A Classification Tree Approach to the Development of Actuarial Violence Risk Assessment Tools

Since the 1970s, a wide body of research has suggested that the accuracy of clinical risk assessments of violence might be increased if clinicians used actuarial tools. Despite considerable progress in recent years in the development of such tools for violence risk assessment, they remain primarily research instruments, largely ignored in daily clinical practice. We argue that because most existing actuarial tools are based on a main effects regression approach, they do not adequately reflect the contingent nature of the clinical assessment processes. To enhance the use of actuarial violence risk assessment tools, we propose a classification tree rather than a main effects regression approach. In addition, we suggest that by employing two decision thresholds for identifying high- and low-risk cases--instead of the standard single threshold--the use of actuarial tools to make dichotomous risk classification decisions may be further enhanced. These claims are supported with empirical data from the MacArthur Violence Risk Assessment Study.     

Using the Referral Decision Scale to Screen Mentally Ill Jail Detainees: Validity and Implementation Issues

The prevalence of serious mental illnesses in jail populations is significantly greater than in the general population. Identifying individuals who warrant psychiatric evaluations is important and benefits correctional staff as well as detainees. One widely used screening instrument intended for this task is the Referral Decision Scale (RDS). This paper reviews the development and validation of the RDS. Using data from a multisite study which assessed postrelease outcomes for detainees with mental illness, various types of validity are addressed. The results confirm that the RDS has some inherent characteristics that seriously limit its practical application as a screening instrument for use by correctional staff.     

Estimating mental health needs and service utilization among prison inmates.

A sample of 3,684 inmates in the New York State prison system was surveyed in May 1986 to determine the prevalence of psychiatric and functional disability and service utilization. It was estimated that 5 percent had a severe psychiatric disability, and 10 percent had significant psychiatric disability. The higher the level of disability, the greater the proportion of inmates that had received mental health services in the last 30 days and in the last year. Still, 45 percent of the severe disability group had no service contacts in the last year. Patterns of utilization differed significantly by sex (a greater proportion of women received services) and by race (a greater proportion of whites received services). The clinical factors associated with receipt of services varied considerably between men and women.     

Enrollment in Mental Health Courts: Voluntariness,Knowingness, and Adjudicative Competence

Mental health courts (MHCs) are rapidly expanding as a form of diversion from jails and prisons for persons with mental illness charged with crimes. Although intended to be voluntary, little is known about this aspect of the courts. We examined perceptions of voluntariness, and levels of knowingness and legal competence among 200 newly enrolled clients of MHCs at two courts. Although most clients claimed to have chosen to enroll, at the same time, most claimed not to have been told the court was voluntary or told of the requirements prior to entering. The majority knew the “basics” of the courts, but fewer knew more nuanced information. A minority also were found to have impairments in legal competence. Implications are discussed.     

The statistical prediction of violent behavior

Considerable attention in the popular and professional literature has focused on the relative costs and benefits of using public protectionist versus civil libertarian models of social control. In this article, we develop statistical analogs of these models to predict violent behavior among samples of defendants found incompetent to stand trial in New York State. The societal and personal costs (errors of prediction) of each model are compared and their implications for clinical practice and social policy are discussed.This research was supported in part by PHS Grant MH 20367 from the NIMH Center for Studies of Crime and Delinquency. The assistance of Thomas Arvanites in data analysis and the comments of Monroe Lefkowitz and Mary Evans Melick on earlier drafts of this paper are gratefully acknowledged.