Multilevel performance probability: a meta-analytic integration of expectancy and self-efficacy.

ABSTRACT

A meta-analysis of one approach to measuring expectancy and self-efficacy was conducted. Although used for over 25 years, this measure has yet to be named or integrated across the two theoretical domains. We proposed to label this measure a Multilevel Performance Probability and conducted a meta-analysis. The search for empirical tests of expectancy and self-efficacy using this procedure yielded 16 studies with ratings which could be subjected to meta-analysis. Five studies with 8 tests were taken from expectancy studies and 11 studies with 47 tests from studies of self-efficacy. In total, the analyses involved 7,444 subjects across 55 tests of the Multilevel Performance Probability to performance relationship. Examination of the measure as a predictor of performance gave a mean r of .51 (p<.001) which is in the same direction and larger than values from other meta-analyses conducted within each of the two theoretical domains (rs of .21 and .38).