WITHDRAWN

A vast literature suggests that gender cues and emotional expression interact to shape emotional judgments of faces, despite earlier evidence indicating independent processing.Specifically, studies consistently reveal that male faces are more likely to be perceived as angry, while female faces are more likely to be perceived as happy.These biases have been attributed to socially-learned associations between sex and emotion or physical resemblance between facial cues.However, most studies in this area have used limited and non-diverse stimuli, such as white, computer-generated faces, which have been criticised for their limited generalizability.Additionally, most related studies make gender-based inferences using binary categories of gender, which disregards the variety in facial sex-typicality between women's and men's faces.Given these two considerations, the present study investigated whether the two sex-expression biases replicate using a more ecologically valid and racially-diverse stimulus set of and how varying masculinity and femininity affect these biases.Consistently, the results illustrated faster categorisation of anger in male faces and happiness in female faces.Feminising faces also enhanced happiness recognition regardless of sex, while masculinisation did not significantly influence anger recognition.These results robustly support the interdependent processing of sex and expression and highlight the importance of both bottom-up and top-down mechanisms underlying the interaction between gender and expression in emotion recognition.