A comparison of clinical teaching evaluations by resident and peer physicians

The purpose of this study was to compare the reliability of inpatient teaching evaluations by resident and peer physicians on Mayo internal medicine hospital services. Three resident and three peer evaluators observed 10 consecutively chosen attending physicians on the Mayo hospital services. Evaluations by resident and peer physicians were compared in terms of mean scores. Kendall's coefficient of concordance (KCC) was used to summarize inter-rater reliabilities and Cronbach's coefficient alpha was used to determine internal consistencies of evaluations by residents and peers. Results of this study revealed that mean scores of the 13 evaluation items were generally higher for resident than peer physicians. None of the items completed by residents had KCC scores >0.5, whereas 10 of the items completed by peers had KCC scores >0.5. Likewise, none of the residents' items had KCC p-values <0.05, whereas nine of the peers' items had KCC p-values <0.05. The overall internal consistency was higher for peers (alpha = 0.76) than for residents (alpha = 0.71). In conclusion, resident physicians uniformly rate faculty highly. Furthermore, peer evaluations yield higher inter-rater and internal reliabilities than resident evaluations, indicating that peer physicians are more reliable than residents for assessing bedside teaching.