Switch costs have been constantly found asymmetrical when switching between two tasks of unequal dominance. We used a combined Stroop-task-switching paradigm and recorded electroencephalographic (EEG) signals to explore the neural mechanism underlying the phenomenon of asymmetrical switch costs. The results revealed that a fronto-central N2 component demonstrated greater negativity in word switch (cW) trials relative to word repeat (wW) trials, and both First P3 and P3b components over the parieto-central region exhibited greater positivity in color switch (wC) trials relative to color repeat (cC) trials, whereas a contrasting switch-related fronto-central SP effect was found to have an opposite pattern for each task. Moreover, the time-frequency analysis showed a right-frontal lower alpha band (9-11 Hz) modulation in the word task, whereas a fronto-central upper alpha band (11-13 Hz) modulation was exclusively found in the color task. These results provide evidence for dissociable neural processes, which are related to inhibitory control and endogenous control, contributing to the generation of asymmetrical switch costs.

Post-error slowing (PES) reflects efficient outcome monitoring, manifested as slower reaction time after errors. Cognitive control account assumes that PES depends on error information, whereas orienting account posits that it depends on error frequency. This raises the question how the outcome valence and outcome frequency separably influence the generation of PES. To address this issue, we varied the probability of observation errors (50/50 and 20/80, correct/error) the “partner” committed by employing an observation-execution task and investigated the corresponding behavioral and neural effects. On each trial, participants first viewed the outcome of a flanker-run that was supposedly performed by a ‘partner’, and then performed a flanker-run themselves afterwards. We observed PES in the two error rate conditions. However, electroencephalographic data suggested error-related potentials (oERN and oPe) and rhythmic oscillation associated with attentional process (alpha band) were respectively sensitive to outcome valence and outcome frequency. Importantly, oERN amplitude was positively correlated with PES. Taken together, these findings support the assumption of the cognitive control account, suggesting that outcome valence and outcome frequency are both involved in PES. Moreover, the generation of PES is indexed by oERN, whereas the modulation of PES size could be reflected on the alpha band.

The prefrontal cortex is responsible for emotional conflict resolution, and this control mechanism is affected by the emotional valence of distracting stimuli. In the present study, we investigated effects of negative and positive stimuli on emotional conflict control using a face-word Stroop task in combination with functional brain imaging. Emotional conflict was absent in the negative face context, in accordance with the null activation observed in areas regarding emotional face processing (fusiform face area, middle temporal/occipital gyrus). Importantly, these visual areas negatively coupled with the dorsolateral prefrontal cortex (DLPFC). However, the significant emotional conflict was observed in the positive face context, this effect was accompanied by activation in areas associated with emotional face processing, and the default mode network (DMN), here, DLPFC mainly negatively coupled with DMN, rather than visual areas. These results suggested that the conflict control mechanism exerted differently between negative faces and positive faces, it implemented more efficiently in the negative face condition, whereas it is more devoted to inhibiting internal interference in the positive face condition. This study thus provides a plausible mechanism of emotional conflict resolution that the rapid pathway for negative emotion processing efficiently triggers control mechanisms to preventively resolve emotional conflict.

The biased competition model proposes that there is top-down directing of attention to a stimulus matching the contents of working memory (WM), even when the maintenance of a WM representation is detrimental to target relevant performance. Despite many studies elucidating that spatial WM guidance can be present early in the visual processing system, whether visual WM guidance also influences perceptual selection remains poorly understood. Here, we investigated the electrophysiological correlates of early guidance of attention by WM in humans. Participants were required to perform a visual search task while concurrently maintaining object representations in their visual WM. Behavioral results showed that response times (RTs) were longer when the distractor in the visual search task was held in WM. The earliest WM guidance effect was observed in the P1 component (90–130 ms), with match trials eliciting larger P1 amplitude than mismatch trials. A similar result was also found in the N1 component (160–200 ms). These P1 and N1 effects could not be attributed to bottom-up perceptual priming from the presentation of a memory cue, because there was no significant difference in early event-related potential (ERP) component when the cue was merely perceptually identified but not actively held in WM. Standardized Low Resolution Electrical Tomography Analysis (sLORETA) showed that the early WM guidance occurred in the occipital lobe and the N1-related activation occurred in the parietal gyrus. Time-frequency data suggested that alpha-band event-related spectral perturbation (ERSP) magnitudes increased under the match condition compared with the mismatch condition only when the cue was held in WM. In conclusion, the present study suggests that the reappearance of a stimulus held in WM enhanced activity in the occipital area. Subsequently, this initial capture of attention by WM could be inhibited by competing visual inputs through attention re-orientation, reflecting by the alpha-band rhythm.

The perceptual load theory in selective attention literature proposes that the interference from task-irrelevant distractor is eliminated when perceptual capacity is fully consumed by task-relevant information. However, the biased competition model suggests that the contents of working memory (WM) can guide attentional selection automatically, even when this guidance is detrimental to visual search. An intriguing but unsolved question is what will happen when selective attention is influenced by both perceptual load and WM guidance. To study this issue, behavioral performances and event-related potentials (ERPs) were recorded when participants were presented with a cue to either identify or hold in memory and had to perform a visual search task subsequently, under conditions of low or high perceptual load. Behavioural data showed that high perceptual load eliminated the attentional capture by WM. The ERP results revealed an obvious WM guidance effect in P1 component with invalid trials eliciting larger P1 than neutral trials, regardless of the level of perceptual load. The interaction between perceptual load and WM guidance was significant for the posterior N1 component. The memory guidance effect on N1 was eliminated by high perceptual load. Standardized Low Resolution Electrical Tomography Analysis (sLORETA) showed that the WM guidance effect and the perceptual load effect on attention can be localized into the occipital area and parietal lobe, respectively. Merely identifying the cue produced no effect on the P1 or N1 component. These results suggest that in selective attention, the information held in WM could capture attention at the early stage of visual processing in the occipital cortex. Interestingly, this initial capture of attention by WM could be modulated by the level of perceptual load and the parietal lobe mediates target selection at the discrimination stage.