The Clinical Prediction of Dangerousness

Abstract

Laws committing dangerous sexual offenders indefinitely are legitimate only insofar as "dangerousness" can be predicted and treated. The court has relied on the clinician to identify and measure the crucial variables in violent behavior and assign a probability of occurrence. Yet most critical studies have concluded that not only has predictive accuracy not been demonstrated but also that such accuracy cannot be achieved. We argue that any conclusion at this time is premature. There are data indicating the soundness of clinical as opposed to statistical prediction of dangerous ness. Both approaches show high false positive error rates, but the im provement in accuracy of clinical studies over statistical studies cannot be denied. In examining the success of treatment, the authors present data indi cating that a program of psychological rehabilitation can have a profound effect on the majority of dangerous sexual offenders. The authors conclude that the indefinite commitment of dangerous offenders is legitimate and appropriate, but the policy carries with it heavy responsibilities. Enforcing such an approach means that the clinician must play an important role in his relationship to the court.