Partner abuse, child abuse and neglect, and suicidal behavior are costly occurrences in American society, in terms? of both human suffering and economics. Healthy People 2010 targets all three problems for reduction. However,? brevention of each of these areas is hampered by inadequate knowledge about the relative relations of risk and? protective factors to the problems.? The purpose of this project is to use a large (over 100,000 respondents), representative archival sample? from 83 communities to provide effect sizes of risk factors at various levels (i.e., individual, family, work, and? community), to test hypotheses about the buffering effects of protective factors, and use these results to create? exploratory models of risk and resilience for these problems.? The archival data set contains detailed assessments of suicidality, partner and child physical assaults and their? impacts, partner and child emotional assaults and their impacts, and child neglectful omissions. We will test the effect? sizes of risk factors using correlations, the buffering effects of protective factors using correlations and backward? stepwise logistic and linear regressions, and overall models using structural equation modeling. We will use statistical? packages that can appropriately estimate standard errors in a complex data set collected at 83 communities.? Prevention science is guided by a five stage, recursive research cycle: (1) defining the problem and measuring its? prevalence; (2) establishing risk and protective factor relations with the problem; (3) efficacy trials; (4) effectiveness? trials; and (5) dissemination trials (Mrazek & Haggerty, 1994). Large gaps remain in the risk factor literature on all? forms of partner abuse and child maltreatment (Slep & Heyman, 2001); very little is known at all about putative? protective factor relations. Because this knowledge provides the bedrock on which to build effective prevention? activities, reaching the Healthy People 2010 objectives is unlikely without improvements. This project is unique and? innovative because (a) it is large enough to have very high power and low standard errors in its estimations of risk and? protective factor effect sizes and additive effects; (b) it will examine risk and protective relations for injury as well as? lower level behaviors (e.g., assault); (c) it will test these relations for partner and child physical abuse, partner and child? emotional abuse, child neglectful omissions (lack of supervision and exposure to physical hazards), and suicidality; (d)? it will examine risk and protective factors for comorbid conditions; and (e) it is large enough to examine these relations? for children in different age groups.