The Clinical Assessment and Prediction of Violent Behavior:

Abstract

An argument is made that many recent studies on the clinical prediction of dangerous behavior have been, given the methodological complexities involved, overly consumed with questions of validity. It is suggested that while issues of accuracy of outcome will and should remain central, there is also value in concentrating on the sheer description of the activities involved in the making of clinical decisions, on the use of contrived experimentation under controlled conditions, and on attempts to establish interclinical agreement with regard to the ascription of dangerousness. It is proposed that studies on accuracy of outcome require edification through the examination of how individual clinicians and members of interdisciplinary groups arrive at their extrapolations and how they infer causality. Several studies, two based upon the author's own work, are reviewed in an attempt to show how a firm base can be laid for future research on the gauging of clinical expertise.