Refining the measurement of Axis II: A Q-sort procedure for assessing personality pathology

The measurement of personality disorders (PDs) has proven to be a difficult enterprise. This article describes two initial studies of the validity and reliability of the Shedler-Westen Assessment Procedure (SWAP), a Q-sort procedure that quantifies clinical judgment, which may be useful both for assessing personality pathology and for empirically refining Axis II categories and diagnostic criteria. In the first study, 153 clinicians from a random national sample used a version of the Q-sort to describe either a prototype or actual patient with either a borderline, antisocial, histrionic, or narcissistic personality disorder. Correlations between aggregated prototype and actual patient profiles provided evidence for convergent and discriminant validity, and a cluster-analytic procedure (Q-factor analysis) produced revised criteria for the four disorders that minimized the problem of comorbidity. In Study 2, a pilot sample of patients were interviewed using a clinical research interview that mirrors the way clinicians assess personality and PDs. The study yielded promising results with respect to the possibility of obtaining reliable Q-sort descriptions based on an interview that resembles a clinical interview rather than the direct-question format used in current Axis II structured interviews. It also produced strong correlations between Q-sort descriptions made by interview and those made independently by the treating clinician, further supporting the validity of the instrument. The findings suggest the potential utility of the SWAP as a measure of PDs and as a method for empirically refining Axis II categories and criteria.