Abstract
The present study used confirmatory factor analysis to evaluate a series of models representing the relations among performance ratings from multiple sources across multiple performance dimensions. The results suggest that the variance in performance ratings can be attributed to three distinct sources. First, there is common variance attributable to performance dimensions or factor loadings. Second, there is variance that can be attributed to the rating source (peer, self, or supervisor). Third, there is unique variance beyond that explained by the performance dimensions and source effects. Further, these results suggest that the rating sources appear to be rating the same performance dimensions, and even tapping the performance constructs to the same extent in some cases. The source effects do account for some variance, but these effects are smaller than the performance dimensions. Additionally, the unique variance accounted for a large component of the variance associated with the performance ratings in the present study.
Sheehan, Mary Kathleen (1999). An assessment of the measurement equivalence of rating sources in a multisource feedback system. Master's thesis, Texas A&M University. Available electronically from
https : / /hdl .handle .net /1969 .1 /ETD -TAMU -1999 -THESIS -S539.