Sprock (2002) sent 89 licensed psychologists brief, hypothetical descriptions of prototypic and non-prototypic cases of the schizoid, antisocial, and obsessive-compulsive personality disorders. Each psychologist received three vignettes and was asked to describe each in terms of the 30 facets of the FFM. Internal consistency of the FFM descriptions was excellent for each of the personality disorders, with average interrater reliability ranging from a low of .51 for the schizoid case to a high of .64 for the obsessive-compulsive and antisocial cases. Furthermore, the FFM descriptions of the three prototypic cases agreed well with FFM descriptions of prototypic cases provided by personality disorder researchers surveyed earlier by Lynam and Widiger (2001). The correlations were .84 with the schizoid, .87 with the antisocial, and .86 with the obsessive-compulsive prototypes. Sprock concluded that these results further “suggest that practicing clinicians can directly apply the dimensions of the FFM to cases of disordered personality with a moderate level of reliability” (p. 417).
Samuel and Widiger (2004) asked clinicians in private practice to describe prototypic cases of each of the ten DSM-IV personality disorders in terms of the 30 facets of the FFM. One hundred fifty-four clinicians each provided ratings on two PDs using a FFM rating form that contained a series of 30 items corresponding to each facet of the FFM. The ratings were then averaged across participants to produce a mean consensus rating for each personality disorder in terms of the 30 FFM facets. The agreement among raters was good, with average inter-rater correlations above .64 for all 10 personality disorders. The consensus ratings converged well with the findings for the three prototypic cases from Sprock (2002), with correlations ranging from .83 for the obsessive-compulsive to .91 for the antisocial. Convergent validity