Modeling electrophysiological measures of decision‐making and performance monitoring in neurotypical children engaging in a speeded flanker task

Abstract

This study aims to use structural equation modeling (SEM) to investigate the role of error processing in behavioral adaptation in children by testing relationships between error‐related and stimulus‐related event‐related potentials (ERPs) obtained from two sessions of a speeded Eriksen flanker task. First, path models of averaged ERP components and mean response times (N1 → P2 → N2 → P3 → RTs) while controlling for trait effects, age, and sex, on each was examined separately for correct and incorrect trials from each session. While the model demonstrated acceptable fit statistics, the four models yielded diverse results. Next, path models for correct and incorrect trials were tested using latent variables defined by factoring together respective measures of ERP component amplitudes from each session. Comparison of correct and incorrect models revealed significant differences in the relationships between the successive measures of neural processing after controlling for trait effects. Moreover, latent variable models controlling for both trait and session‐specific state variables yielded excellent model fit while models without session‐specific state variables did not. In the final model, the error‐related neural activity (i.e., the ERN and Pe) from incorrect trials was found to significantly relate to the stream of neural processes contributing to trials with the correct behavior. Importantly, the relationship between RT and error detection in the final model signifies a brain‐and‐behavior feedback loop. These findings provided empirical evidence that supports the adaptive orienting theory of error processing by demonstrating how the neural signals of error processing influence behavioral adaptations that facilitate correct behavioral performance.

The impact of punishment on error‐related brain activity in children

“…It is possible that the Flankers task may have been too difficult and thus less engaging for young children, as previous studies with children around 7-11 years old have tended to use Go/No-go tasks, which appear to have a lower proportion of cases excluded due to noise or poor data quality (e.g., 3% data loss;. However, one recent study using a standard Flankers task among children ages 8-12 yielded a slightly lower loss of data across two EEG sessions (approximately 16%;Lin et al, 2021), so it is plausible that the loss of data in the current study may also be related to the addition of the punishment condition. Future studies should assess the potential utility of optimizing the Flankers or adapting Go/No-go tasks with punishment manipulations for use with young children.…”

Developmental trends of auditory novelty oddball P3 while accounting for N2 in 7‐ to 25‐year‐olds

“…Furthermore, young children have a greater negativity in their mean N2 B2P amplitudes than adults, also consistent with previous findings of larger N2 amplitudes in younger children (Kihara et al, 2010). As postulated in the previously mentioned modeling papers (Lin et al, 2022; Taylor et al, 2019), N2 represents target discrimination and P3 represents cognitive evaluation. Therefore, the finding in the current study that young children have a greater negativity in their mean N2 B2P may reflect increased attentional allocation required to process and discriminate infrequent stimuli.…”

“…While much of the research examining the oddball and novelty oddball task has focused on the P3, measuring a single ERP component does not account for the influence of prior components. Previous studies using child data to model ERP components as sequential phases of neural processing have demonstrated that earlier ERP components predict later components, for example, N2 predicts P3 (Lin et al, 2022; Taylor et al, 2019). The subtraction of N2 from P3 to produce a peak‐to‐peak measure of P3 amplitude can capture the inherent relationship between the two components.…”

Development of auditory change‐detection and attentional capture, and their relation to inhibitory control

“…Finally, although we report the relations between EEG measures, the current study examined each of their relations with inhibitory control mostly independent from each other. We encourage future studies to investigate the relations between EEG measures in a multivariate model to better understand the temporal relations between measures (e.g., Lin et al, 2022) or by examining profiles of children with unique patterns across EEG measures.…”