Proceedings of the Annual Meeting of the Cognitive Science Society

When deciding how to allocate cognitive control to a giventask, people must consider both positive outcomes (e.g.,praise) and negative outcomes (e.g., admonishment). How-ever, it is unclear how these two forms of incentives differen-tially influence the amount and type of cognitive control a per-son chooses to allocate. To address this question, we had par-ticipants perform a self-paced incentivized cognitive controltask, varying the magnitude of reward for a correct responseand punishment for an incorrect response. Formalizing controlallocation as a process of adjusting parameters of a drift diffu-sion model (DDM), we show that participants engaged in dif-ferent strategies in response to variability in reward (adjustingdrift rate) versus punishment (adjusting response threshold).We demonstrate that this divergent set of strategies is optimalfor maximizing reward rate while minimizing effort costs. Fi-nally, we show that these dissociable patterns of behavior en-able us to estimate the motivational salience of positive versusnegative incentives for a given individual.