“Oops, I Did it Again?” Behavioral and Electrophysiological Indices of Error Processing Predict Adaptive Post-Error Compensations

Paul J. Beatty

Advisor: Craig McDonald, PhD, Department of Psychology

Committee Members: James Thompson, Martin Wiener

Online Location,
March 10, 2021, 11:00 AM to 01:00 PM

Abstract:

Empirical research suggests that when an error is committed during forced-choice decision-making tasks, a network of brain regions including the anterior cingulate cortex (ACC), pre-supplementary motor area (preSMA), and dorsolateral prefrontal cortex (dlPFC), will enable us to pause from the ongoing task, attend to and identify the error, and help to prevent us from making the mistake again once we re-engage in the task. Although researchers often examine the relation between electrophysiological indices of error processing and the magnitude of response time (RT) slowing following error commission (post-error slowing; PES), the utility of PES remains a subject of debate. Are participants slower because they respond more cautiously and consequently bring their performance back up to acceptable levels? Or are participants slower because they were distracted by the error and consequently continue to perform sub-optimally when re-engaging in the task? The present dissertation presents two experiments, which provide support for adaptive post-error adjustments and argue for the inclusion of a time-frequency approach to better understand the role of medial frontal cortex in action monitoring. Experiment 1 demonstrates that evoked neural responses (an event related potential component known as the error related negativity; ERN) and brain waves (i.e. rhythmic oscillatory activity in the theta frequency range) that take place at the time of error commission, can be dissociated in terms of their ability to predict post-error behavior, as well as their sensitivity to response conflict associated with an erroneous response. More specifically, induced theta power appears to be more closely related to adaptive post-error behavior than the ERN. Experiment 2 replicates these findings as well as demonstrates that the tendency for participants to immediately correct for their errors predicts adaptive post-error behavior. Even when a secondary, corrective response is not explicitly made following an error, the amount of motor activation underlying the subthreshold error correction, indexed by lateralized suppression of rhythmic oscillatory activity in the beta frequency range, can predict post-error behavioral adjustments. More specifically, we observed that greater activation for subthreshold corrections predicted post-error speeding and increased post-error accuracy at the single trial level. This finding provides novel evidence that the response conflict associated with subthreshold error corrections facilitates adaptive post-error adjustments.