Feedback Valence, Prediction Error, and the Limits of Computational Attribution in Symptom-Related EEG

Greater trait anxiety is linked to a larger negative-minus-positive FCz theta feedback response in 46 participants with elevated depressive symptoms, after adjusting for depression severity.

• Why it matters: Signed prediction error in binary feedback tasks can be tightly tied to feedback valence, risking misattribution of EEG effects to computational variables rather than valence.
• What they did: The authors reanalyzed two open probabilistic-learning EEG datasets, testing reliability, theta specificity versus delta and P3-minus-N2, and participant-level recovery using exact-design simulations and participant-specific model refits.
• The result: Feedback valence and signed prediction error were highly correlated (r = -0.964), theta specificity was not established, and model discrimination could work at the population level despite weak individual slope recovery.

AI summary of the abstract · may contain errors · not medical advice · read the paper.