Though physical abuse probably causes both violent and nonviolent offending, the review presented here strongly suggests that physical abuse distinguishes between violent and nonviolent offenders. It was beyond the scope of this chapter to determine under what circumstances this occurs or whether measures of severity of abuse, or the particular outcomes measured, or the samples used matter. We did find that the association was robust with respect to the application of a long series of control variables, including controls for parental warmth and child effects, and even nonviolent offending.
The findings were not as consistent for “mixed” indicators of abuse experience. This may be due to the fact that those indicators usually include neglect, and the evidence for an association between neglect and violence is weaker and less consistent, though more research would be needed to draw any firm conclusions. Sexual abuse does appear to be consistently associated with violent behavior and also remains a good prospect for identifying those at high risk for violent offending, as opposed to general offending. The findings for sexual abuse also seem to withstand the application of controls to test rival hypotheses, except most controls for parental warmth which may indicate that a protective influence of parental warmth is at work. The research on trauma thus far suggests an association with violence and the possibility that trauma differentiates between violent and nonviolent offenders. However, there are too few studies to draw any strong conclusions, and those studies use disparate operationalizations of trauma and violence, which make them a less cohesive whole than our tabular presentation makes them appear.