Randomization tests to determine significance levels for microanalytic congruences between self-efficacy and behavior

Abstract

Self-efficacy research has employed a "microanalytic congruence" measure to assess relationships between self-judged efficacy and behavior. This measure assesses the percentage of tasks for which self-efficacy and behavior are congruent. Though obtained congruence valued have often been quite large, they have been reported without significance levels indicating the probability of obtaining these values by chance. This is especially problematic in that the congruence expected by chance may be considerably greater than 50%. The present paper critically analyzes previous approaches to estimating chance congruence levels. A randomization test is presented for obtaining significance levels for the microanalytic congruence measure, irrespective of the form of the self-efficacy scales. This method, applied to data reported in the literature, shows the obtained congruences to be highly significant.