[FieldTrip] cluster-based permutation test - baseline correction

[Herring, J.D. (Jim)]

Dear Jose,

If you assume there to be (perhaps trivial) differences in the baselines of conditions A and B, but you are interested in the difference in change from baseline then doing a baseline correction is indeed what you want to do. If you do not assume a difference at baseline it is in principle not necessary to correct for it (It might even hurt: http://datacolada.org/39) .

With regards to ‘demeaning’ (and detrending), have a look at these two pages:
http://www.fieldtriptoolbox.org/faq/why_does_my_tfr_look_strange
http://www.fieldtriptoolbox.org/faq/why_does_my_tfr_look_strange_part_ii

It might be beneficial to demean and detrend before doing your frequency analysis.

Best,

Jim
From: fieldtrip-bounces at science.ru.nl [mailto:fieldtrip-bounces at science.ru.nl] On Behalf Of Jose
Sent: Monday, July 18, 2016 6:08 PM
To: fieldtrip at science.ru.nl
Subject: [FieldTrip] cluster-based permutation test - baseline correction

dear community,

I'm performing analysis of CTF MEG data. I have two conditions of interest: A and B, and I have performed spectral decomposition with the wavelet method to obtain the power (frequencies from 1 to 50 Hz). I would like to perform a cluster-based permutation test to assess if there are significant power differences between my two conditions of interest between the onset of my stimulus and 1 sec after,

While I understand the principles of cluster-based permutation test, with their distinct versions: between trials, within trials and within subjects, what I am interested in is to test differences between my two conditions in their baseline-corrected versions. Since I haven't seen this procedure in the tutorial, but it seems logic to me, I would like to verify the way to proceed is do baseline correction (for each condition and subject) and then run my cluster-permutation test. Is there anything else I should consider?

Also, since I haven't used 'demean' during my trial definition (I was advised not to do this), I was wondering if this could impact my baseline correction, and subsequent statistical analysis. I've seen some messages in the forum referring to this procedure,

Thanks in advance,
Jose
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20160719/d657ce4d/attachment-0002.html>