Abstract
This study investigated how clinicians characterize patients whom they have diagnosed as having a dependent personality disorder (p.d.). For each subject, the clinician assembled an attribute profile by choosing from a list of distinguishing diagnostic criteria for each of eight clinically relevant personality dimensions (i.e. behavior, interpersonal conduct, cognitive style, etc.). The criteria list represented personality prototypes for each of the eleven DSM-III personality disorders. Approximately 30 clinicians from 20 sites completed forms on 310 patients. For 44 of these patients, a base rate of 14%, the most prominent Axis II diagnosis was dependent p.d. One major method of analysis was the positive predictive power statistic, which gives the probability of the diagnosis given the symptom, a process similar to the process of formulating a diagnosis. It was found that the theoretically derived attributes of the dependent p.d. prototype were empirically associated with the clinical diagnosis, with behaviorally incompetent, interpersonally submissive, inept self-perception, and introjection defense mechanism the best predictors. The dependent p.d. prototype was less well agreed upon than the compulsive but better agreed upon than the narcissistic, histrionic, or borderline personality disorder prototypes. Other findings included that although depression was the principal Axis I disorder and marital or family tension was the chief psychosocial stressor, these were not differentially associated with the dependent p.d. Also investigated was the pattern of co-occurrence with other personality disorders. The final area investigated was sex bias in this diagnosis, and it was found that the presence of the theoretical dependent p.d. attributes were more predictive of the diagnosis in women than in men, supporting a hypothesis of sex bias in this diagnosis.