Dataser for "Effects of irrelevant speech on cognitive and behavioural outcomes in open-plan offices"

This dataset accompanies the study "Effects of irrelevant speech on cognitive and behavioural outcomes in open-plan offices." Twenty-four participants (12 male, 12 female) performed a Sustained Attention to Response Task (SART) under four within-subject conditions of irrelevant speech sound pressure level (SPL): quiet control (35–40 dB(A)), low (40–45 dB(A)), medium (45–50 dB(A)), and high (50–55 dB(A)). Electroencephalogram (EEG) and electrodermal activity (EDA) were recorded simultaneously. After excluding participants with low EEG sampling rates or artefact-contaminated recordings, 16 valid EEG datasets were retained.

The file dataset.xlsx contains four sheets. The aggregated_behavior_eda sheet provides condition-level behavioural and EDA data for all 24 participants, including reaction capacity (Cr), correct rate (Rc), overall performance (P), and the relative change in skin conductance level (dSCL). The time_series_behavior sheet provides the same behavioural metrics binned into 20-second intervals across the 4.3-minute task. The aggregated_eeg sheet provides condition-level EEG task engagement (TE), defined as the ratio of beta power to the sum of alpha and theta power at prefrontal electrodes (AF7, AF8, FP1, FP2), for the 16 valid participants. The time_series_eeg sheet provides the same TE index in 20-second bins. Across all sheets, subject_id is an anonymised participant identifier, spl_level encodes condition (0 = quiet, 1 = low, 2 = medium, 3 = high), gender is coded as 0 = female and 1 = male, and noise_sensitivity is the Weinstein Noise Sensitivity Scale score.