Abstract
The present study attempted to evaluate the validity of Paraphilic Coercive Disorder (PCD) by studying the psychometric and statistical properties of the recently developed Paraphilic Coercive Disorder Checklist (PCDC; Zinik & Padilla, 2010). At present the PCDC is the only extent scale that purports to differentiate diagnostically offenders with PCD from other types of sexual offenders. In an attempt to assess the validity of the PCDC, this study used archival files and an already existing extensive database to rate the items of the scale on a large sample of incarcerated sex offenders who were part of the Massachusetts Treatment Center (MTC) research program [n = 812]. A subset of this offender sample [n = 138] that had also been administered the Multidimensional Assessment of Sex and Aggression (MASA), which is a well validated self-report inventory containing scales assessing multiple domains relevant to PCD (Knight & Cerce, 1999; Knight, Prentky, & Cerce, 1994). Identical analyses were also performed on subsets comprised solely of rapists [n = 336] and child molesters [n = 337]. Principal components analyses were calculated on the full sample, as well as on the MASA, pure rapist, and pure child molester subsets. The results of these principal components analyses showed between three to five factors depending upon the sample tested. In both the full and subsamples many items, especially the frequency of offending items, correlated negatively with well-validated MASA scales that theoretically should capture constructs related to PCD (e.g., sexualization and sadism). Although some significant correlations emerged with other factors (most notably with Sexual Deviance and Coercive Sexual Behavior factors), most MASA correlations were weak and non-significant in the full sample and subsamples. Last, although several of the emergent PCDC factors in the full sample and each subsample correlated positively with offender scores on the Psychopathy Checklist-Revised (PCL-R), Factor 2 scores reflected stronger associations than Factor 1.