[FieldTrip] depsamplesregrT help

[Jonas Obleser]

Hi Stephen,

from my understanding (greatly facilitated by Eric Maris a year ago or so),
depsamplesregrT does only test for linear increases or decreases in your data, depending on the sequence in which you input your data conditions.
For example, freqstatistics(cfg, cond1{:}, cond2{:}, cond3{:}) will result in positive clusters for data increasing (quasi-)linearly from cond1 to cond3, and negative clusters for data behaving vice versa. 

For regression/correlation with an independent (e.g., behavioural) regressor, we use a statfun Nathan Weisz (also active on the list here) once wrote and kindly shared with us. Hope this helps somewhat.

Best wishes, J.



Dr. Jonas Obleser
Auditory Cognition Group
Max Planck Institute for Human Cognitive and Brain Sciences
Leipzig, Germany
(p) +49 (0)341 9940 114
(e) obleser at cbs.mpg.de