Group influences on probability judgments concerning social and political change

Abstract

This study was intended to inquire into the generality of group-induced extremization (shift away from the neutral, middle point of a given response scale). The stimulus task for subjects was to estimate the probability that certain desired social, political, or economic conditions would exist by the year 2000. Another purpose of this study was to test the generality of the discussion-induced shift toward pessimism found in a previous study. Using nine discussion groups (size 3 or 4) and 13 control subjects (simple test-retest), the following findings were obtained. ( 1) Group means shifted toward higher extremity after discussion. (2) There was no shift when the extremity of group members' individual positions were compared. (3) The control subjects shifted toward less extreme judgments, and their judgmental confidence decreased. (4) No group shift toward pessimism was obtained. The discussion centers on subsidiary (correlational) evidence suggesting one cause of this group extremization was the disproportionate influence of extreme group members.