Abstract
Voluntary task switching (VTS) provides a useful paradigm for examining how recent reward outcomes are associated with the balance between cognitive stability and flexibility. We investigated whether reward-history effects on task choice varied as a function of cue–outcome contingency. Participants (N = 616) completed a VTS task with performance-contingent rewards. Reward contingency was manipulated between-subjects at probabilities of100\%, 80\%, 70\%, and 60\%. Switch probability was modeled using logistic mixed-effects regression. The relationship between prior reward outcomes and the current reward prospect in predicting switch behavior differed across probability conditions. Reward transition effects were larger at 100\% and 80\% than at 70\% and 60\%. Secondary analyses further indicated that outcomes that were unexpected given the cue–reward mapping were associated with subsequent switching in some comparisons, although evidence that these associations systematically varied with probability was mixed. Because task choices did not influence reward outcomes, the observed effects are more consistent with carry-over associations between reward history and task selection than with reward-guided exploration or reward maximization. Exploratory individual-differences analyses suggested that reward transition effects were more pronounced in males than in females; additionally, trait-level reward sensitivity moderated the magnitude of reward transition effects on switch behavior. Overall, the findings are consistent with the hypothesis that cue-outcome contingency modulates the extent to which recent rewards are followed by particular task-choice patterns, although further research is required to clarify the mechanisms through which reward uncertainty may contribute to these effects.