This study provides an experimental investigation of the consequences of conflict between children's personal identities and experimentally manipulated group identities. Elementary-school-aged children (N = 82, ages 5-11) attending a summer school program rated their own academic and athletic abilities and were then randomly assigned to one of two novel groups. Children's views of the academic and athletic skills of the novel groups were assessed both before and after information about the groups' academic and athletic skills was manipulated via posters placed in their classrooms. Following the manipulation, children's self-views, ingroup identification, and intergroup attitudes were assessed. Results indicated that (a) in the absence of information about the novel groups, children projected their personal identities onto their ingroup identities, (b) children maintained their ingroup identities in the face of new information that should have altered their ingroup identities, and (c) more positive personal identities predicted ingroup bias, which in turn predicted happiness with one's ingroup membership. The latter finding suggests that a tendency for children to generalize from their idiosyncratic positive self-views, rather than an indiscriminate desire for self-enhancement or positivity, may be responsible for ingroup bias. Copyright © 2010 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]