2.3 EEG Preprocessing
Brain Vision Analyzer (version 2.2.1.8266; Brain Products, Gilching, Germany) was used for data preprocessing and waveform analyses. The data preprocessing steps for each participant began with segmentation, followed by filtering (0.3 Hz high-pass with a slope of 24 dB/oct and 30 Hz low-pass with a slope of 48 dB/oct), bad channel selection, raw data inspection, independent component analysis (ICA) for corrections of eye blink and lateral eye movement artifacts using a restricted Infomax algorithm, bad channel interpolation with spherical splines, re-referencing to the average signals recorded from all electrodes, and artifact rejection (if an extreme voltage (>+80 or <-80 μV) was detected). A fixed delay of 18 ms was corrected due to the anti-aliasing filter of the amplifier (see the advisory notice about the timing affected by anti-alias filter effects used in Net Amps 300 amplifiers, 26 November 2014, EGI for more details). The epoch ranged from 168 ms pre-stimulus to 632 ms post-stimulus. Baseline correction was conducted based on pre-stimulus period data.
Based on these preprocessing steps, 24% of trials were rejected on average. For the adaptation and MMN analyses, only the final tones in the 4th to 30th positions were included, ensuring a balanced comparison between standards and deviants by maintaining a similar number of each. The average remaining trial numbers in the three conditions related to the adaptation effects, namely the deviants (1st tones), the 2nd tones, and the final tones in the 4th to 30th positions, were 230 (range: 152–297), 197 (range: 133–250), and 139 (range: 98–172) respectively. The average remaining trial numbers for the conditions of the 3rd to 10th tones, which are relevant to the adaptation pattern analyses, are detailed in Table S1 in Supplementary Material II.