How the timing of performance feedback impacts individual performance

Using a multi-step task setting where learning can help improve individual task performance, I experimentally examine the effect of the timing of performance feedback in an initial period on future task performance when this feedback is absent. I find an inverted-U relation between the timing of feedback and future performance. When feedback is provided before implementation of an initial decision, high learning costs discourage individuals from learning in the initial period to the detriment of future performance. Further, when feedback is provided after extended delays beyond implementation of a decision, learning costs increase relative to those present when feedback is provided after a short delay, resulting in lower learning and future performance. As such, I find that providing feedback immediately following implementation of a decision most effectively promotes learning and future performance as this is the point at which learning costs are lowest. My study extends prior research on feedback timing by incorporating the notion that learning costs fluctuate throughout the phases of a multi-step task and offers practical implications for designing performance evaluation and feedback systems.