Last time, I reviewed a set of experiments that analyzed the effect of minimal groups on racial biases. This time, I investigate a follow up experiment to this research, and briefly elaborate on how it has influenced my research.
In an extension of their previous research, Jay Van Bavel and William Cunningham took advantage of a common finding in categorization research: individuals are more likely to remember faces of those who share their race, rather than other-race individuals. They arbitrarily assigned white participants to one of two mixed race groups. Participants were then shown pictures of members from both groups, and then, importantly, asked questions about how much they identified with their group. In a final task, they were shown faces from both of the groups along with new faces and asked whether they had seen each face before.
When examining memory accuracy, in general participants had better memory for members of their own group (the ingroup), regardless of member race. They also demonstrated generally poor memory for members of the other group (the outgroup). However, this effect was moderated by level of identification with their ingroup. For participants who valued their membership in the ingroup, recall for members of the outgroup was poor, but for individuals who did not value their ingroup membership, recall for both groups was equal.
Interestingly, this effect was also moderated by ingroup role. When participants were told that they were to act as a “spy” on the outgroup, they recalled outgroup members as accurately as ingroup members, even when their commitment to the ingroup was high.
The authors propose that these manipulations are both manipulations of “social identity,” that is, not only the knowledge that an individual belongs to a certain group, but the value of that identity to the individual and the role the individual plays in that group. Accordingly, variations of social identity, such as low levels of identification with the ingroup, can reduce or even remove the benefits prescribed to ingroup members. In these experiments that benefit was limited to recall, but it is likely that other such benefits (such as reductions in race bias towards ingroup members) would disappear as well.
Accordingly, the research I am currently conducting in the Cesario lab examines these results from a motivated cognition perspective. Restructuring these results in motivated terms, ingroup biases result because groups facilitate goal striving. Groups are seen as reliable and helpful entities that assist and individual goal pursuit. We are currently examining other factors that moderate these group biases. One of our goals is to see if by changing the structure and role of different arbitrary groups, we can extend the benefits of group membership (such as a reduction in racial bias) to outgroup members.