Corrected EEG data was then transformed into reference-free current source density (CSD) estimates using a spherical Laplacian algorithm (Nunez et al. 1997; Mima and Hallett 1999), which addressed several issues that may confound coherence between two scalp electrodes. For one, many of the commonly used referencing approaches, namely nose reference, linked earlobe or mastoid reference, and average reference, have all been shown to result in incorrect coherence estimates to a varying degrees (e.g., Fein et al. 1988; Nunez et al. 1997; Mima and Hallett 1999; Schiff 2005). Using this reference-free measure would thus circumvent problems associated with the choice of reference (Nunez et al. 1997; Mima and Hallett 1999). Further, this step removed volume conduction effects, which ensures the observed local coherence was not the consequence of two neighboring electrodes receiving signals from the same source (Nunez et al. 1997). In other words, CSD estimates more accurately index the source of neuronal activity from the underlying generators (Nunez et al. 1997; Tenke and Kayser 2012). Given minimal effects of reference and volume conduction on CSD estimates, coherence measures based on
