Using Discrepancies to Predict the Perceived Quality of Work Life

• Published in: Journal of business and psychology 1991-09, Vol.6 (1), p.39-55
• Main Authors: Rice, Robert W., Peirce, Robert S., Moyer, Reed P., McFarlin, Dean B.
• Format: Article
• Language: English
• Subjects: College students; Commuting; Correlations; Job satisfaction; Modeling; Parsimony; Predictive modeling; Questionnaires; Smiles; T tests
• ISSN: 0889-3268; 1573-353X
• DOI: 10.1007/BF01013684

Two job satisfaction surveys tested the hypothesis that discrepancy models provide stronger predictions of satisfaction than do more parsimonious models relying solely on a single component of discrepancy scores. Questionnaire data from samples of 180 working college students and 125 human service workers focused on specific facets of the job, such as: pay, promotion opportunity, and hours of work. Wanted amount of specific job facets was the standard of comparison considered in both surveys. The questionnaires provided measures of facet satisfaction, facet amount, wanted amount, the perceived discrepancy between facet amount and wanted amount as well as the calculated discrepancy between facet amount and wanted amount. Results from 60 tests of differences between correlations provided mixed support. Discrepancy scores were consistently stronger predictors of satisfaction than were wanted amount scores; 24 of 30 tests provided significant support and only 2 of 30 tests provided significant differences in the direction opposite to the hypothesis. Discrepancy scores were not, however, consistently stronger predictors of satisfaction than were facet amount scores; only 8 of 30 tests provided significant support and 8 of 30 test provided significant differences in the direction opposite to the hypothesis. Discussion focused on the importance of addressing these findings in efforts to improve discrepancy theories of satisfaction.