Exposure to Domestic Violence, Gender Ideology, Rape Myths and Sexual Assault Perpetration: A Structural Equation Modeling Analysis

Abstract

The prevalence of male perpetrated sexual assault and its antecedents have long remained topics of great interest and discussion among researchers, educators and practitioners. This study sought to extend and refine existing research on the links between masculinity and sexual assault by exploring their relation to gender ideology and prior exposure to domestic violence in childhood. Specifically this thesis examined the link between exposure to domestic violence in childhood, perpetrated by fathers, gender ideologies, rape myth acceptance and sexual assault among undergraduate men, using previously collected longitudinal data. Based on previous literature, grounded in various theoretical frameworks, I proposed a linear pathway linking these variables. I theorized that young boys exposed to parental violence in the home would internalize traditional gender role beliefs; these stereotypical gender beliefs would then increase their likelihood of developing hypermasculine attitudes characterized by callous sexual attitudes and an increased probability of accepting rape myths. I proposed that the convergence of these antecedents would lead to a higher risk of sexual assault perpetration. Results from a bivariate correlation analysis and structural equation modeling provided strong support for my linear model. Findings indicated that young men exposed to domestic violence in childhood were more likely to internalize traditional gender role beliefs. Subsequently those young men who internalized traditional gender role beliefs were more likely to adopt a hypermasculine gender ideology. Higher hypermasculine attitudes among respondents led to significant increases in rape supportive beliefs. Consequently high levels of rape supportive attitudes to significant increases in reported sexually assaultive behavior. Findings also revealed that higher alcohol consumption had a positive direct effect on almost every variable in the model, while race and relationship status both had minor to zero effects. The implication of these findings, intervention strategies, limitations and suggestions for future research were also discussed.