Predicting problem solving outcomes from expectancies of outcome, self-efficacy, and compliance

The relationship between problem solving and a variety of expectancies was explored to assess and compare the accuracy of three types of expectancies. Before performing an anagram test, students completed questionnaires to assess their expectancies that (a) a technique would be effective if used (outcome), (b) they would be able to perform the technique (self-efficacy), and (c) they would actually implement the technique (compliance) in solving a problem. Results indicated that for certain groups of problem solving strategies, after experience with the task, outcome and compliance expectancies were significantly related to subsequent performance, but only self-efficacy expectancies were related to both prior and subsequent performance. After extensive experience, self-efficacy and outcome expectancies lost their predictive edge over compliance expectancies while compliance expectancies demonstrated significant ability to predict performance. This suggests that subjects may find it more cognitively economical to converge both outcome and self-efficacy expectancies into a single compliance expectancy. Additionally, certain strategies rated by most subjects as ineffective were rated high in effectiveness by subjects who were later less successful at solving anagrams, creating an inverse outcome expectancy-performance relation. This replicated an earlier finding in Yates, Casey, Schoonover, and Gray (1990).