Reliability and Construct Validity of the General Factor of Personality Disorder

Criterion B of the alternative model of personality disorder in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), outlines maladaptive trait dimensions that characterize personality disorders. Emerging evidence from bifactor confirmatory factor analyses suggest these traits are related at a higher order level by a general factor of personality disorder (g-PD). Further, emerging evidence points to traits most closely related to borderline personality disorder as underpinnings of g-PD. Further investigation is required to better understand the shared basis of personality disorder, with attention to the reliability and validity of g-PD. The g-PD theory was examined in a clinical (n = 242), and community sample (n = 252) of adults, using a brief form of the Personality Inventory for DSM-5 (PID-5). Structural analyses supported a correlated 6-factor model and a bifactor solution, validating the g-PD structure. Reliability indices supported the unidimensionality, reliability, and replicability of the g-PD factor. The strongest loading and most unidimensional items on the g-PD factors were from the Negative Affectivity and Disinhibition trait domains, partially replicating the trait profile of borderline personality disorder traits. In validity analyses, the nomological network of the general and specific factors were examined. g-PD was more strongly correlated with internalizing measures and impairment than specific factors, but specific factors were more strongly correlated with thought disorder and externalizing measures than g-PD. Our results support the nature and reliability of a general factor characterized by Negative Affectivity and Disinhibition unifying personality disorder traits in a brief form of the PID-5. Implications for the alternative model of personality disorder, PID-5, and g-PD theory are discussed.