The purpose of the proposed project is to identify the relative contribution of situational stressors, support, and structure as moderators of violent behavior in specific ecological settings: peer, work, family, and treatment. Based on a reciprocal causality model and analysis from prior research, a prospective design with repeated measures and a cross-validation procedure will be utilized to answer specific research questions regarding situational predictors of violent outcome. A total of 1056 adult males presenting at the Western Missouri Mental Health Facility (WMMHF) will be interviewed to obtain predictor measures (Violence Proneness Index; Ecological Participation Patterns; stress, structure, and support in specific settings) and will be categorized into a potentially violent group (744) and a non-violent control group (312). For subjects in the potentially violent group, post-release follow-up interviews will be conducted at three six-month intervals. Violent/non- violent outcome after release will be measured by an arrest or admission for violence or self-reported violence during these three consecutive follow-up periods. Potentially violent subjects will be divided into a calibration and cross-validation sample. Structural equation models will be used to test the proposed model in the calibration sample and the final model in the cross-validation sample. The anticipated final product is a violence risk scale that will include critical situational factors as well as individual proneness; the investigators suggest that such a scale will be useful in treatment delivery settings as a basis for decision-making with respect to treatment, release, follow-up, and case management plans.
