Date of Award

Fall 12-2013

Degree Type

Masters Thesis

Degree Name

Master of Arts (MA)



Committee Chair

Randolph Arnau

Committee Chair Department


Committee Member 2

Bradley Green

Committee Member 2 Department


Committee Member 3

Patrick Carnes


How do pathological personality traits relate to specific expressions of compulsive sexual behaviors? Previous studies have concluded that there are only small relationships between these two constructs, but such studies have usually conceptualized sexual compulsion as a unidimensional construct. The current study used a correlational design to evaluate the relationship between pathological personality traits and six general domains of sexually compulsive behaviors and cognitions in an inpatient and outpatient clinical sample (N = 540) of males seeking treatment for sex addiction. The pathological personality trait results were assessed using the MMPI-2 Psychopathology Five (PSY-5) domain and facet traits. The sexually compulsive behaviors and cognitions were measured by seven higher-order factors of the Sexual Dependency Inventory-4.0 scales. Based on the findings of previous studies and theory, it was expected that the PSY-5 factors and facet traits Disconstraint, Negative Emotionality/Neuroticism, Psychoticism, and Mistrust would have small to moderate associations with sexual compulsivity in the sample. Data were analyzed at three different levels. First, zero-order correlations were used to analyze the relationship between the PSY-5 domain and facets scales, SDl-4.0 higher order factors, and SDI-4.0 individual scales. Next, multiple regression analyses were conducted to see how each PSY-5 domain and facet scales related to each of the higher-order factors of sexual compulsive behaviors. Finally, canonical correlation analyses were used to explore the relationships between the PSY-5 domain and facet scales with the SDI-4.0 higher order factors at the multivariate level.

The results were largely consistent with what was hypothesized. The PSY-5 domains primarily related to the measures of sexual addictions and cognitions were Disconstraint, Negative Emotionality/Neuroticism, and Psychotic ism. The PSY-5 facets added greater definition to the results and accounted for more variance in the canonical correlations analysis compared to the canonical correlation analysis of the PSY-5 domains. The findings suggest that the PSY-5 facets may have clinical utility above and beyond the PSY-5 domains due to their greater specificity, enabling clinicians to target very specific problematic traits in therapy.