Goals, Strategy Development, and Task Performance: Some Limits on the Efficacy of Goal Setting

Specific, difficult goals enhance performance in many tasks. We hypothesize, however, that this effect disappears or reverses for novel tasks that allow multiple alternative strategies. We report findings from three laboratory experiments using a stock market prediction task with these characteristics. In the first study, 34 students made predictions concerning the value of 100 companies' stock based on three manipulated cues after receiving either a "do your best" or a specific, difficult goal (come within $10 of the actual stock price) concerning the accuracy of their predictions. In the second study, 88 students making stock market predictions received one of the following goals: do your best, specific-easy (come within $30), specific-moderate (come within $20), specific-hard (come within $10), or a tapering, specific goal (decreasing from $30 to $10 in $5 increments every 20 predictions). Finally, the third study (n = 30) replicated the first study by using a different prediction algorithm for the stock market simulation. The results of repeated measures multivariate analyses of variance conducted on indexes of prediction accuracy and predictor weightings supported the hypothesis that specific, difficult goals (prediction accuracy) increase an individual's strategy search activity and reduce prediction accuracy for the stock predictions.