[FieldTrip] design for ft_freqstatistics

[Elena Krugliakova]

Dear Fieldtrippers,

I have a question about design for ft_freqstatistics.
I have only 4 subjects with 2 conditions*300 observations.
So far I did a single - subject statistics, but I would like to estimate
the effect on the group level by pulling subjects together without
averaging across trials. In this case I need to control for the fact that 4
subsets of my data are coming from different subjects.
I assume that design matrix should look like this (2 conditions, 4
subjects, 3observation(as an example)):

[image: Inline images 1]
*(I hope you can see the picture...)

cfg.ivar  = 1;  % 2 conditions
cfg.cvar  = 2; % I assume "subject" can be treated as a control variable
cfg.uvar  = 3; % observations

Do you think the control variable is a proper way to control for "subject"
effect?

Kind Regards,
Elena
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20170630/b5513e4a/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 4011 bytes
Desc: not available
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20170630/b5513e4a/attachment.png>