Dear Experts,

I successfully run two 2-sample t-tests (10 and 17 subjects, independen, unequal variance) using the swmc1 and swmc2 images separately.

Now I would like to use the whole (=nonsegmented) brain, but still making use of the warping to the common template.

I therefore "created warps" for the native T1 (= wmT1), but when running the 2-sample ttest, it gives significant voxels all over the brain when contrasting the larger group > smaller gorup... Is it because of the unequal number of subjects? Intercept? What would I have to do?

How would one normally do 2 sample ttests without the segmentation? Is there a reason no to do it?

Thanks for any hint!
Markus