Effects of Training on Decision-Making

Abstract

The experiment was designed to test whether training on one simulated problem would upgrade decisions on an entirely different one. The training problem required groups to contribute three solutions. E then demonstrated a method for appraising solutions in terms of the facts that had been supplied. The test problem required only the making of a decision about a promotion in which a choice between three alternatives was allowed: Encourage, Discourage and Insufficient Information (compromise decision). In Condition I, 222 Ss were given the test problem and worked as individuals; in Condition II, 153 Ss were given the test problem and worked in 3-person groups; and in Condition III, 189 Ss were first given the training problem, working in 5-person groups in the morning, and the test problem in the afternoon, working in 3-person groups as in Condition II. Comparison of Conditions I and II showed the 3-person groups to be more likely to reach the compromise alternative, resulting both in fewer high (Discourage) and low quality decisions. Comparison of Conditions II and III showed a sharp increase in high quality decisions (7.8% to 36.5%), indicating that the training had a significant effect. Since the training and test problems had no overlapping content, the only relevant link between them was the exercise in evaluating solutions in terms of objective criteria and hence avoiding emotional bias based on liking the person or using promotion as a reward for past performance.