Endogene und exogene Kontrolle im Aufgabenwechselparadigma

Abstract

Switching between two tasks in randomized order leads to costs both in response times and errors when compared to a condition where task is held constant. These global switch costs (Mayr, 2001) are even observable in trials where tasks are locally repeated. A possible explanation for this phenomenon could be that under switching conditions subjects use an endogenous control strategy to counteract interference from exogenous control or to ensure relatively fast switching of task sets. Both assumptions imply that this strategy results from the randomized execution of multiple tasks within one block. Results from four experiments showed that it is not the changing execution of tasks but rather the changing indication by explicit cues that triggers such a strategy. Invalid cues were used to construct blocks in which both tasks needed to be executed although only one task is being cued. Global switch costs, as measured by the comparison of blocks with constant and randomized cues, were found even under these invalid cuing conditions. The costs varied merely with the frequency of task cues although this type of frequency did not relate to the frequency of task execution. It was thus concluded that the strategy was not triggered by the interference itself but rather by the expectation of this interference. These and other results (e. g. local switch costs, costs of invalid cues) were integrated in a model describing the interaction of endogenous and exogenous control in the task switching paradigm.