Gender differences in understanding the meanings of affect cues, often labeled emotion recognition, have been studied for over a century. Past reviews of the literature have concluded that girls and women score higher than boys and men on tests of accuracy in decoding affect cues, which are most often tested in the cue modalities of face, body, and content-free voice. The present meta-analysis updates knowledge on this topic by including many more studies (1188 effect sizes in 1011 studies; total = 837,637) and examining a wide range of moderators such as health status of sample, international location, cue channels of the test, and other sample and test characteristics. Indeed, the gender difference favoring girls and women still exists, and evidence for publication bias was weak. The difference is not large ( = 0.12, = 0.24), but it is extremely consistent across many moderators, which, even when significant, show minor differences. Health status was the only moderator to produce groups without a significant gender difference.
