Cross-sex theory of mind in the domain of sexual violence: upset, fear, and perceived likelihood

Abstract

An evolutionary perspective on theory of mind (ToM) leads us to hypothesize that inferences about others' beliefs, desires, and emotions are somewhat domain specific. In domains in which women and men have recurrently confronted different adaptive problems, selection should favor sex-differentiated ToM design features. One such domain centers on sexual violence; men have been the primary perpetrators and women the primary victims over time and across cultures. Using a mixed-subjects design (N = 781; 39% men), we tested two preregistered competing evolutionary hypotheses of sex-differentiated ToM inferences. The byproduct hypothesis posits that mind-reading errors occur because individuals use their own sexual psychology as a reference point, leading to an egocentric bias when inferring emotional reactions in the other sex. The adaptation hypothesis posits selection has favored adaptive inferential biases, analogous to men's oversexualization bias, that function to minimize more costly errors even if they result in more frequent low-cost errors. These biases could function to facilitate some sexual strategies. Underestimating victims' upset from sexual assault, for example, could facilitate a sexually exploitative strategy. We tested several predictions from both the byproduct hypothesis and the adaptation hypothesis across dimensions of cross-sex inference—upset, fear, and perceived likelihood of sexual victimization in a sample collected from social media and a public university in the southern United States. Men, on average, statistically underperceived women's self-reported upset following intimate partner sexual assault (Cohen's d = 0.48). Women, on average, statistically overperceived men's self-reported upset (Cohen's d = 0.64). In contrast, men and women were relatively accurate in predicting the opposite sex's fear of sexual violence. Men accurately inferred women's perceived likelihood of sexual victimization; whereas women statistically overestimated men's perceived likelihood of victimization (Cohen's d = 0.69). We tested individual differences theoretically relevant to the adaptation hypothesis. Discussion focuses on evidence bearing on adaptation and byproduct hypotheses.