Understanding the Psychopathic Personality Inventory (PPI) in terms of the unidimensionality, orthogonality, and construct validity of PPI-I and -II.

Comments on an article by Marcus et al. (see record 2011-23134-001). Marcus et al. (this issue) hold that PPI-I and PPI-II are generally orthogonal (uncorrelated). Examination of the primary PPI scales calls this into question. Specifically, the scales that make up Psychopathic Personality Inventory (PPI)-I and PPI-II are differentially correlated in offenders (Neumann et al., 2008) and college students (as reported in Lilienfeld, 1990, final sample). This pattern of correlations indicates that the socalled orthogonal association between PPI-I and PPI-II may be attributable to a statistical "washout" effect, rather than zero correlations across scales. The scales that make up PPI-I and PPI-II may also be correlated differentially with external correlates, rendering interpretation of PPI-I and PPI-II associations with such correlates ambiguous. The overlap of PPI scale scores among forensic and nonforensic samples also is problematic, particularly for the PPI-I scales.