Clinical Prediction of Violent Behavior

• NCJ Number: 78339
• Author(s): J Monahan
• Date Published: 1981
• Length: 143 pages
• Contract Number(s): 282-77-0176-KS
• Sponsoring Agency: US Dept of Health, Education, and Welfare
• Sale Source: Superintendent of Documents, GPO; National Institute of Justice/
• Publication Type: Statistics
• Language: English
• Country: United States of America
• Note: Crime and Delinquency Issues - A Monograph Series.

Annotation

This monograph is intended to assist practicing mental health professionals in understanding the issues involved in violence prediction so that they may improve the appropriateness and accuracy of their clinical predictions.

Abstract

Noting that the prediction of violent behavior has played an important role throughout legal history, the monograph defines violent behavior and describes three current major criticisms of violence prediction by mental health professionals: that violence cannot be predicted with any satisfactory level of accuracy, that any attempt to do so violates the civil liberties of the subjects, and that the societal protection function is at variance with the traditional helping role of the mental health professions. The monograph examines the concepts of predictor and criterion variables, outcomes of positive and negative prediction, decision rules, and base rates. It presents a model of the clinical prediction process and outlines variables that psychiatrists and psychologists rely upon to predict violent behavior. Common clinical errors in prediction such as lack of attention to base rates of violent behavior and reliance upon erroneous predictor items are described. In addition, a review of research on clinical prediction is presented that underscores both the importance of past violence as a predictor of future violence and the problem of accuracy in predicting violent behavior among institutionalized populations. Two additional approaches to improving the accuracy of clinical predictions are depicted: the incorporation of statistical concepts into clinical decisionmaking and the increased emphasis on situational or environmental predictors of violence such as a disturbance or deficit in family, peer, or job-support systems. Finally, a model format for the clinical prediction of violence is proposed that offers 14 questions for the clinician to consider in making a prediction of violent behavior. A hypothetical case study is used to describe how such questions might be answered, and a clinical report, based upon these answers, is presented. Tables and figures are given. A bibliography of 130 references is appended.