UC Berkeley

Individuals demonstrate implicit evaluative biases with respect to multiple dimensions of socialcategorization. However, little is known about how such implicit biases manifest toward targets displaying simultaneously intersecting social categories. Across four studies (N = 4,314) we used Single-Target IATs (Studies 1-4) and Evaluative Priming Tasks (Study 4) to test competing hypotheses concerning implicit evaluations of multiply categorizable targets varying in race, gender, social class, and age. Overall, we observed a dominant pro-female/anti-male bias, which accounted for more target-level variation in implicit evaluations than race-, class-, or age-related biases. We also documented smaller and less consistent pro-upper-class/anti-lower-class biases, and pro-Asian/pro-White/anti-Black racial biases. We observed little evidence of consistent interactions between social categories, or of effects differing between student samples (Studies 1-3) and a representative US sample (Study 4), or as a function of presenting targets as full-body or upper-body photographs (Studies 3 & 4). Taken together, these results suggest that implicit evaluations of multiply categorizable targets may operate according to a category dominance hierarchy, with a single category (here, gender) predominantly driving evaluations, but ancillary categories producing compounding levels of bias toward individuals displaying multiple stigmatized or positively-valued social identities.