Threat and the group attribution error: When threat elicits judgments of extremity and homogeneity

In two studies, the authors investigated the impact of threat on the percep tion of extremity and homogeneity among the members of a group. They relied on the Group Attribution Error paradigm and asked participants to estimate the attitudes in a group of voters. Depending on the condition the decisio n rules were such that the proposition allegedly passed or failed. The degr ee of threat of the group was varied by informing participants that the gro up represented a small (4%) or substantial (40%) proportion of the populati on living in the area. Consistent with the Group Attribution Error literatu re, the outcome of the vote influenced participants' inferences about the e xtremity of the voters' attitude. More positive (negative) attitudes were i nferred in the case Of a positive (negative) outcome of the vote. This effe ct was qualified by the level of threat. Participants inferred relatively m ore extreme and more homogeneous attitudes among the members of a threateni ng group.