A Dimensional Comparison of a Self-Report and a Structured Interview Measure of Conduct Disorder

Eligibility for a diagnosis of antisocial personality disorder (ASPD) requires evidence of antecedent conduct disorder (CD). Accurately identifying CD may be influenced by various factors, including assessment methodology. The present study used a two-parameter latent variable model to examine the relative performance of a self-report measure and a structured clinical interview in retrospectively detecting the CD spectrum among adult male offenders (N = 1,159). Self-report and clinical interview tended to converge regarding the rank order of severity indicated by CD symptom criteria. In addition, at relatively low levels of CD severity, self-report provided more information about the CD spectrum than did clinical interview. At relatively higher levels of CD severity, however, clinical interview provided more information about the CD spectrum than did self-report. Latent variable models offer a potential means of combining multiple assessment methods in a way that maximizes information gleaned by capitalizing on the contextual strengths of each approach.