, 2001) and community (Glover, Nicholson, Hemmati, Bernfeld, & Quinsey, 2002; Grann, Långström, Tengström, & Kullgren, 1999) aggression indicate that the PCL-R may be predictive of both. However, the relationship between PCL-measured psychopathy and subsequent community violence is more robust than the relationship between psychopathy and insti- tutional misconduct (Edens, Petrila, & Buffington-Vollum, 2001). In fact, the aspect of institutional misconduct emphasized in many of the studies on the topic bears little resemblance to the physical violence that is of primary con- cern to clinicians and policy makers. In the Buffington-Vollum et al. (2002) study, for instance, the PCL-R predicted verbal institutional aggression but not physical institutional aggres- sion, and in the Kroner and Mills (2001) study, the PCL-R correlated only weakly with “major misconduct,” a category that included rioting, drug/alcohol abuse, and refusing a direct order. Questions have also been raised about the PCL-R’s ability to predict serious assaultive behavior during confinement (Edens, Buffington-Vollum, Keilen, Roskamp, & Anthony, 2005). It is becoming increasingly apparent that not all aspects of the PCL-R are equally effective in predicting violence. In a meta-analysis comparing the predictive efficacy of the two PCL-R factor scores, Walters (2003) ascertained that Factor 2 (Chronically Unstable Antisocial Lifestyle) was a significantly better predictor of institutional misconduct and recidivism than Factor 1 (Selfish, Callous, and Remorseless Use of Others). More recently, Walters, Knight, Grann, and Dahle (2008) discovered that one of the two facets that com- prises Factor 2 of the PCL-R, Facet 4 (Antisocial), predicted violent and general recidivism above and beyond the first three facets, whereas the first three facets generally failed to predict recidivism above and beyond Facet 4. The Walters et al. (2008) study also revealed that of the two Facet 4 parcel scores, Parcel H (Criminality) was a stronger predic- tor of general recidivism than Parcel G (General Acting 1 Federal Correctional Institution Schuylkill, Minersville, PA, USA 2 Drexel University, Philadelphia, PA, USA Corresponding Author: Glenn D. Walters, Psychology Services, FCI-Schuylkill, PO Box 700, Minersville, PA 17954-0700, USA Email: gwalters@bop.gov at DREXEL UNIV LIBRARIES on March 8, 2015 asm.sagepub.com Downloaded from