A LONGITUDINAL-STUDY OF UPWARD FEEDBACK

This study followed 92 managers during four administrations of an upwa rd feedback program over 2.5 years. Managers' whose initial level of p erformance (defined as the average rating from subordinates) was low i mproved between Administrations 1 and 2, and sustained this improvemen t 2 years later. Most of the performance improvement occurred between the first and second administrations of the program. The observed perf ormance improvement could not be attributed solely to regression to th e mean and was unrelated to the number of times, or when, the managers received feedback. Our results suggest that the continued administrat ion of an upward feedback program can result in sustained change over a fairly long period of time and that actually receiving feedback may be less important than the exposure to the valued behaviors. We use co ntrol theory and goal setting theory as frameworks for explaining the results and suggesting directions for future research.