Learned predictiveness and group perception: An associative analysis of biases in stereotype formation

Stereotypes are beliefs about traits that are shared by members of a social group. Categorisation on the basis of group membership provides cognitive economy, condensing our experience of many group members into a general impression of that group. But generalisation comes at a price if it causes us to overlook an individual's unique characteristics and prejudge them on the basis of group membership. Application of negative stereotypes in particular (eg "members of group X are lazy") has damaging effects on society, generating fear, anger and resentment towards certain groups. It is thus important to understand how stereotypes form, and by extension how they might be unlearned. To the extent that stereotypes form on the basis of our experience with group members, they can be analysed using models of associative learning developed on the basis of studies of animal conditioning. Formation of a stereotype can be modelled as formation of an association between a group label (group X) and a trait (lazy), with the strength of the stereotype corresponding to the strength of this association. An associative view of stereotypes allows us to bring a wide body of research on animal and human associative learning (along with formal models derived from this research) to bear on the question of how stereotypes are formed. In particular, it suggests ways in which the stereotype formation process might be biased, causing us to develop irrational and objectively incorrect views of social groups. This idea forms the focus of the proposed research project.