EOG reduction algorithms for two reasons: 1) these algorithms sometimes introduce artifactual data, principally because the eye recording does not accurately represent the actual EOG effect on the EEG, and 2) there were very few trials that had EOG contamination. Further, those averaged waveforms either with ‘atypical’ morphology of ERP waves/components (e.g., flat P3 or P2 larger than P3) or with suppressed/flat response (e.g., very low voltage ERP components shaped like uniform sine waves of about 2-4 μV in amplitude) were also rejected from further analysis. Only the trials with proper response (i.e. button press within the selection window) were averaged.
