Reconfiguration of response-set in task switching: Event-related potential evidence.

In task switching, an interaction between task and response is often observed, with response repetition (RR) benefits in task-repeat trials and RR costs in task-switch trials. The theoretical accounts of the RR effect remain controversial, and neuroscience evidence is scarce. The present study utilized the event-related potentials (ERPs) method to explore the neural mechanism underlying the RR effect by adopting a cued task-switching paradigm. The ERP results revealed the interaction between task and response in the P3b time window, with a response switch positivity under task-repetition conditions and an RR positivity under task-switching conditions. In addition, there were RR positivity in the N2 irrespective of task transition and in the late component (LC, 550-600 ms) that only under the task repetition condition. On the individual level, the RR benefit positively correlated with the RR positivity in the LC, while the RR costs negatively correlated with RR positivity in the N2/P3 component. These results suggest that both response reconfiguration and episodic-retrieval make contributions to the RR effects, which were also discussed in terms of predictive model for a domain-general inference and learning of perceptual categories.

Proactive and reactive control differ between task switching and response rule switching: Event-related potential evidence.

The distinction between task-switching (T-switch) and response-rule switching (RR-switch) has been reported in previous studies. However, it is unclear whether the neural correlates of proactive and reactive control differ between T-switch and RR-switch. In this study, a modified cue-target task was adopted. When the cue in the current trial differed from that in the preceding trial in shape (or color), the participants had to perform a T-switch (or RR-switch). Otherwise, they performed the same task following the same response rule. The behavioral results showed that the switch cost was greater for the RR-switch than for the T-switch. The event-related potential results indicated that (1) for cues, the switch-positivity in the late positive component (LPC) (500-800 ms) was more enhanced for the RR-switch than for the T-switch over the central to parietal regions, reflecting increased proactive control for the RR-switch compared with the T-switch; (2) for targets, the P3 amplitude was more attenuated in the RR-switch than the T-switch over the central and parietal regions, reflecting increased reactive control for the RR-switch; and (3) under the T-switch, the switch-positivity in the cue-LPC was negatively correlated with accuracy cost, while under the RR-switch, the switch negativity in the target-P3 was positively correlated with the reaction time cost. These findings suggest that similar proactive and reactive control are recruited in the T-switch and RR-switch, whereas cognitive control efforts clearly differ between them, perhaps due to different sub-processes.

Neural Dynamics of Cognitive Control in Various Types of Incongruence.

Conflict-control is a core function of cognitive control. Although numerous studies have considered cognitive control to be domain-general, the shared and distinct brain responses to different types of incongruence or conflict remain unclear. Using a hybrid flanker task, the present study explored the temporal dynamics of brain activation to three types of incongruence: flanker interference, rule-based response switch (rule-switch), and action-based response switch (response-alternation). The results showed that: (1) all three types of incongruence evoked larger N2 amplitudes than the congruent condition in the frontal region, with the N2 amplitudes and topographical distribution of the N2 effect differing between the different types of incongruence; and (2) in the P300 time window, the flanker interference condition yielded the most delayed P300 latency, whereas the rule-switch and response-alternation conditions yielded smaller P300 amplitudes with a longer interval from P300 peak to a keypress. These findings suggest that different types of incongruence are first monitored similarly by the cognitive control system and then resolved differently.

Not inertia but reconfiguration: Asymmetric switch cost in a hierarchical task.

Numerous studies have investigated asymmetric switch cost between different types of tasks. However, the underlying neural mechanism of asymmetric switch cost in a hierarchical task remains unknown. In the present study, we used a representation-nested paradigm and defined three hierarchical levels of number comparison. Participants were required to switch between different levels according to the color of the stimuli. Behavioral result showed a longer response time in the upward switch (from the lower level to the middle level) than in the downward switch (from the higher level to the middle level). The event related potential results showed that (1) a larger N2 was elicited by the upward switch, which reflects the response inhibition process; and (2) the sustained potential component was most negative for the upward switch and least negative for the repeat condition, thereby reflecting the different degrees of reconfiguration of task set. The reverse asymmetric switch cost in the hierarchical task was explained using task-set reconfiguration theory and related to the activation of the left frontal region.

More abstract, more difficult to shift: Behavior and electrophysiological evidence.

Previous studies on task or rule shifting have shown that the switch cost is affected by the hierarchical level of abstraction. The present study aimed to explore the neural correlates of hierarchical modulation in rule shifting. Participants were required to respond to number stimuli (1-9, excluding 5) while adhering to alternative rules cued by different perceptual features of the stimuli. Across trials, the rules might repeat or shift at either high or low hierarchical levels. Behavioral results indicated that the reaction times were significantly slower in the high-shift than in the low-shift condition. The electrophysiological results revealed (1) significant difference in P300 between high- and low-shift conditions, with larger P300 in the high-shift condition; and (2) increased amplitude of the late positive component (LPC) elicited by the high-shift condition as compared to that by the low-shift condition. These findings demonstrate that the task-set reconfiguration and rule selection is more difficult for higher hierarchical level of rule shifting as compared with lower level.

Brain responses associated with different hierarchical effects on cues and targets during rule shifting.

Numerous studies have explored the neural mechanisms related to task switching or rule shifting, but few have revealed the neural substrates related to rule shifting at different hierarchical levels. The purpose of this study is to explore the different event-related potentials (ERPs) elicited by higher and lower hierarchical levels of rule shifting in a cue-target task. In the task, participants were presented with Arabic digits (1-9, excluding 5), and performed a parity or magnitude judgment according to the preceding cues, a letter R. The rules were constructed in a hierarchical set, and the rule was either repeated or shifted between neighboring trials. Importantly, the rule shifting would be at a higher or lower hierarchical level. The results showed that, for the cues, the higher hierarchical shifting evoked increased P2, P3, late positive component (LPC), and decreased N2 relative to lower hierarchical shifting, reflecting the increased demand in selective attention, rule processing, and proactive control during the rule acquisition stage. For the targets, the hierarchical effects were reversed, and emerged later than the rule shift effect. These findings imply that, in the cue-target rule-shifting task, the brain responds differently to different hierarchical levels of rule shifting.

Distinct brain responses to different inhibitions: Evidence from a modified Flanker Task.

Whether inhibition is a unitary or multifaceted construct is still an open question. To clarify the electrophysiological distinction among the different types of inhibition, we used a modified flanker paradigm, in which interference inhibition, rule inhibition, and response inhibition were compared to non-inhibition condition. The results indicated that, compared to the non-inhibition condition (1) the interference inhibition condition induced larger negativities during N2 epoch at the frontal region, (2) the rule inhibition condition elicited a larger N1 at the posterior region, followed by a larger P3a at the frontal region, reflecting the function of proactive cognitive control in the new stimulus-reaction (S-R) association, and (3) the response inhibition condition evoked a larger P3b at the posterior region, reflecting the process of suppressing the old response and reprogramming the new action. These findings provide new evidence that distinct neural mechanisms underlie different types of inhibition.

The Effect of Feedback and Operational Experience on Children's Rule Learning.

This study aimed to examine the relative effect of feedback and operational experience on children's rule learning in a balance scale task, in which 88 children under the age of 7 years were asked to judge the state of equilibrium under four conditions. In the Control condition, children were required to observe the scale and predict which side would tilt down or keep balance, without feedback on the correctness of their answer. In the Operation (Op) condition, children were required to place the weights on the scale just like the experimenter did before they made predictions. In the Feedback (Fe) condition, feedback was provided for each prediction, but children were not allowed to operate the scale. In the Op-Fe condition, children could operate the scale and they were provided feedback for each prediction. The results showed that, (1) children in Control condition merely adopted the lowest level of rule, the Weight Rule; (2) when they were either given feedback or the opportunity to operate the scale, they used a higher level rule, such as the Distance Rule, more frequently; and feedback was more effective than the operational experience was in promoting rule learning; (3) when they were allowed to operate the scale, and were simultaneously provided feedback, rule learning increased markedly, suggesting that feedback-based operation is the most efficient method for facilitating children's rule learning.