[Eeglablist] Spectral analysis assumptions and EEG

[Virgil Stokes]

In the last decade there have been considerable progress in the analysis of 
human EEG. However, if one examines the methods currently being used for 
spectral analysis, many of them are based on two important assumptions:  the 
signal is stationary (in the wide-sense)  and infinite in length.

First, what is the recommended test for stationarity of EEG signals?
Second, where can one find a list of spectral methods that are *not based on 
stationarity and finiteness*.
Third, is it possible to rank the plethora of spectral methods in order of being 
"nearest" to handling signals that are nonstationary and finite in length?

Best regards!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20111017/43681bb0/attachment.html>