All recordings were digitally re-sampled offline at 256 samples per second. In order to extract the N2 component that would be independent of the effects of the initial ascending wave of the P3 and other slow deflections, a band-pass filter of 2-16 Hz was applied to the waveform (Figure 3, results section). The EEG segments were divided into epochs of 1625 ms (187.5 ms pre- and 1437.5 ms post-stimulus, similar to Kamarajan et al., 2005a). The EEG activity of 125 ms prior to stimulus onset served as baseline. All epochs exceeding ±75μV amplitude were automatically excluded from further processing. The eye blink, eye movement, and other artifacts were visually inspected and subsequently removed manually. Each subject had a minimum of 20 good (i.e., artifact-free) trials in each condition (Go/NoGo). Averages were computed for each condition and subject separately.
