Development of an Actuarial Model for Prediction of Dangerousness in Forensic Patients (From Forensic Psychiatric Patient in Texas - Historical Perspective and Normative Research on Dangerousness, Appendix 6, 1980 - See NCJ-85168)

• NCJ Number: 85174
• Author(s): J M Mullen; M Mason; R Reinehr; H K Dudley
• Date Published: 1980
• Length: 23 pages
• Sale Source: National Institute of Justice/
• Language: English
• Country: United States of America
• Note: Available on microfiche as NCJ-85168

Annotation

Actuarial prediction of the dangerousness of a sample of Texas forensic patients using demographic and psychological test data was made to two criteria: the judgment of a research board and the postrelease history of patients released during the time frame of the study.

Abstract

All male patients admitted to the maximum security unit of Rusk State Hospital during a 12-month period (n=269) were administered a battery of psychological tests, and demographic information was obtained. For each person in the sample, a research panel composed of a psychologist, social workers, and a psychiatric aide made a dichotomous decision about the dangerousness of the patient. Agreement by two or more panel members was the criterion for placement of a person in the dangerous or nondangerous group. Differences between the dangerous and nondangerous groups on demographic and psychological test data were evaluated by chi-square, analysis of variance, and discriminate analyses. A subsequent sample of subjects was collected and similar analyses computed for cross validation purposes. Stepwise discriminate analysis showed that nine variables provide the most efficient model for differentiating between the dangerous and nondangerous groups. The variables in their contributing order are marital status; Holtzman Inkblot Technique variables of hostility, penetration, popular, verbal hostility, fighting, and crime; and the Buss-Durkee negativism and irritability scales. Tabular data and 10 references are provided.