Kayser, 2001) and activity at bridged electrodes was replaced by spherical spline interpolation from unaffected sites. Only one control subject showed bridging at any of the sites reported here (only O1 to Oz), and exclusion of this subject's data did not affect our findings. The filtered, continuous EEG was then blink corrected using a spatial, singular value decomposition filter generated from identified blinks and artifact-free EEG periods (NeuroScan, 2003). Additional artifact detection and correction was performed using a reference-free approach (Kayser and Tenke, 2006b). Trials containing artifacts in more than 8 channels were rejected; otherwise, contaminated channels were interpolated from the data of the artifact-free channels by means of spherical splines (Perrin et al., 1989). Artifact detection and electrode replacement was verified by visual inspection. Artifact-free data were averaged for correct responses to targets, novels and nontargets. The number of trials contributing to averages did not differ between groups for targets (controls: mean = 42, S.D. = 4; patients: mean = 41, S.D. = 6; t = 0.86, df = 38, ns), novels (controls: mean = 39, S.D. = 5; patients: mean = 37, S.D. = 7, t = 1.03, df = 38, ns), or nontargets (controls: mean = 272,
