Two surveys were conducted to explore the effects of incumbents’ task performance on their job analysis ratings, using four job analysis scales (importance and level rating scales of generalized work activities and skills) from Occupational Information Network (O*NET). We examined these while controlling for such demographic variables as gender, age, tenure and education. In the first study, skill importance, level ratings, and self-rated performance were obtained from 38 power plant designers in one organization. The results of hierarchical regression analyses showed that, after controlling for the individual demographic variables, task performance still had significant impacts on the level ratings of technical skills. However, the partial correlation coefficient between task performance and technical skill importance ratings became marginally significant after controlling for the demographic variables. The second study involved 88 book editors from one publishing company, with task performance ratings collected from their direct supervisors. Hierarchical regression analyses showed that, after controlling for the demographic variables, editor’s task performance had significant effects on both importance and level ratings of information processing activities.
These two studies extended existing research on job analysis ratings of identical jobs in several ways. First, we examined the effect of job performance on job analysis ratings with individual demographic variables controlled for. The resulting partial correlations from the first study were different or moderately different from the zero-order correlations without partialling out the demographic variations. Second, following the suggestion by Lindel et al. (1998) and Van Iddekinge et al. (2005) that organizational level variables may affect job analysis ratings of the same job in different organizations, we explored the influence of task performance on job analysis ratings of one job in one organization. Therefore, in each study, the potential effects of possible organization-level variables on job analysis ratings were controlled for. In many ways, our analyses ensured a relatively stringent evaluation of the effects of job performance on job analysis ratings. Third, the findings indicated that task performance influenced job analysis ratings of many scales, including level ratings and importance ratings. Fourth, consistent with Borman et al (1992), the present results suggest that differences in job analysis ratings may reflect real differences, either among tasks assigned to different job incumbents under identical job titles or differences in ways by which job incumbents complete the same task. One practical implication is that, when conducting job analysis in organizations, practitioners need to consider the potential influence of task performance on incumbents’ job analysis ratings as well as individual demographic variables.
