Human behavior is more strongly driven by the motivation to avoid losses than to pursue gains (loss aversion). However, there is little research on how losses influence the motivation to exert effort. We compared the effects of loss and gain incentives on cognitive task performance and effort-based decision making. In three experiments, participants performed a cognitively effortful task under gain and loss conditions and made choices about effort expenditure in a decision-making task. Results consistently showed significant loss aversion in effort-based decision making. Participants were willing to invest more effort in the loss compared to the gain condition (i.e., perform a longer duration task: Experiments 1 and 2; or higher task load: Experiment 3). On the other hand, losses did not lead to improved performance (sustained attention), or higher physiological effort (pupil diameter) in Experiments 1 and 2. In Experiment 3, losses did enhance working memory performance, but only at the highest load level. Taken together, these results suggest that loss aversion motivates higher effort investment in effort-based decision-making, while the effect of loss aversion during a performance may depend on the task type or effort level.

No abstract

The motivation to avoid losses is often considered a strong drive of human behavior, affecting decisions in the context of risk, temporal delay, and effort provision. However, studies measuring cognitive performance under loss and gain incentives have yielded mixed findings. In a recent study, we found evidence that losses motivated better working memory performance than gains. Since these results deviated from previous studies, we decided to perform a preregistered, high‐powered, replication, and extension study. Eighty‐six participants performed a working memory task (N‐Back) under Gain and Loss conditions. Pupillometry was recorded during task performance, and subjective effort ratings were taken after each task run. Results showed that contrary to our previous findings, losses did not enhance working memory performance, relative to gains, at any N‐Back level. Similarly, while pupil diameter was parametrically modulated by working memory load, no differences between the Gain and Loss conditions were found. Participants did believe that they were exerting more effort during Loss versus Gain conditions, regardless of N‐Back level. However, this did not translate to performance differences between Gain and Loss conditions.