This week, Roland Imhoff visits from the University of Mainz to give a talk as part of the SWE Colloquium.

Using data-driven approaches to rethink the definitions of stereotypes and prejudices

Textbook definitions of intergroup attitudes typically distinguish between stereotypes and prejudices. The term ‘stereotypes’ denotes beliefs about how individuals are based on certain category memberships (X is shy because X is a librarian), thus an association between category and an attribute (not necessarily implying any valence). This is markedly different for prejudices that are often conceptualized akin to the affective aspect of attitudes. The term ‘prejudices’ is reserved for individuals’ affective reaction to people because of their category membership, i.e., an association between category and valence. Despite this relatively clear-cut definition, I will argue, the current state of the literature is muddled. Partially responsible for this confusion is the prominent Stereotype Content Model and its assertion that the primary dimension on which social groups are stereotypes is warmth. Inspired by a data-driven alternative stereotype content model (ABC) I will argue that warmth impressions are better conceptualized as prejudices than stereotypes as (a) warmth is almost synonymous with valence and (b) unlike other stereotype dimensions (e.g., agency, beliefs) warmth judgments are not consensual, but idiosyncratic. Reconciling the SCM and the ABC model, I will show that warmth (or communion) is of utmost importance to understand people’s emotional reactions and concrete behavior towards groups, but that to detect this, warmth has to be treated as idiosyncratic rather than socially shared. This has important implications not only for our theoretical understanding of category-based attitudes but also for concrete research practice.