Faculty Evaluations by Medicine Residents Using Known versus Anonymous Systems

Abstract

PURPOSE This study examined the extent to which faculty evaluation results differed, based on whether residents were required to submit ratings anonymously or not. METHOD We used a retrospective analysis of existing records representing Internal Medicine residents' evaluation of 51 faculty members in an anonymous and known (non-anonymous) rater system on an inpatient medicine service. RESULTS Mean scores for 48 of 51 individuals were lower for anonymous than non-anonymous evaluations. The mean scores were as follows: Anonymous = 5.4 (95 % CI, 5.2-5.6); Nonanonymous = 6.1 (95 % CI, 5.9-6.3). Regression analysis of mean scores for non-anonymous evaluations against those for anonymous evaluations revealed a significant relationship (r = 0.83, p < 0.001). CONCLUSIONS Faculty evaluations completed anonymously by residents are significantly lower than those for which resident identities were known. Given the strong, significant relationship between individual faculty members' evaluation ratings from both systems, other factors influencing evaluations should be considered. No subgroups suffered more under anonymous rankings.