Abstract
The minimal group paradigm has consistently shown that people will discriminate to favor their own group over an out-group, even when both groups are created arbitrarily by an experimenter. But will people actually form groups that are so arbitrary? And could something as trivial as a randomly assigned name tag color serve as a fault line during group formation? In this study, we use in vivo behavioral tracking (IBT) to precisely and unobtrusively track samples of participants as they assort repeatedly into groups. We find that participants do form groups on the basis of their randomly assigned name tag colors, but that name tag homophily emerges over time, becoming stronger in subsequent groups. Our results suggest that people are unconsciously or consciously biased toward group similarity, even when similarities are essentially meaningless. Our study has implications for theories of intergroup relations and social identity. It also demonstrates the utility of applying real-time tracking to study group formation.