Abstract
Background Social anxiety symptoms are most likely to emerge during adolescence, a developmental window marked by heightened concern over peer evaluation. However, the neurocognitive mechanism(s) underlying adolescent social anxiety remain unclear. Emerging work points to the error-related negativity (ERN) as a potential neural marker of exaggerated self/error-monitoring in social anxiety, particularly for errors committed in front of peers.However, social anxiety symptoms are marked by heterogeneity and it remains unclear exactly what domain(s) of social anxiety symptoms are associated with ERN variation in peer presence, particularly within the adolescent period.
Methods To advance and deepen the mechanistic understanding of the ERN’s putative role as a neural marker for social anxiety in adolescence, we leveraged a social manipulation procedure and assessed a developmentally-salient domain of social anxiety during adolescence—Fear of Negative Evaluation (FNE). Adolescents residing in Hanzhong, a small city in the southwestern region of mainland China, had EEG recorded while performing a flanker task, twice (peer presence/absence); FNE, as well as global social anxiety symptoms were assessed.
Results Overall ERN increases in peer presence. FNE specifically, but not global levels of social anxiety symptoms, predicted ERN in peer presence.
Conclusions These data are the first demonstration that the ERN relates to a specific domain of social anxiety in adolescents, as well as the first evidence of such relations within a non-WEIRD (Western, Educated, Industrialized, Rich and Democratic) sample. Results have important implications for theory and research into adolescent social anxiety.
Competing Interest Statement
The authors have declared no competing interest.
Footnotes
Major changes: 1. we now employ the ERP Reliability Toolbox (Clayson & Miller, 2017) to determine the minimum number of trials necessary to achieve adequate reliability and reflect results in the updated manuscript; 2. We have now shifted away from a traditional statistical approach and employ a mixed effects framework throughout the revised manuscript. Reasons are below: within the context of traditional statistical approaches (e.g., those employed in our original manuscript), dealing with unequal trial counts across conditions involves weighing trade-offs. On the one hand, there is the possible imbalance in signal-to-noise ratios across conditions (although for mean amplitude analyses, see: Luck, 2014). On the other hand, matching trial counts via random selection of a subset of trials removes useful data (and variance), with results further dependent on the random seed. Thus, to address unequal trial counts, we have now shifted away from a traditional statistical approach and employ a mixed effects framework throughout the revised manuscript. Specifically, all ERP (and behavioral) data are now modelled within a mixed effects framework at the single trial level to account for variation in trial counts (see Heise et al., 2022 for a complete justification of this approach as the preferred method for dealing with unequal trial counts in ERP research). All changes are reflected in pages 12-16 of the revised methods and results section to view these changes.