Measurement of negative affectivity: A comparison of self-reports and observer ratings

Both self-report and non-self-report measures (co-worker ratings) of negative affectivity (NA) were obtained. Based on previous research, it was hypothesized that these two measures of NA would be moderately correlated. It was also expected that self-report measure of NA would be more strongly correlated with self-reported stressors and strains than would the co-worker measure. Finally, it was predicted that statistically controlling for self-reported NA would attenuate stressor-strain correlations more than would controlling for co-worker ratings of NA. A sample of 121 full-time employees who were pursuing graduate degrees completed self-report measures of stressors, strains, and negative affectivity. These individuals also asked two co-workers to assess their level of negative affectivity. Results supported all three hypotheses. Specifically, there was modest convergence between self-reported and co-worker NA. In addition, self-reported NA was more strongly correlated with self-reported stressors and strains than was the co-worker NA measure. Finally, statistically controlling for self-reported NA attenuated stressor-strain correlations much more than controlling for co-worker NA. These results suggest that observer ratings may be of some use in future NA research. Other implications of these findings are discussed.