Predictive neural activity scales with noise but not performance in speech-in-noise

Notice: This is an earlier version with known errors. For use and citation, refer to the corrected version at {doi: 10.5281/zenodo.17373690}. This record contains de-identified derivative data for an EEG study of the sustained posterior negativity (SPN) under noise vs. silence conditions. The package includes:•spn_electrode_long.csv : Trial- or participant-level SPN values by Participants Sex, Condition, and Electrode (µV).•electrode_metadata.csv : Mapping from Electrode to ROI (and optionally Hemisphere).• speech_accuracy.csv : Speech-in-noise accuracy per SNR_dB with n_trials and n_correct.• variable_dictionary.csv : Column definitions and units.• README_DATA.md, LICENSE-DATA.txt. IDs are pseudonymous; dates and direct identifiers are removed. Condition labels are normalized to in noise / in silence . All files are UTF-8 CSV with “.” as decimal separator; SPN is in µV and SNR in dB. Derivatives were generated with analyze_spn.py (GitHub: https://github.com/kokamoto46/analyze_spn), ensuring that figures and statistics reported in the manuscript can be reproduced from these tables. License: CC BY 4.0. Please cite the dataset as:Kazuhiro Okamoto*, Kengo Hoyano, Tomomi Nomura, Keisuke Irie, Naoya Obama, Narihiro Kodama, and Yasutaka Kobayashi (2025). Predictive neural activity scales with noise but not performance in speech-in-noise (v1.0). Zenodo. https://doi.org/10.5281/zenodo.17173946 Raw EEG is not posted due to privacy; access may be available on request under a data-use agreement.