A contrastive analysis of the factorial structure of the PCL-R: Which model fits best the data?

The aim of this study was to determine which of the factorial solutions proposed for the Hare Psychopathy Checklist-Revised (PCL-R) of two, three, four factors, and unidimensional fitted best the data. Two trained and experienced independent raters scored 197 prisoners from the Villabona Penitentiary (Asturias, Spain), age range 21 to 73 years (M = 36.0, SD = 9.7), of whom 60.12% were reoffenders and 73% had committed violent crimes. The results revealed that the two-factor correlational, three-factor hierarchical without testlets, four-factor correlational and hierarchical, and unidimensional models were a poor fit for the data (CFI ≤ .86), and the three-factor model with testlets was a reasonable fit for the data (CFI = .93). The scale resulting from the three-factor hierarchical model with testlets (13 items) classified psychopathy significantly higher than the original 20-item scale. The results are discussed in terms of their implications for theoretical models of psychopathy, decision-making, prison classification and intervention, and prevention.