Patterns of Sexual, Physical, and Emotional Abuse of Women by College Men: Multivariate Examination of Attitudes Toward Women, Social Environment, Psychopathology, and Childhood Exposure to Abuse

Abstract

This study tested whether attitudes toward women, social environment, psychopathology, and childhood exposure to abuse each contribute unique, nonredundant information in the prediction of men's patterns of abusing women. Individual sexual abusiveness, group rape/group sex involvement, physical abusiveness, and emotional abusiveness were examined. Data were analyzed from 255 college student men who anonymously reported their abuse of women since the age of 14. All four types of predictor variables were significantly related to total abusiveness after controlling for the other three types of predictors. However, only attitudes toward women and involvement in an alcohol centered social environment were each related to sexual abusiveness after controlling for the other three predictor groups. With respect to emotional abusiveness, that same was true for childhood exposure to abuse and attitudes toward women, but not for involvement in an alcohol centered social environment. Physical attacks were not predicted by any variable group after controlling for all others. A stepwise discriminant function analysis differentiated (1) nonabusive men, (2) men who were sexually abusive only, (3) men who were nonsexually abusive only, and (4) men who were abusive toward women in multiple ways. Men who were sexually abusive only were distinguished by the highest levels of alcohol use, finding dates at parties and bars, and fraternity affiliation. Men who were nonsexually abusive only were distinguished by the highest levels of hostility toward women; furthermore, their levels of antisocial tendencies and cognitive-emotional disturbance were almost as high as those of the multiply abusive group. The multiply abusive group also had the highest levels of sexual arousal to rape vignettes, was the most reactive to peer pressure to put their girlfriend "in her place," and had the highest levels of childhood experiences of having witnessed their fathers behaving in disrespectful ways toward their mothers. The nonabusive group was characterized by much lower scores on all predictor variables than the other three groups on all but two of the discriminating variables. The discussion addresses implications of the findings for intervention programming, theory, and future research.