ABSTRACT

Despite advances in violence risk analysis, which have resulted in a variety of reliable and valid norm-based measures, opinions regarding the clinical utility of violence prediction remain split. Predicting non-violence accurately (true negatives) has never been difficult for persons with no history of violence. This is due to the typically low base rate of violence when population base rates are utilized. Accurately predicting violence (true positives) for those persons with previous violence has met with substantial failure when unstructured clinical interviewing and judgment were employed as the primary method. A disproportionate number of false positives can be expected, in part, because of the low base rate phenomenon. U.S. Supreme Court rulings and findings suggest structured interviewing and empirically based testing with their built-in base rates will be helpful in legal situations where dangerousness is an issue. This chapter reviews some of these challenges including the reluctance of forensic professionals to conduct a deception analysis as an integral part of the violence risk analysis. Suggestions for increasing the accuracy of true positive predictions in forensic settings and situations are rendered.