Hello-
I have gotten very strange results and I am not sure what i might be doing wrong. I have two groups of subjects each with two visits spaced two weeks apart. I ran TBSS and used glm_gui in order to make a design matrix for group stats. I included the contrasts and design. I know the design is a little peculiar, but i have to make it in the order of the scans from the TBSS folder and the order was:

 Group1_sub1_visit1, Group1_sub1_v2, Group1_sub2_v1, Group1_sub2_v2, Group1_subn_v2, ...Group2_sub1_v1,...Group2_subx_v2


so that is why EV1 (in the EVs figure) looks like 1, -1, 1, -1 ... with a few exceptions.

There are several issues with the tstats. First, contrasts that shouldn't show anything (such as a control group scanned 2 weeks apart). The results show all white matter fibers through out the entire brain. As I increase the threshold, they all disappear at once. So, seemingly they are all equally significant--which isn't reasonable. This is the case for all contrasts that show significance.

Also, the contrasts expected to show changes either don't or show the opposite. Clearly I shouldn't question the validity based on not getting the results I expected, but with the addition of the previous oddities, i think something is wrong.

Is it something in my design?

Thanks so much and sorry for the multiple posts,

Greg