This is three sets of data from the same underlying EEG, all with varying coherence results, and with the weighted average showing the alpha hypercoherent pattern with better fidelity than any other for this data.
These results are from 300 seconds of linked ear EEG data, note the dominant slower alpha peak frontally…. And the raw coherence values of that linked ear data. The raw EEG file sample is also included, so you can see the waveforms these values are being drawn from.
The same exact 300 seconds of EEG data, reprocessed now with the weighted average montage. Note the difference in spectra, and waveform!!! The temporal slower alpha is now seen as the source of that slower alpha content.
The alpha hypercoherence in the EEG is easily seen in this data, but not in the linked ears.
This shows that you need to find the EEG montage that shows the actual EEG data for your case first, and THEN calculate coherence.
The images below show the Spectral plot, coherence plot and raw EEGs. Just like the other montages did. The Cz coherences are so inflated with field effects they are at 0.8 across the full spectrum at some sites, obviously artifactually high.
You can see that the commonly stated though incorrectly held opinion that weighted average ruins coherence is not at all correct. The processed EEG data show the reality of the real world of EEG, which shows that when temporal sites are involved, linked ears is not the rational choice.
When doing coherence measurements, the first step is to find the right montage to show the EEG “correctly”, and then the coherence can be calculated off the time series provided from that data.