[FieldTrip] Correlating reaction times with time-frequency activity

Emilio A. Valadez eavaladez at gmail.com
Fri Nov 8 20:05:10 CET 2019


Dear FieldTrip community,

I would like to use ft_freqstatistics to test correlations between trial
reaction times and time-frequency power values. I know that this can be
done on a between-subject level (correlating subjects' mean reaction times
with their mean time-frequency responses), but I am wondering if a similar
analysis can be done on a single trial level across all subjects. That is,
can ft_freqstatistics be used to test whether, for a given sample of
subjects, there is a within-subject correlation between single-trial RT and
single-trial time-frequency activity? If so, what should cfg.design look
like and which cfg.statistic is most appropriate?

For cfg.design, I was thinking it might need to be something like this:

cfg.design(1,:)  = [400 432 450 500 540 325 456 299 302 487]; %
single-trial RTs in ms
cfg.design(2,:)  = [1 2 3 4 5 6 7 8 9 10]; % global trial numbers across
all subjects
cfg.design(3,:)  = [1 1 1 1 2 2 2 3 3 3 ]; % subject numbers

...with cfg.ivar = 1, cfg.uvar = 2, and cfg.wvar = 3, but I'm not sure
whether this is a correct use of cfg.wvar. Am I on the right track?

Thank you,
Emilio


*--*
Emilio A. Valadez, Ph.D.
Postdoctoral Fellow
Child Development Lab
University of Maryland, College Park
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20191108/af212fee/attachment.html>


More information about the fieldtrip mailing list