The present study investigates a dimensional scheme for assessing the domain of Personality Disorder (PD) and compares such an approach to the current categorical system. Three samples of patients (state hospital inpatients, metropolitan hospital in- and out-patients, and university counseling center patients; n = 50 each) will be extensively evaluated using traditional methods of clinical diagnosis (e.g., structured interviews), measures of social and occupational functioning, and scales specifically developed to measure dimensions underlying the personality disorders. These three diverse samples are included to assure a broad representation of the relevant pathology, because they are expected to show differing prevalence rates of the various personality traits and disorders. The date will permit direct comparison of the categorical (diagnoses derived by interview) and dimensional (trait profiles generated by the scale scores) approaches to assessment of personality disorder. Given expected correspondences between certain diagnoses and scales (e.g., Paranoid PD with Mistrust), one question of primary interest will be whether the scales are bimodal, thus supporting a categorical approach; or normally distributed, indicating that continuous dimensions may more accurately describe personality pathology. Examination of the distributions will be followed by chi-square analyses to test whether the traits are specific to particular diagnoses or are broadly distributed across diagnoses. Relations between the diagnoses and trait scores will be further examined through a series od multivariate analyses, including factor and cluster analyses, and multiple regression. Finally, relations between various indicators of social/occupational adjustment and both the diagnoses and trait scores will be examined, also using multivariate techniques. It is possible that each approach--the dimensional and categorical--will provide appropriate for assessing some types of personality disorder, but not for all. In any case, the trait scale data will provide valuable information regarding personality disorder, and even if a categorical approach is supported, the scales (used individually or in configural combinations) will likely offer a convenient, cost-effective method of assessment. If, on the other hand, a dimensional approach is suggested, the scales can provide the foundation for a reconceptualization of the domain of personality disorder in subsequent DSM revisions.
