Executive control of cognitive processes in task switching.

Abstract

Current theories of frontal lobe function suggest the presence of a central executive system which coordinates complex novel behaviors, and controls the activation of rules in unfamiliar tasks. Seven experiments investigated the operation of the frontal central executive by examining how rule complexity and goal memory affect task-switching performance. In all experiments, each subject completed a pure block of one task (A) and a pure block of a second task (B). Task switching was studied by having subjects complete a mixed block in which they alternated between two tasks (i.e., A, B, A, etc.). Goal memory limitations were studied by manipulating the presence of task cues on each trial. Three experiments used stimuli and tasks similar in design to the Wisconsin Card Sorting Test. These experiments had subjects sort geometric patterns based on shape, number, shading, and size. An easy rule condition required sorting based on single-dimension rules (e.g., sort according to number). A hard rule condition required sorting based on conjunctive-dimension rules (e.g., sort according to both number and shape). Four additional experiments had subjects solve arithmetic problems. Addition and subtraction were the two easy rule conditions, while multiplication and division were the two hard rule conditions. Previous researchers have claimed that task difficulty does not affect switching time. This claim was shown to be only partly correct. The present experiments demonstrated that switching time depends on rule complexity. Subjects took longer to switch between tasks with hard rules than between tasks with easy rules. The present experiments also showed that task cues affect switching time. Goal memory limitations are reduced when subjects can see cues reminding them of the current task. The effects of rule complexity and task cues do not interact, suggesting that these two factors affect task switching independently. Additional analyses revealed that task switching is independent of other forms of task difficulty. Double dissociations between task response times and switching times imply that the switching process is independent of the task decision process. The results of these experiments reveal key properties of the central executive control of task switching.