Source: (2000) Journal of Interpersonal Violence. 15(6): 631-650.

This study was a test of general hypotheses on the significance of resources to risk-only models. Its purpose was to test whether adding resources to models of risk would move some participants from high-risk to lower risk categories. To do this test, the authors used a known-group logistic hierarchical regression analysis composed of 1,311 male and female prison inmates, and two different samples of male and female noninmates totaling 1,703 comparison participants. The authors found through four separate tests that adding resources significantly increased the fit of the model and the percentages of correctly classified inmates and noninmates. The results of the present research provide empirical support for the idea that risk-only models leave out important information related to resources. Assessments that include personal, familial, and community resources and risks will be more accurate and comprehensive than those that focus on risks alone. Author's abstract.