The Use of Violence in an Offense in a Sample of Convicted Rapists Within a Population Heterogeneity/State Dependence Framework

Abstract

The use of violence in a rape offense is an important but neglected question. This study seeks to enumerate the variables that predict the use of moderate and severe violence in a rape offense within a population heterogeneity/state dependence framework. Population heterogeneity asserts that crime or violence is caused by an underlying propensity while state dependence argues that prior crime or violence can increase or decrease the likelihood of future crime. A mixed model asserts that time-stable traits predispose an individual towards a certain level of crime or violence and time-varying characteristics can amplify or diminish this underlying risk. A sample of 222 convicted rapists from the Massachusetts Treatment Center was assessed on developmental, relationship and job attributes. A multinomial logistic regression analysis was performed on three levels of the outcome: no violence, moderate violence or severe violence. The results support a mixed model. Theoretical and policy implications are discussed.