Post-error Slowing Reflects the Joint Impact of Adaptive and Maladaptive Processes During Decision Making

Impact of motor error processing on performance on a working memory task: effect of modulating cognitive load in high and low span groups

Making a motor error, hitting the wrong key for example in responding to a stimulus, impacts perceptual, motor, and cognitive processes by implying a variation in performance and response time on the next trial (Post Error Slowing, Dutilh et al., 2012; Wessel, 2018). Wessel et al. (2022) showed that an error produced on a flanker-type task had a negative impact on a working memory span recognition. This effect is called ERIAM (Error Related Impairment of Active Working Memory) and consists of degradation of memory span following the production of an error on a conflict task concurrent to a memory task. In their study, the ERIAM effect is more pronounced in Low Span individuals. This suggests that these individuals have less stable Working Memory representations and are more likely to be impacted by an error. The presented study replicates the ERIAM protocol of Wessel et al. (2022) by adding variation of cognitive load. In addition, span recognition task is modified with a recall task. Results show a replication of ERIAM effect in our sample. Moreover, ERIAM effect is only significant in lowspan group in the moderate cognitive load condition and in highspan group in the increased load condition. In line with Lavie's load theory (Lavie et al., 2004), we explain these results by the fact that High and Low Span groups do not process cognitive load increase in the same way.

Transcutaneous vagus nerve stimulation boosts accuracy during perceptual decision-making

BACKGROUND

The locus coeruleus-norepinephrine (LC-NE) system is a well-established regulator of behavior, yet its precise role remains unclear. Animal studies predominantly support a "gain" hypothesis, suggesting that the LC-NE system enhances sensory processing. In contrast, human studies have proposed an alternative "urgency" hypothesis, postulating that LC-NE primarily accelerates responses.

METHOD

To address this discrepancy, we administered transcutaneous vagus nerve stimulation (tVNS) in two experiments. In the first experiment (n = 22), we showed that 4-s tVNS trains reliably induced greater pupil dilation compared to SHAM condition, indicating increased LC-NE activity. In the second experiment (n = 21), we applied tVNS during a random dot motion task to assess its impact on perceptual decision-making.

RESULT

tVNS improved accuracy without affecting reaction times, which appears inconsistent with the "urgency" hypothesis. Exploratory drift-diffusion model analyses further support the "gain" hypothesis, revealing that tVNS increased the drift rate, indicative of enhanced evidence accumulation. Both accuracy and drift-rate improvements were most prominent following errors and especially pronounced in participants who exhibited post-error declines in these measures under SHAM.

CONCLUSION

Our findings align with the "gain" hypothesis, with tentative evidence suggesting that the impact of LC-NE activity adapts to task demands. Accordingly, tVNS showed the strongest effects in contexts prone to accuracy declines, possibly reflecting attentional disengagement, which points to a role of LC in mitigating lapses of attention.

Task goals shape the relationship between decision and movement speed

The speed at which we move is linked to the speed at which we decide to make these movements. Yet the principles guiding such relationship remain unclear: whereas some studies point toward a shared invigoration process boosting decision and movement speed jointly, others rather indicate a trade-off between both levels of control, with slower movements accompanying faster decisions. Here, we aimed 1) at further investigating the existence of a shared invigoration process linking decision and movement and 2) at testing the hypothesis that such a link is masked when detrimental to the reward rate. To this aim, we tested 62 subjects who performed the Tokens task in two experiments (separate sessions): experiment 1 evaluated how changing decision speed affects movement speed, whereas experiment 2 assessed how changing movement speed affects decision speed. In the latter experiment, subjects were encouraged to favor either decision speed (fast decision group) or decision accuracy (slow decision group). Various mixed model analyses revealed a coregulation of decision (urgency) and movement speed in experiment 1 and in the fast decision group of experiment 2 but not in the slow decision group, despite the fact that these same subjects displayed a coregulation effect in experiment 1. Altogether, our findings support the idea that coregulation occurs as a default mode but that this form of control is diminished or supplanted by a trade-off relationship, contingent on reward rate maximization. Drawing from these behavioral observations, we propose that multiple processes contribute to shaping the speed of decisions and movements.NEW & NOTEWORTHY The principles guiding the relationship between decision and movement speed are still unclear. In the present behavioral study involving two experiments conducted with 62 human subjects, we report findings indicating a relationship that varies as a function of the task goals. Coregulation emerges as a default mode of control that fades when detrimental to the reward rate, possibly because of the influence of other processes that can selectively shape the speed of our decisions or movements.

Post-error adjustments occur in both reaching and grasping

Post-error slowing (PES), the tendency to slow down a behavioral response after a previous error, has typically been investigated during simple cognitive tasks using response time as a measure of PES magnitude. More recently, PES was investigated during a single reach-to-grasp task to determine where post-error adjustments are employed in a more ecological setting. Kinematic analyses in the previous study detected PES during pre-movement planning and within the grasping component of movement execution. In the current study (N = 22), we increased the cognitive demands of a reach-to-grasp task by adding a choice between target and distractor locations to further explore PES, and other post-error adjustments, under different task conditions. We observed a significant main effect of task condition on overall reaction time (RT); however, it did not significantly impact PES or other post-error adjustments. Nonetheless, the results of this study suggest post-error adjustment is a flexible process that can be observed during pre-movement planning and within the onset and magnitude of the reaching component, as well as in the magnitudes of the grasping component. Considering the sum of the results in the context of existing literature, we conclude that the findings add support to a functional account of error reactivity, such that post-error adjustments are implemented intentionally to improve performance.