Task Switching in Predictable and Unpredictable Cases

Abstract

Fourteen experiments have been run in order to provide evidence regarding the cognitive processes that underlie the switching between simple cognitive tasks. Central to these experiments was the predictability factor; in half of the cases, the upcoming task could be predicted in advance with absolute certainty while in the other half no foreknowledge regarding the upcoming task was provided. In all of the experiments, switch costs were found to be smaller when no task foreknowledge was provided relative to when task foreknowledge was available. Chapter 2 provided evidence regarding the interplay of endogenous and exogenous control in task switching. Top-down and bottom-up processes are not completely insulated from one another. Chapter 3 revealed that both task difficulty and task expectancy play a central role in determining performance on unpredictable cases. Based on the results so far, a task switching model was developed and discussed. Chapter 4 concentrated on the effects of task similarity on performance. It seems that in some cases when tasks are similar at a conceptual level then this results to interference increasing switch costs. Finally, on Chapter 5 behavioral and neuroimaging data provided further evidence that expectancy (in the form of trial expectancy) has a central role on task switching performance. In addition, the neuroimaging data revealed brain regions that could be linked with central components of the proposed task switching model. Concluding, in contrast to many task switching approaches, evidence is provided in the thesis in favor of the presence of endogenous control on unpredictable cases. This control, in the form of expectancies regarding the upcoming task or trial, plays a central role on task switching performance.