Grand average ERP waveforms for each stimulus condition (Fig. 1) and average waveforms for each patient and control subject were carefully inspected to select time windows that bracketed the peaks for the ERP components and optimized their measurement across subjects: 70–145 ms N1, 150–240 ms for N2, and 220–470 ms for P3. These windows were used to compute mean integrated amplitudes for each component at each recording site. The mean latency of P3 peaks to novel and target stimuli were also computed for patients and controls at the midline parietal site (Pz) where P3 amplitude is maximal. Although there was no significant difference in the P3 latency for patients and controls, P3 latency was shorter for novels (mean = 312 ms, S.D. = 36) than for targets (mean = 324 ms, S.D. = 34; F = 5.36, df = 1,38, P = 0.026). The waveforms in Fig. 1 illustrate, however, that there was considerable overlap of the P3 components for novel and target stimuli, and therefore the same broad window (220–470 ms) was used to incorporate both the novelty P3 and target P3 for each patient and control.
