These are data from the primary dependent measures from my ESRC grant examining how intergoup bias influences the formation of novel lab-based stereotypes via a process of cumulative cultural evolution. The data were collected from undergraduate participants who were tested individually. Each participant was a single 'generation' within a 'diffusion chain' of four generations. Participants were asked to try and learn some information about some novel social targets. The responses each participant produced during a test phase were used as the learning materials for the next participant (i.e., generation). The dependent measures of interest were the accuracy with which people completed the task and the amount of structure that was present in their responses. We were interested in whether a stereotype-like categorical structure would develop as information was passed down the chains and whether this would result in associated increases in accuracy.The proposed research will establish how the membership and status of social groups influences how cultural stereotypes form and change. Cultural stereotypes are template-like depictions of social categories whereby group membership is associated with the possession of certain attributes (e.g., scientists are geeky, Scottish people are miserly, men like the colour blue). Stereotypes exert substantial influence on us as individuals and on our society: when people endorse stereotypes it leads to prejudice and discrimination towards members of minority groups; even when people refute stereotypes the mere knowledge of their content can lead to unconscious bias in thoughts and behaviour. Yet, in the face of an infinitely complex social environment stereotypes play a vital social cognitive role by efficiently organising and structuring social information. Given their ubiquity and influence it is perhaps surprising that relatively little is known about how cultural stereotypes form and change. We propose that stereotypes form and change via a process of cumulative cultural evolution. Because people possess shared biases that influence how information is remembered and communicated, when knowledge is repeatedly passed from person to person these biases accumulate causing the content of information to change in predictable ways. Research has shown that when information is passed down chains of individuals - a bit like the children's game often called 'Chinese whispers' or 'telephone' - it becomes increasingly simplified and structured. For example, we recently demonstrated that as novel social information passes from person to person it develops a stereotype-like structure that was not previously present. Thus, through the process of cumulative cultural evolution, even very small amounts of bias at the level of individual people can translate into much bigger societal biases like cultural stereotypes. The proposed research will establish whether individual biases associated with a person's membership of social groups influences the formation and evolution of cultural stereotypes. Whether we perceive others as belonging to the same social group as ourselves (the in-group) or a different social group (the out-group) has profound implications for our thoughts and behaviours. Group membership tends to lead to intergroup bias, with people more likely to favour in-group members and discriminate against out-group members. The proposed research will determine whether repeatedly communicating social information about in-group and out-group members results in the formation of relatively positive in-group stereotypes and negative out-group stereotypes. In addition, the proposed research will also establish whether it is possible to predict how the content of stereotypes will evolve based on the perceived status of different out-groups (e.g., whether they are perceived to be high status or low status). The proposed research will therefore help establish whether cumulative cultural evolution leads to the unintentional but inevitable formation of stereotypes, whose content is largely determined by the shared biases of perceivers rather than the actual properties of the groups themselves.

Lab-based experimental psychology using a diffusion chain methodology. The data were collected on personal computers. A sample of undergraduate students was recruited to take part in the experiments.