Are losses more effective than rewards in improving performance in a cognitive task?

It is widely assumed that individuals are more driven to avoid losses than to gain rewards. Yet, this assumption has not been explicitly tested in commonly used cognitive tasks. Here, we compared the influence of both incentives on performance in the Stroop task, in which fast and correct responses resulted in increased rewards or reduced losses. No-incentive trials served as control condition. If losses indeed have a higher subjective weight than equivalent rewards, performance should be improved most by the prospect of loss reduction. Although both incentives clearly improved performance as compared with no-incentive trials across three experiments, we observed no significant advantage of loss compared with reward trials. Rather, our data points in the opposite direction, with overall significantly slower responses in loss compared with reward trials, which may reflect a conflict between an inherent avoidance tendency and the required immediate response. Overall, these findings suggest that losses are not more, but slightly less effective than rewards in improving performance in this cognitive task. Since our data does not support the notion of an asymmetry under which avoided losses are more effective than equivalent rewards to improve performance in a prototypical cognitive control task, both incentives should be fully matched to provide a fair comparison in such task contexts.