Erroneous and correct actions have a different affective valence: Evidence from ERPs.

Abstract

The accuracy of actions is swiftly determined through specific monitoring brain systems. Event-related potential (ERP) studies have shown that error commission is associated with the generation of the error-related negativity (ERN/Ne), while correct actions with the correct-related negativity (CRN). Although the exact functional meaning of the ERN/Ne (and CRN) component remains debated, some authors have suggested that it reflects the processing of the emotional significance of actions. However, no study to date has directly linked amplitude changes at the level of the ERN/Ne-CRN to the affective processing of actions. In order to decode the emotional valence of actions performed during a Go/noGo task, we used an evaluative priming method in this study. After each action following the Go/noGo stimulus, participants had to categorize an evaluative word as either positive or negative. Behavioral results showed that response errors (i.e., False Alarms -FAs) performed during the Go/noGo task led to a faster categorization of negative than positive words.Remarkably, this evaluative priming effect was related to the magnitude of the ERN/Ne component generated during the Go/noGo task. Moreover, ERP results showed that the processing of evaluative words following FAs was influenced early on after word onset (early posterior negativity-EPN effect), while it was influenced later following correct as well as incorrect actions (late positive potential-LPP effect). Altogether, these ERP results suggest that the action-related ERN-CRN component encodes the perceived emotional significance of actions, such that early stages of evaluative word processing following these actions are influenced by this automatic process.3

Evaluative priming reveals dissociable effects of cognitive versus physiological anxiety on action monitoring.

“…To this aim, in Experiment 2, we induced worry by means of an anagram task, while worry was treated as a trait variable in Experiment 1. The results of these two experiments replicate the findings previously put forward by Aarts et al (2012Aarts et al ( , 2013. More specifically, in these two experiments, we found evidence for a significant interaction effect at the group level between the putative affective value of the action (correct/positive vs. incorrect/negative) and the valence of the subsequent emotional word (positive vs. negative).…”

“…Accuracy and RTs (for correct responses) were analyzed separately using a repeated measures ANOVA with (i) the valence of the target word (either positive or negative) and (ii) the type of action (False Alarms and Fast Hits) preceding word presentation as within-subject factors. We did not include in these analyses trials corresponding to Slow Hits because this action type did not lead to any significant and consistent evaluative priming effect in previous studies (Aarts et al, 2012(Aarts et al, , 2013, nor in the present one, t < 1.This may be due to the fact that the putative valence of slow hits (unlike either Fast Hits or False Alarms) was somehow ambivalent in the sense that a slow hit was a correct action in absolute terms but performed too slowly relative to the arbitrary response deadline (hence probably carrying also a negative connotation). Even though the dichotomy between Fast Hits and Slow Hits could appear somewhat arbitrary at first sight, earlier studies using the same speeded Go/noGo task already showed that these two trial types elicited different neurophysiological effects (at the CRN level specifically) and could be dissociated from one another (see Vocat et al, 2008;Aarts et al, 2013).…”

“…We did not include in these analyses trials corresponding to Slow Hits because this action type did not lead to any significant and consistent evaluative priming effect in previous studies (Aarts et al, 2012(Aarts et al, , 2013, nor in the present one, t < 1.This may be due to the fact that the putative valence of slow hits (unlike either Fast Hits or False Alarms) was somehow ambivalent in the sense that a slow hit was a correct action in absolute terms but performed too slowly relative to the arbitrary response deadline (hence probably carrying also a negative connotation). Even though the dichotomy between Fast Hits and Slow Hits could appear somewhat arbitrary at first sight, earlier studies using the same speeded Go/noGo task already showed that these two trial types elicited different neurophysiological effects (at the CRN level specifically) and could be dissociated from one another (see Vocat et al, 2008;Aarts et al, 2013). Furthermore, auxiliary analyses (not reported here) performed on the RT distributions for the three main action types (False Alarms, Fast Hits and Slow Hits) revealed only partial overlap between Fast Hits and Slow Hits, suggesting that these two trial types actually corresponded to different behavioral outcomes in terms of RT speed during the speeded Go/noGo task (see Pourtois, 2011for a similar analysis and conclusion).…”

“…Likewise, worry and arousal were also entered as main (and concurrent) predictors in the regression model used for the evaluative priming effect obtained for emotional faces. (Aarts et al, 2012(Aarts et al, , 2013.…”

The cognitive mechanisms underlying deception: An event-related potential study

“…Related to this explanation is a second possibility. Aside from error and conflict monitoring, the ERN has also been suggested to be sensitive to the emotional valence of errors (Aarts et al, 2013;Luu et al, 2000). Within this line of reasoning, smaller CRNs would correspond to a more positive evaluation of an action.…”