The psychometric properties of the Personality Assessment Inventory-Adolescent's Borderline Features Scale across two high-risk samples.

The present study sought to assess the performance of the Borderline Features (BOR) Scale of the adolescent version of the Personality Assessment Inventory (Morey, 2007b) in 2 high-risk samples: inpatient and justice-involved adolescents. This study is the first to evaluate the BOR scale in high-risk adolescent samples, outside the initial standardization studies. Across both samples (NClinical = 327, NForensic = 151), results indicated good internal consistency. Confirmatory factor analysis indicated poor fit of the 4-factor structure proposed by the measure's authors. Convergent validity and receiver-operating characteristics analyses, conducted in the clinical sample, indicated that the adolescent version of the Personality Assessment Inventory BOR scale had good diagnostic accuracy for predicting a borderline personality disorder diagnosis (via structured interview). Findings suggest that the BOR scale has adequate internal consistency, convergent validity, and clinical utility, although areas for future measure evaluation (including factor structure) remain. Still, the BOR scale may partially address the current hesitation to assess borderline personality disorder features in high-risk youth because it is embedded within a broadband psychopathology measure. (PsycINFO Database Record