Abstract
Group effects on individual inferences were studied to determine whether risky shifts in individual judgments would follow group decisions. Ss were asked to make likelihood ratio estimations in a non-cumulative condition and cumulative odds estimates in another condition. Ss were either in individual followed by group conditions or the reverse. Both natural and ad hoc 5-person groups were used. The normative model provided by Bayes' theorem was used to examine groups with effects on individual judgments. Natural and ad hoc groups did not differ. Noncumulative likelihood-ratio groups were veridical compared with the normative model. Although cumulative odds group were conservative in their judgments, no systematic shifts in a risky or conservative direction were found for either task. The group had an anchoring or conformity effect on later judgments of individuals, but this effect was not in the direction of veridicality.
Get full access to this article
View all access options for this article.
