coherence and inter-trial coherence questions

[Eric Maris]

Dear Matteo,





I have two questions regarding statistical tests: first one regarding
coherence differences, second one pertaining testing inter-trial coherence.

We are analysing MEG data from an auditory experiment with 5 subjects. We
considered the deviant stimulus as the only stimulus in the recordings,
while the standard one was considered as baseline activity.

We analysed the TFR data in both designs: within-trial and within-subject
using respectively "actvsblT" and "depsamplesT" statistics. In both cases we
compared a time interval extracted from the baseline versus a time interval
from the deviant stimulus response.

We wish to assess coherence differences using "indepsampleZcoh" statistic in
both single-subject and multiple-subject cases. In a previous post
<https://listserv.surfnet.nl/scripts/wa.cgi?A2=ind1004&L=FIELDTRIP&D=0&P=108
12>  Dr. Maris replied that this statistic only works in a between-trial
design (for single-subject case) and suggested to compare the different
conditions as they were un-paired.
We exploited this idea for the single-subject case and we would like to know
if it is legitimate to extend the same idea for the multiple-subject
situation?



In the multiple-subject situation, you should first calculate the two
coherences (one for the deviant stimulus and one for the baseline condition)
for every subject. These coherences should be compared using depsamplesT
(and not using indepsamplesZcoh, which only works for a single-subject
study). I'm fairly sure that the testing of coherence in a multiple subject
study has already been addressed in the Fieldtrip discussion list. I think
it was Jan-Matthijs Schoffelen who gave a extensive description on how to do
it.




Furthermore we would like to know if there is a way to test inter-trial
coherence differences in Fieldtrip?
In a previous
<https://listserv.surfnet.nl/scripts/wa.cgi?A2=ind0803&L=FIELDTRIP&D=0&P=180
> post Stephan suggested how to deal with the problem; is it already
implemented?

If it has been implemented, I'm not aware of it. However, if you are a
little bit proficient in Matlab programming, you can write you own statfun
and have Fieldtrip apply it to your data. By debugging a statistical
analysis that uses a built-in statfun, you can learn how the algorithm
works.





Good luck,



Eric Maris













Thanks in advance for your time

Matteo


----------------------------------

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.

http://listserv.surfnet.nl/archives/fieldtrip.html

http://www.ru.nl/fcdonders/fieldtrip/


----------------------------------
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.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20101001/561516f1/attachment-0001.html>