DSM-5 Section III model of personality disorders: Developing a measure of Criterion A

Date of Award

Fall 2015

Document Type


Degree Name

Doctor of Philosophy (PhD)


Psychology and Behavioral Sciences

First Advisor

Tony R. Young


Disagreement has long existed within the mental health field about the nature of personality dysfunction (Livesley, 1986; Millon & Davis, 1996). In 2013, the Personality and Personality Disorders Work Group (PPDWG) for DSM-5 proposed an empirically supported, dimensional model for Personality Disorder (PD) diagnosis. Criterion A of the model was comprised of problems of Self (as related to Identity, Self-Direction) and Interpersonal Problems (as related to Empathy, Intimacy). Despite the model's empirical support, it was placed in Section III of DSM-5, as an appendix (Widiger, 2013).

The purpose of the present study was to develop and validate a measure of Criterion A and to examine its relation to other components of the Section III model and to known sequelae of personality pathology. Initial items for the measure were compiled and written by the author and reviewed by a small content validation sample. Results of this sample were used for initial item reduction and revision. The Criterion A measure was then completed by a 210-participant pilot sample, along with the Personality Inventory for DSM-5-Brief Form (PID-5-BF; Krueger et al., 2013), the Psychological Well Being (PWB; Ryff, 1989), and the Measure of Disordered Personality Functioning, 20-item version (MDPF; Parker et al., 2004). Principle Components Analysis, parallel analysis, and internal reliability analyses were used to derive the final, four-factor, 43-item measure, the factors of which corresponded ideally to the four Criterion A domains. The resulting measure was then completed by both a non-clinical sample (N = 204) and a clinical sample (N = 107). Confirmatory Factor Analyses (CFA) of the non-clinical and clinical data indicated poor fit with the four factor model, though participant scores from the Criterion A measure in both the non-clinical and clinical sample continued to demonstrate expected relationships with other measures of personality dysfunction. Results indicate that the Criterion A measure possesses promising relationships with the Section III model and the other experimental measures, and has a four factor structure for targeting all four domains of Criterion A. Additional research of this kind is recommended to further examine the measure's capabilities.