Cognitive Multitasking: Inhibition in Task Switching Depends on Stimulus Complexity.

Stefanie Schuch,Otmar Bock,Klara Freitag,Luca Moretti

doi:10.5334/joc.115

Abstract

We report a series of three experiments investigating inhibition in task switching, using N-2 repetition costs as an empirical marker. The experiments were structurally identical, employing a standard experimental paradigm where participants switch between three different categorization tasks. The experiments differed with respect to the stimulus material. According to prominent theories of cognitive control, N-2 repetition costs should be observed in all three experiments. To our surprise, this is not what we observed: N-2 repetition costs did not occur in Experiment 1, where we used static pictures from a driving simulator environment showing an oncoming car, embedded in a car-driving scene. In contrast, we observed robust N-2 repetition costs in Experiment 2, where we used static pictures of faces, and in Experiment 3, where the identical car stimuli from Experiment 1 were used, but without the surrounding visual scene. These results suggest that N-2 repetition costs depend on the complexity of the stimulus material. We discuss two aspects of complexity: 1) When the relevant stimulus feature is embedded in a complex visual scene, task-irrelevant features in that scene might trigger additional task sets, and thus induce additional task switches, attenuating N-2 repetition costs among the instructed task sets. 2) The presence of distractors might lead to additional covert or overt shifts of spatial attention, which in turn might reduce the size of N-2 repetition costs. On a more general level, the results illustrate the difficulty of transferring laboratory tasks to settings that bear more similarity to everyday life situations.

Highlights

In cognitive psychology, paradigms have been developed to investigate cognitive processes in constrained, minimalized laboratory settings, in order to isolate the cognitive processes in question
The absence of N-2 repetition costs is surprising, given that the structure of Experiment 1 was very similar to other studies in the cognitive psychology literature, where robust N-2 repetition costs were observed: Participants switched between three different categorization tasks on a trial-by-trial basis; the stimuli were multivalent, such that the relevant categorization task cannot be determined on the basis of the presented stimulus, but needs to be inferred from the task cue presented prior to the stimulus; responses were multivalent, such that the same set of response keys is used for the different tasks
One may argue that two aspects of Experiment 1 differ from typical task-switching and N-2 repetition cost paradigms: First, the task cue disappeared upon stimulus onset in Experiment 1, whereas in many taskswitching and N-2 repetition cost paradigms, it remains present during stimulus presentation, especially when visual task cues are used

Summary

Introduction

Paradigms have been developed to investigate cognitive processes in constrained, minimalized laboratory settings, in order to isolate the cognitive processes in question. It is often assumed that findings from laboratory studies can serve to explain human behavior in everyday life situations outside the lab. This assumption is not always valid: Laboratory and daily life settings differ in a variety of aspects (e.g., Bock & Hagemann, 2010; Bock et al, 2019), and empirical effects observed in a laboratory context cannot always be transferred to contexts outside the lab. We systematically manipulated aspects of stimulus material and stimulus presentation, in order to determine possible causes for these discrepancies

Methods

Results

Conclusion