Familiarity breeds overconfidence: Group membership and shared experience in the closeness-communication bias

Though both spouses and strangers understand a speaker equally well in a novel communication task, people overestimate their success of communication more with spouses than strangers (Savitsky, Keysar, Epley, Carter, & Swanson, 2011). This phenomenon, termed the closeness-communication bias, is presumed to occur because people anchor their common ground assessments egocentrically and only minimally adjust when they perceive their interlocutors to be similar to themselves. In four experiments on college undergraduates, we investigated whether (1) friendship, (2) college affiliation, (3) minimal groups, and (4) familiarity (as defined as shared experience) affected the closeness-communication bias. As predicted, participants predicted greater prosody comprehension for friends than strangers (ηp2 = 0.56), as did minimal group members (ηp2 = 0.220) and individuals who engaged in a short online chat (ηp2 = 0.116). However, we did not find evidence of the bias when comparing undergraduates speaking to another student from the same college vs. speaking to a student from a rival college (ηp2 = 0.04). Thus, people are more likely to overestimate common ground both when they have prior personal experience with their interlocutors or when they have minimal knowledge about their interlocutors' group characteristics. In addition, the amount of personal experience or prior knowledge influences the magnitude of the bias.