Meta-analyses of sex differences in physical aggression to heterosexual partners and in its physical consequences are reported. Women were slightly more likely (d = -.05) than men to use one or more act of physical aggression and to use such acts more frequently.Men were more likely (d = .15) to inflict an injury, and overall, 62% of those injured by a partner were women. The findings partially support previous claims that different methods of measurement produce conflicting results, but there was also evidence that the sample was an important moderator of effect size.
Four variables were also used to assess a specific hypothesis. I predicted that in dating relationships, w hich are typically found at younger ages and in student sam ples, m en w ould be m ore inhibited about using physical aggression toward their partners than would be the case for m en in m ore established (m arried or cohabiting) relationships, w hich are typically found at older ages. The ratio- nale is that w om en in dating relationships can term inate these m ore easily than they can a cohabiting or m arital relationship and also can make it widely known to the peer group that the man had been violent. A study com paring rates of m en-only and w om en-only partner physical aggression in dating, cohabiting, and marital relationships (Stets & Straus, 1989) found a much greater disparity in dating relationships. Here, women were about four times more likely than men to be the only one of the couple aggressing. In the other two cases, the proportions were similar for the two sexes.
One possible consequence of men being inhibited about using physical aggression tow ard a dating partner w ould be to m ake it safer for w om en to use acts of physical aggression. Fiebert and Gonzalez (1997) found, among a sample of female college stu- dents, that 29% admitted initiating assaults on a male partner. Of these, around half said that they had no fear of retaliation or that, because m en could easily defend them selves, they regarded their own physical aggression as not a problem. This reasoning would lead to the prediction that a larger effect size in the female direction w ould be associated w ith relationships that involve a lower proportion of men who show physical aggression to a partner and that these relationships typically involve dating rather than cohabitation or marriage. Two confounds in this analysis are that dating relationships occur at younger ages than cohabiting or marital relationships and that they have typically been studied in student rather than community samples. These two variables were therefore included in the analysis, the sam ple being dum m y coded as student or com m unity.
Therefore, according to self-reports, women are more likely than men to commit acts of physical aggression, whereas accord- ing to partner reports, their respective levels are sim ilar, although this is attributable to the outliers. This discrepancy posed a problem of how to proceed with the meta-analysis. Several studies used com posite m easures, derived from both self- and partner reports, and the majority of other studies included both values. Therefore, I considered that a com posite value w ould best reflect the overall central tendency across all studies. A t the sam e tim e, so as not to ignore the discrepancy between the sources of information, I calculated separate values for self- and partner reports when car- rying out the categorical m odel testing.
When measures were based on specific acts, women were sig- nificantly more likely than men to have used physical aggression toward their partners and to have used it more frequently, although the effect size was very small (d =.05). When measures were based on the physical consequences of aggression (visible injuries or injuries requiring medical treatment), men were more likely than women to have injured their partners, but again, effect sizes were relatively small (d = .15 and .08)
https://pdfs.semanticscholar.org/2f5...332.1614345879