Binding of a Task-Irrelevant Context in Voluntary Task Switching: Increasing Context Variability

Previous research provided evidence for feature binding and episodic retrieval processes in task switching, also involving task-irrelevant features (the context henceforth). A binding and retrieval account postulates that repeating the context retrieves the bound features, for example the task and/or the response. In the present study, we examine whether repeating the context retrieves the previous task (suggesting that the two are bound) by measuring the probability that the same task is executed when the context repeats. We employed a voluntary task-switching paradigm, wherein participants are free to pick the task to perform in each trial among two possible tasks. To ensure a sufficient number of task switches, we employed a procedure devised by Mittelstädt et al. (2018). Specifically, in each trial, we delay the onset of the task-repetition stimulus, with a delay that increases the more consecutive task repetitions are performed. Since the more active a task set is, the more likely it is selected, we predict that repeating the context will increase n-1 voluntary task repetitions. In this experiment, we used eight different colorful and complex pictures as context. Compared to previous experiments where we used only two contexts and simpler (i.e., red or blue boxes), in this experiment we expect a larger effect of repeating the context due to fewer bindings being formed with each context instance and to a greater saliency when repeating a picture out of eight possible pictures.