On Using Residual Difference Scores in the Measurement of Congruence:
The Case of Met Expectations Research
P. Gregory Irving and John P. Meyer
Despite meta-analytic support for the met expectations hypothesis, Irving and Meyer (1994, 1995) suggested that methodological problems such as the use of difference scores and retrospective measures of met expectations have resulted in an overstatement of this support. In a recent article, Hom, Griffeth, Palich, and Bracker (1998) simultaneously tested several potential psychological mediating mechanisms of realistic job preview (RJP) effects. These authors suggested that met expectations is a critical mediating mechanism, having direct effects on job satisfaction and indirect effects on organizational commitment, withdrawal cognitions, and actual turnover through job satisfaction and other mediating mechanisms such as coping efficacy and perceived employer honesty. However, they used ``residual gain scores" to measure met expectations. In this article, we demonstrate that the use of residual scores for the purposes of operationalizing met expectations creates the same problems as does the use of difference scores, a technique that has been widely criticized in the literature.