Using visually-presented common words or unknown faces, schizophrenia patients showed significantly poorer RM and WM performance than healthy controls, which agrees with evidence of impaired memory processing in schizophrenia (e.g., Barch, 2005; Pelletier, Achim, Montoya, Lal, & Lepage, 2005). Despite poorer performance, the associated sequence of cortical activation was highly comparable in patients and controls, with matching N1 sink peak latencies (approximately 150 ms) and topographies (inferior lateral-parietal maximum), replicating previous findings (Kayser et al., 1999, 2006, 2009). Patients and controls revealed distinct, stimulus-dependent N1 sink asymmetries (left-larger-than-right for words and vice versa for faces). However, auditory hallucinators had markedly reduced N1 sink for words and faces compared to healthy controls and nonhallucinators, independent of memory paradigm. Avoiding known limitations of reference-dependent ERP measures (Kayser et al., 2006; Kayser & Tenke, 2010; Tenke & Kayser, 2012), the neuronal generators underlying N1 amplitude were concisely summarized by inferior-parietal sinks and quantified by temporal PCA after aligning the individual CSD waveforms for N1 sink peak latency.
