baseline and normalization question

[Reyna L. Gordon]

Dear FieldTrip list,

I am a new user of the toolbox and am currently using it for Wavelet
analysis of my EEG data, in preparation for the cluster randomization
analysis.

My question concerns normalization: my understanding of normalization
is that it serves to either preserve or eliminate differences,
depending on what parameters you choose. So if I want to put
individual subjects on equal footing (de-emphasizing their individual
differences in spectral power, in the case of this time frequency
analysis), as well as different frequencies (to control for the fact
that the different bands tend to have different power), I would
normalize across subjects (by, for instance, using the change in power
instead of their raw power) and do this separately for each frequency
of interest. The freqbaseline function in the FieldTrip toolbox seems
to calculate a separate baseline for each electrode channel (in
addition to each frequency), whereas our thinking was that in order to
find out where (i.e. in which channel) the power is more prominent, it
would be preferable to obtain one baseline power value (for each
frequency) from averaging together the power from all the channels.

I have prepared some code to accomplish this on my FieldTrip data, but
I am unsure about the theoretical implications of performing frequency
baseline correction in this manner. Already I can see one downside of
what I am trying in my data, which is that ambient noise in certain
frequency bands shows up more on certain channels than others. Does
anyone have insight on whether there are other disadvantages or
violations of assumptions associated with "our" normalization method?

Likewise, if I instead use the freqbaseline tool and normalize
separately for each channel, [how] does the cluster randomization
statistical analysis compensate for the (de-emphasized intrinsic)
differences between channels?

Thank you very much for any all input.

Regards,
Reyna

-----------------------
Reyna L. Gordon
PhD Candidate
Center for Complex Systems & Brain Sciences
Florida Atlantic University
gordon at ccs.fau.edu

----------------------------------
The aim of this list is to facilitate the discussion between users of the FieldTrip  toolbox, to share experiences and to discuss new ideas for MEG and EEG analysis. See also http://listserv.surfnet.nl/archives/fieldtrip.html and http://www.ru.nl/neuroimaging/fieldtrip.