The asymmetric bias: race-based lightness perception fades only for female faces, not for male.

Previous research has shown that race and gender information can influence the perceived lightness of faces, as social categorization may introduce biases in perception and memory. However, individuals cannot easily process one facial attribute in isolation. To address this, this study systematically examined the joint effects of race and gender on perceived face lightness across two experiments: the first employed a simultaneous lightness matching task, while the second utilized a delayed lightness matching task. The results revealed a critical gender-based asymmetry. For male faces, the classic race-based lightness bias (with Caucasian faces perceived as lightest, African faces as darkest, and Asian faces as intermediate) was observed and remained stable across both perceptual and memory stages. In contrast, the influence of race on female faces was more complex: a racial lightness bias (with Caucasian faces perceived as lightest, African faces as intermediate, and Asian faces as darkest) was present during initial perceptual processing, but this bias was not maintained in the subsequent memory-based task. Regarding participant gender effects, faces were generally perceived as lighter by male participants than by female participants during perceptual processing. In the memory-based task, however, these participant gender differences in perceived lightness were no longer observed. These results provide the novel insight that social category influences are not uniform but are gated by category intersectionality (race ✕ gender) and processing stage, revealing a uniquely dynamic perceptual status for female faces.