Abstract
56 male and 56 female undergraduates played a modified Prisoner's Dilemma game, two teams playing on each occasion. Ss chose a competitive or cooperative strategy toward both their opponents and partners on 20 trials. Partners received more cooperative choices (p < .01). Ratings on ability, sociometric, and motive traits improved for both ingroup and outgroup, however, ingroup ratings tended to improve more. The ingroup-outgroup differential was significantly stronger for game-relevant motive traits than for any other type of trait. The results support a “selective bias” hypothesis, which implies that bias in intergroup perceptions is especially great in the case of motives aroused by intergroup contact. A projective mechanism which defends against the perception of reprehensible motives in the ingroup might account for this selective bias.