The Event-Related Potentials Responding to Outcome Valence and Expectancy Violation during Feedback Processing

Area of Science:

• Neuroscience
• Cognitive Science
• Decision Making

Background:

• Feedback-related negativity (FRN) is a neural signal often linked to reward prediction error (RPE).
• Debate exists whether FRN reflects unsigned prediction error (UPE) or signed prediction error (RPE).
• Limited research compares FRN across appetitive and aversive domains or examines parametric effects of prediction error (PE).

Purpose of the Study:

• To investigate how outcome valence and parametric prediction error influence FRN.
• To determine if FRN encodes UPE or RPE under different experimental conditions.
• To clarify the role of valence and expectancy violation in FRN generation.

Main Methods:

• Utilized a probabilistic learning task with varied prediction and feedback magnitudes.
• Conducted experiments in both valence (gain/loss) and non-valence (digits) contexts.
• Manipulated emphasis on outcome valence versus predictive accuracy in a third experiment.

Main Results:

• Experiments 1 and 2 indicated that FRN encodes UPE.
• Experiment 3 showed FRN encodes RPE when valence is emphasized and UPE when predictive accuracy is highlighted.
• FRN's sensitivity to both outcome valence and expectancy violation was confirmed.

Conclusions:

• FRN's coding is flexible and context-dependent, not solely RPE or UPE.
• Neural responses can shift based on task demands and emphasized information dimensions.
• This highlights the brain's adaptive mechanisms in processing feedback signals.