Date created: | Last Updated:

Category: Project

Description: Recent evidence suggests that our perception of other people is literally, and not just figuratively influenced by knowledge about the social categories to which they belong [1,2]. As a result, even the perceived brightness of a person’s skin is said to be automatically biased by knowledge about their presumed racial group [1]. Here, for the first time, we examine the claim that such biases are automatic: do they operate quickly, are we aware of them, and can they be overcome by higher-level cognitive states? These questions play key roles in both the moral issue of where responsibility for prejudice lies – with the individual, or the society, for instance – and the practical issue of how bias, prejudice, and discrimination can be reduced in the real world. Participants compared the luminance of pairs of faces, and we independently manipulated actual luminance, and the presence of African or European morphological features. We found from participants’ mouse cursor movements that the influence of morphology on judgements arose as early as that of luminance itself did, and from their confidence ratings and response times that they were not aware that their responses were biased. In a second experiment, we told participants about the bias, and asked them to try to avoid it, finding reduced reliance on morphology on all measures, including initial movements. Therefore, while the bias occurred quickly, and apparently without awareness, just making people aware of their bias was sufficient to reduce it. This finding has wide ranging consequences on how we assess people’s responsibility and attempt to reduce the impact of social biases on everyday decisions.