Unique motor plans facilitate learning during task switching, but at the expense of greater switch costs

Many factors influence task switching, including attention, motor resources, and the different types of memory systems required for each component task. Even so, prior task-switching studies have focused almost exclusively on tasks that do not require significant learning. An experiment is reported that investigates how attention, motor planning, and memory systems influence the learning of two categorization tasks when those tasks are: (1) completely novel to participants, and (2) must be switched between on a random trial-by-trial basis during the learning process. Results indicate that associating each category label with a unique motor response facilitates the learning of each task during task switching. Participants were more accurate on switch trials with unique motor responses, but at the cost of longer response times. A novel cognitive control model is proposed that successfully accounts for these results. Finally, these results are discussed within the context of existing category-learning and cognitive-control theories, none of which predict the observed pattern of results.