Studies of attitudes toward mental illness indicate that community residents fear the mentally ill--they perceive them to be dangeous. Studies comparing arrest rates of former mental patients to those of the general public have led to the proposal that a "kernel of accuracy" characterizes the publics' perception. However, these studies have been challenged for their exclusive use of arrest data, their use of state hospital patients only, and their failure to adjust for sociodemographic and contextual variables.

The proposed research will add to our knowledge about the violent/illegal behavior of the mentally ill in three ways. First, both self report and arrest rate data will be used as indicators of violent/illegal behavior. Second, the inquiry will use samples collected in a psychiatric epidemiology study of the Washington Heights section of New York City that includes a variety of patient types (N=218), samples of persons in treatment for drug, alcohol or antisocial behavior problems (N=165) and a sample of community residents (N=267). Third, the research will control for relevant social and psychological differences between patient and non-patient groups.

For this proposal differences between patients and community residents self reports of violent/illegal behavior were computed. Consistent with arrest rate studies current mental patients (excluding drug, alcohol and antisocial respondents) were more likely than community residents to engage in violent/illegal behavior. The specific aims address four questions to help explain why this is so: 1) Do patients have sociodemographic characteristics that are associated with higher rates of violent/illegal behavior? 2) Do patients have higher rates of violent/illegal behavior because they are selected from neighborhoods where such behavior is more prevalent? 3) Are higher rates of violence/illegal behavior among patients due to specific aspects of psychopathology such as drinking or drug use? and 4) Which diagnostic groups account for the higher rate of violence among patients and do untreated cases of psychopathology show high rates of violence? Answers to these questions will have relevance to theory and will provide more refined knowledge for public education purposes. To answer them funds are needed to 1) to merge census tract data to individual records, 2) to obtain and computerize arrest/conviction data and 3) to support data analysis and write up costs.

R03MH043610

1987-09-15

1989-08-31