Abstract This article reports results of a field study which tested the descriptive validity of expectancy theory predictions about job performance and satisfaction in a public accounting environment. The correlations between overall job satisfaction and predicted performance were interpreted as indicating that the four expectance models were significant predictors of job satisfaction for the audit staff accountants. Results obtained were sufficiently dissimilar to those obtained in other settings so that the ability of the expectancy models to predict job performance levels in an accounting context was questioned. The study used a between subjects research design for testing expectancy theory predictions of work behavior. This design, in which expectancy valence scores are correlated with concurrent individuals, has rather convenient statistical properties and has been extensively employed. A within-subjects research design will be required in order to properly test the descriptive validity of expectancy.
W. Bruce Johnson (Sun,) studied this question.