Conference on Cognitive Computational Neuroscience

Humans procrastinate. Procrastinators usually make little progress at the start (start late) and a significant increase in progress shortly before the deadline (rush to complete). Yet, the cognitive mechanisms underlying this time course of progress remain poorly understood. Here, we show how the timing of rewards, affects procrastination in terms of the time course of progress. We did this by creating a novel experimental paradigm where people work on a self-paced, week-long task consisting of a large number of units of work. We found that offering an immediate reward upon task completion helped people start the task earlier. Moreover, it helps people who generally procrastinate complete the task earlier and complete units of work earlier. A computational model using reinforcement learning theory qualitatively predicts the effect of reward timing on procrastination.