The treatment of relevant and irrelevant outgroups in minimal group situations with crossed categorizations

Among a sample of Belgian high school boys, the author attempted to determine on the basis of social identity theory (H. Tajfel & J. C. Turner, 1979) whether an outgroup's relevancy in intergroup comparisons influences the amount of intergroup discrimination. Relevancy was manipulated by the use of 2 trivial categorization dimensions that were orthogonally crossed. In 1 of the crossed categorization conditions, this procedure resulted in the formation of 4 groups, whereas in the 2nd condition, only 2 of these groups were effectively formed. Nevertheless, the participants in both conditions were instructed to evaluate 4 groups--either 4 actual groups (1st condition) or 2 actual and 2 hypothetical groups (2nd condition). In both conditions, the intergroup evaluations exhibited the same pattern--that is, they did not differentiate their ingroups from the partly overlapping groups but were clearly biased against the double outgroup.