Abstract
A sample of 133 employees from a large oil company was used to investigate the theory that job satisfacion is a function of the discrepancy between what a person wants from a job and what the person gets from the job. Five job attributes—skill‐utilization, influence, variety, pressure, and interaction—were used. The hypothesis that the discrepancy between the desired and perceived levels of the job attributes would be a better predictor of job satisfaction, if weighted by the importance of the job attribute, received only slight support. Attempts to overcome methodological problems associated with previous tests of the hypothesis also had little effect on the ability to predict job satisfaction. Overall, the best predictor of job satisfaction was the perceived level of the job attributes, especially skill‐utilization. The subjective ratings of the importance of job attributes were found to differ from an empirical determination of importance. The results suggest a need to further investigate work values and to reconsider their relationship to job satisfaction.