Dissociating stimulus-set and response-set in the context of task-set switching

Abstract

The primary aim of the present research was to determine how stimulus-set and response-set components of task-set contribute to switch costs and conflict processing. Three experiments are described wherein participants completed an explicitly cued task-switching procedure. Experiment 1 established that task switches requiring a reconfiguration of both stimulus- and response-set incurred larger residual switch costs than task switches requiring the reconfiguration of stimulus-set alone. Between-task interference was also drastically reduced for response-set conflict compared with stimulus-set conflict. A second experiment replicated these findings and demonstrated that stimulus- and response-conflict have dissociable effects on the "decision time" and "motor time" components of total response time. Finally, a third experiment replicated Experiment 2 and demonstrated that the stimulus- and response- components of task switching and conflict processing elicit dissociable neural activity as evidence by event-related brain potentials.

Figure 1

Schematic depiction of (A) a conventional task switching procedure and (B) a procedure used by Kieffaber and Hetrick (2005) to isolate stimulus-set switches. The boxes indicate target stimuli. Moving from left to right, the first and second set of arrows are intended to depict the internal structure of stimulus-set, selectively gating information that corresponds with the current task-relevant decisions. The third set of arrows depict the internal structure of response-set, selectively activating task and decision-appropriate responses.

Figure 2

Schematic depiction of the current task. Again, arrows are intended to depict cognitive representations of attentional selection, decision-making, and response-mapping. Open symbols = red, closed symbols = blue.

Figure 3

Schematic depiction of the four conflict types that are possible in the context of the current task. Note that the relevant task is “Shape” and that solid arrows depict only the correct routes for information processing. Open symbols = red, closed symbols = blue.

Figure 4

Procedural schematic of Experiment 1.

Figure 5

Accuracy (A) and response times (B) across conflict types for each of the three switch conditions.

Figure 6

Task switching costs across levels of congruency (bars depict standard error of the mean).

Figure 7

Procedural schematic of Experiment 2.

Figure 8

Experiment 2 response times shown separately for each of the three switch and four congruency conditions and for each of the DT and MT components (error bars reflect ± 1 SE).

Figure 9

Experiment 3 response times shown separately for each of the three switch and four congruency conditions and for each of the DT and MT components (error bars reflect ± 1 SE).

Figure 10

A, Grand-average, cue-locked ERPs from select midline channels. Highlighted regions correspond to the time courses of the N1 (N1_C) and LPC (LPC_C) clusters. B, Topographical distribution of the mean effect size (η²) across samples for each of the electrodes included in the cluster (“halo” extending beyond head boundary represents electrodes on the lateral surfaces of the scalp).

Figure 11

Mean (± 1 SE) cluster-level amplitudes averaged across time and electrodes for each cluster.

Figure 12

A, Target-locked grand-average ERPs from select midline channels. The time course of each cluster is highlighted in gray. B, Topographical distribution of the mean effect size (η²) across samples for each of the electrodes included in the cluster (“halo” extending beyond head boundary represents electrodes on the lateral surfaces of the scalp).

Figure 13

Mean (± 1 SE) cluster amplitudes for each Cluster across each condition.