[Eeglablist] magnitude squared coherence - trial averaging

[Germano Gallicchio]

Dear EEGLAB users,

I am computing the magnitude squared coherence between electrode pairs.

(1) Is there any EEGLAB function to compute this specific kind of coherence
measure?

(2) I am currently using the MATLAB function mscohere(), which takes
several arguments including the 2 vectors that I want to compare.
Naturally I have several trials per condition and I am not sure which of
the following procedures is the correct one:
a. average trials per condition (like I would do for ERPs) and then compute
the coherence over these trial-averaged vectors
b. compute coherence for each trial and then average the single-trial
coherence, per condition
Could any of you please help me select the right procedure?

(3) Do the number of trials per condition influence the computation of the
magnitude squared coherence?

Thanks for your time.

All the best
Germano
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20140702/7f391422/attachment.html>