EEG data were recorded on a 128-channel eego system (ANT Neuro, Enschede, The Netherlands) with a 1000-Hz sampling rate. CPz was used as the online reference and the ground electrode was placed close to the left mastoid. Electrode impedance was kept under 20kΩ during the experiment.
Preprocessing was conducted with EEGLAB (Delorme & Makeig, 2004 (link)) and ERPLAB (Lopez-Calderon & Luck, 2014) (link) in MATLAB (Mathworks, Natick, MA). Data were re-referenced off-line to the average reference. Band-pass filtering (0.01-30Hz) was applied to the continuous EEG data, which were then divided into 1000-ms epochs starting at 200ms before and ending at 800ms after the presentation of the object. Since the signals near the end of the stimulus presentation was not of interest, the last 200ms of stimuli presentation (i.e., 800-1000ms) was not used for artifact rejection to avoid rejecting epochs due to artifacts during this time window. Incorrect trials were also excluded from further analyses. Epochs with ocular artifacts were removed by visual inspection and by the moving-window peak-to-peak function in ERPLAB on VEOG, HEOG and the channels selected in the decoding analysis with a threshold of 100 μV, a window size of 200ms and a step size of 50ms. On average, 4.57% and 6.24% of the trials were rejected due to incorrect responses and artifacts respectively.
Preprocessing was conducted with EEGLAB (Delorme & Makeig, 2004 (link)) and ERPLAB (Lopez-Calderon & Luck, 2014) (link) in MATLAB (Mathworks, Natick, MA). Data were re-referenced off-line to the average reference. Band-pass filtering (0.01-30Hz) was applied to the continuous EEG data, which were then divided into 1000-ms epochs starting at 200ms before and ending at 800ms after the presentation of the object. Since the signals near the end of the stimulus presentation was not of interest, the last 200ms of stimuli presentation (i.e., 800-1000ms) was not used for artifact rejection to avoid rejecting epochs due to artifacts during this time window. Incorrect trials were also excluded from further analyses. Epochs with ocular artifacts were removed by visual inspection and by the moving-window peak-to-peak function in ERPLAB on VEOG, HEOG and the channels selected in the decoding analysis with a threshold of 100 μV, a window size of 200ms and a step size of 50ms. On average, 4.57% and 6.24% of the trials were rejected due to incorrect responses and artifacts respectively.