Relative performance feedback: Effective or dismaying?

In this experiment, I analyze whether the provision of relative performance feedback differently affects the performance of subjects when provided in various feedback environments. Subjects were ranked either relative to the performance of many subjects in the past or relative to three subjects working alongside themselves. Results indicate that the response from subjects in the former varies with how they perceived their own ability to solve the task. Those reporting low ability reduce their performance when provided with the feedback, whereas those reporting high ability improve. For subjects who were ranked relative to others working alongside themselves, no one respond negatively, but only those reporting high ability improve their performance. An important implication from this, especially for managers who design feedback policies in organizations, is that the way relative feedback is designed may lead to different behavioral reactions. In particular, the choice of benchmark used to relatively rank employees may result in responses that are not beneficial and lead to inefficient use of resources.