Initial applications of spectral analysis for finding the relationship between the EEG recordings from different regions of scalp dates back to 1960's.
[1] Corticocortical coherence has since been extensively studied using EEG and MEG recording for potential diagnostic applications[2] and beyond.
While the consensus suggests that the functional neural communication between distinct brain sources leads to synchronous activity in those regions (possibly connected by neural tracts, in either direct or indirect way),[3][4][5] an alternative explanation emphasises on single focal oscillations that occur at single brain sources that eventually appear connected or synchronous in different scalp or brain source regions.
[6] Corticocortical coherence has been of special interest in delta, theta, alpha, beta and gamma frequency bands (commonly used for EEG studies).
[8] Statistical significance of coherence is found as function of number of data segments with assumption of the signals' normal distribution.