Does Measurement Bias Explain Sex Differences in Self-Reported Empathy?

Self-reports of trait empathic concern suggest that women are more empathic than men, while self-reports of state empathy yield small, inconsistent sex differences. Researchers have inferred from this pattern that sex differences in empathic concern reflect measurement biases that are elicited by personality questionnaires but not reports of emotion experience in real time. However, sex differences in trait empathy need not correspond to sex differences in empathy in any one situation because trait questionnaires measure empathy at a higher level of aggregation. Moreover, the hypothesis that trait questionnaires of empathy are more biased than state measures has never been empirically tested. We conduct measurement invariance tests on ten studies from our laboratory (N = 6,982). The effect of non-invariance upon observed sex differences were generally small, and we found some evidence that trait measures are actually more biased than state measures. A meta-analytic summary of bias-free estimates revealed that women on average do experience more state and trait empathy than men, although there was substantial heterogeneity in effect sizes.