Is Risky Shift Due to Disproportionate Influence by Extreme Group Members

The extremity-influence explanation of risky and cautious shifts is examined in two ways. First, mathematical models based on the assumption that extreme initial responders have more influence than moderates are derived from congruity theory. Their success at predicting group decisions is compared with the simple averaging of initial responses. Second, influence is defined in terms of talking time and correlated with initial extremity. Both analyses fail to support the hypothesis that extreme responders to an item exert greater influence than less extreme responders.