We hypothesize that conflicts with group opinion evoke feedback-related negativity (FRN) a component often associated with a performance monitoring and reinforcement learning prediction error signal and its magnetic counterpart mFRN. We also suggest that conflicts with normative group opinion induce an increase in theta-band oscillatory activity, which is associated with negative behavioral outcomes . Moreover in the present study we demostrate that the magnitude of both signals could predict individual proneness to social influence.
20 female subjects rated the trustworthiness of female faces. At the end of each trial subjects were informed about an ‘average group rating’ of the face given by a large group of students from the same Russian university (feedback). Actual group ratings were systematically manipulated during the experiment. Feedback-locked ERFs and time-frequency maps were calculated separately for the conflict trials (i,e, the subject’s ratings mismatched the group rating) and for the no-conflict trials (Figure 1) . A separate analysis was performed for the trials that led to change of subjects’ ratings towards the group rating (conformity) vs trials where conflict with group opinion didn’t lead to conformity (no conformity).
In line with the social influence hypothesis, participants changed their ratings of trustworthiness to align themselves with the group ratings
We have also found a significant cluster of activation in the posterior medial prefrontal cortex (pMPFC) at 204-240ms time window in conflict vs no conflict condition.
ERS/ERD analysis performed on a cluster of fronto-central gradiometers revealed significantly higher theta power increase (against -400 to -200 pre stimulus baseline) for conflict than for no-conflict trials.
The increase in fronto-central theta power following conflict trials was more pronounced as compared with no-conflict trials.
The study has demonstrated that a mismatch between individual and group opinions triggered both evoked and induced theta-band activity in the posterior medial frontal cortex similar to FRN, implicating the involvement of general performance-monitoring during the social influence.
In general, our results support the hypothesis that forms of social influence are mediated by activity of the posterior medial frontal cortex as a part of the general performance-monitoring circuitry.