Comparing self- and supervisor evaluations: a different view.

Abstract

Studies comparing self- and supervisor evaluations of performance during medical training have generally indicated a lack of congruence between the two sets of ratings. A possible explanation is that supervisors may make evaluations on the basis of an overall impression whereas self-evaluations are more sensitive to differences between performance in different areas. Support for this explanation was indicated by consistently higher correlations among supervisor ratings of 13 performance areas for first-year residents. Support was also indicated by separate factor analyses of supervisor and self-ratings of 13 areas, in that two factors were identified for supervisors whereas there were three for self-ratings. It was concluded that the data supported the notion that self-ratings were more discriminating than supervisors' in evaluating resident performance across a number of performance areas.