Assessing the automaticity of intergroup bias

Intergroup bias is a pervasive feature of human societies, underlying many forms of prejudice, stereotyping, and discrimination (Schneider, 2003). However, its developmental origins have not been well-established. The most prominent account predicts the gradual decline of intergroup bias with increasing cognitive maturity (Aboud, 1988), but this prediction is hard to reconcile with the widespread presence of bias in adults. In this thesis, I contrast predictions made by social learning theories with those made by 'automaticity' theories, which instead suggest that intergroup bias may emerge automatically as soon as ingroup/outgroup categorizations are made.;If this latter account is correct, then it would predict that: (a) intergroup bias should be present from the earliest moments of social categorization, (b) intergroup bias will appear early across a wide range of human populations, irrespective of the social status of the ingroup or the outgroup, and (c) intergroup bias should be relatively easy to induce in the laboratory, by placing children or adults in artificial social groups. In this thesis, I present experiments designed to test each of these three predictions.;To test the first prediction, I utilize a new method of attitude assessment, which allows the measurement of race attitudes in children as young as age 3. I found that children of this age showed adult-like ingroup favoring race bias, but only if they could successfully perform adult-like racial classifications. This finding reinforces the notion that categorization and evaluation are tightly linked.;To test the second prediction, I investigated intergroup bias in a US minority population. Minority populations present a particularly strong test of the automaticity hypothesis because ingroup status and the wider social value of the group are in direct conflict. However, contrary to the mere membership prediction, minority children did not exhibit ingroup preference.;To test the third prediction, I conducted a minimal-group experiment in which I randomly assigned five year-old children to groups about which they had no prior knowledge. Supporting the automaticity hypothesis, they exhibited intergroup bias on multiple measures. These results suggest that intergroup bias does emerge automatically following ingroup/outgroup categorization, at least in members of the majority.