On how to be unpredictable: evidence from the voluntary task-switching paradigm

The voluntary task-switching paradigm requires subjects to select randomly between tasks and promises to provide a window into executive task selection independent of exogenous influences present in standard task-switching situations. We show here that the degree to which subjects perseverate on tasks across trials captures unique individual differences variance, but also that the switch rate is under strong stimulus-driven control: "Voluntary" switches are much more frequent when the stimulus changes than when it repeats. Most important, we show that individuals whose no-switch trials are selectively slowed exhibit less perseveration and stimulus-driven effects (and thus more voluntary selection) than individuals whose no-switch trials do not show this slowing. We suggest that selective slowing indicates a strategy of treating trials as discrete events--possibly through inhibition of the preceding task set. These results not only demonstrate massive nonvoluntary influences on voluntary selection that are largely untapped by standard task-switching measures, but also show how such influences can be counteracted through strategic adaptations.