The purpose of the studies was to investigate the effects of group membership on the evaluations of ingroup and outgroup others, and to explore the potentially mediating effect of a personal trait. The "contextualism" was examined as the personal trait since this was hypothesized to concern the degree to which group membership is integrated into one's self concept-the central issue of the idea of social identity. A preliminary study examined validity and reliability of a scale to tap this individual difference with the previous (n = 3883) and new data (n = 99). In the main study, 46 subjects participated in a minimal group experiment after answering the contextualism scale. The primary dependent measures were evaluative ratings of anonymous people and their writings from ingroup and outgroup. On the whole, evaluative ingroup bias was evident. While the degree of ingroup bias was positively related to the contextualism for person evaluation, no such relation was found for writings evaluation. Mechanisms to produce ingroup bias were discussed.