Abstract
Individuals adapt their readiness to shift spatial attention or switch tasks, referred to as cognitive flexibility, according to changing environmental demands (Dreisbach & Haider, 2006; Sali et al., 2020). However, the domain-specificity of flexibility learning and the resulting implications for attentional capture are unknown. If cognitive flexibility learning is domain-specific, high switch-readiness should be unrelated to distractibility. Alternatively, learned task-switching flexibility may be associated with an increase in the likelihood of attentional capture, reflecting domain-independent learning. Across two experiments, we manipulated the likelihood of task switches across consecutive trials within visual search paradigms to test whether learned flexibility is associated with increased distractibility. In Experiment 1, participants completed an additional singleton paradigm, searching for the unique shape in each array, and in Experiment 2, participants completed a feature search paradigm, searching for a shape-defined target. In both experiments, a color singleton was present in half of the trials and participants periodically switched between two categorization rules. Experiment 2 also included rare probe trials in which participants reported letters that had briefly flashed at each location (as in Gaspelin et al., 2015) to test whether task-switching flexibility weakened singleton distractor suppression. In both studies, behavioral task-switching costs were smallest in blocks of trials with frequent trial-by-trial task-switching, indicating learned adjustments in cognitive flexibility. While participants were overall slower on distractor present trials than distractor absent trials in Experiment 1, there was no difference in the magnitude of capture according to the likelihood of task-switching. In Experiment 2, response times did not differ between distractor present and distractor absent trials in the feature search task and we observed no difference in probe recall at the singleton distractor location according to switch likelihood. Together, our results suggest that flexibility learning in the domain of task-switching does not carry consequences for moment-by-moment distractibility.