As intended, the unrestricted PCA for these CSD waveforms yielded optimized factors representing N1 sink activity, with highly comparable time courses of factor loadings (RM vs. WM, peak latencies 141 vs. 148 ms, explained variance 16.4% vs. 14.8%; Figure 2B). The corresponding factor score topographies of N1 sink were, for each paradigm, restricted to posterior, inferior-lateral regions, revealing a robust left-larger-than-right hemisphere asymmetry for words, and the opposite, less robust asymmetry for faces (Figure 2B). The overall N1 sink amplitude was substantially reduced for auditory hallucinators, although the task-dependent asymmetry was nevertheless preserved.
