A Bayesian Inference Model of Corollary Discharge Dysfunction in Schizophrenia: Quantifying Efference Copy Precision from EEG Data

Read full text Summarise with VIRAT
Authors: Narendra Bayutama Wibisono
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Abstract Reduced suppression of the auditory N100 to self-generated sounds is a candidate marker of impaired corollary discharge in schizophrenia, and motor preparation is hypothesised to supply the predictive signal that drives this suppression. Whether motor preparation scales N100 suppression, and whether that coupling differs between patients and controls, remains unresolved. We reanalysed publicly available button-press EEG data from 81 participants (32 healthy controls, HC; 49 people with schizophrenia, SZ). N100 suppression was defined as the corrected active response (button-tone minus button-alone) minus the passive response at Fz, FCz, and Cz. Group differences were examined with Welch tests and HC3-robust regressions adjusted for education and age. The association between motor preparation (mean C3_B1; secondary analysis: a C3−C4 LRP-style index) and suppression was examined with Bayesian Student-t regressions including a motor preparation × group interaction, prior-sensitivity analyses and posterior predictive checks. Suppression was lower in SZ at all three electrodes (mean differences 1.21–1.54; Hedges’ g = 0.51–0.70; Holm-adjusted p ≤ .034). After adjustment for education, the group coefficients shrank by 29–58% and their 95% confidence intervals included zero (Fz −1.10 [−2.30, 0.11]). In the Bayesian models, every 95% HDI for the interaction included zero (e.g., Cz −0.49 [−1.42, 0.41]), and the sign of the estimated interaction reversed when motor preparation was indexed by the LRP-style measure. Lower N100 suppression in SZ is therefore present in unadjusted analyses but is sensitive to adjustment for education, and these data do not provide credible evidence that the motor preparation–suppression association differs between groups. Conclusions about this association depend on how motor preparation is operationalised. We position the findings as exploratory and outline the data needed to test efference-copy accounts more directly. Keywords: corollary discharge, efference copy, N100 suppression, schizophrenia, Bayesian regression, motor preparation, predictive coding Code and Data Availability Data: The raw EEG dataset analyzed in this preprint is publicly available via Kaggle, deposited by the original authors of the research publication "Did I Do That? Abnormal Predictive Processes in Schizophrenia When Button Pressing to Deliver a Tone" (Ford et al., 2014) at: https://www.kaggle.com/datasets/broach/button-tone-sz Code: The custom Python scripts used to calculate the Bayesian precision weights and ANCOVA residuals for this manuscript are openly available in Zenodo at: https://doi.org/10.5281/zenodo.24082010

Fields of Study & Subjects

Schizophrenia research and treatment Free Will and Agency Neural dynamics and brain function

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
19 March 2026
Language
EN
Document type
Preprint