On Thu, Mar 5, 2009 at 8:29 PM, John Colby <[log in to unmask]> wrote:

Hello all,

I am doing a pretty basic 3 group comparison in TBSS, investigating whether
there is any group effect on FA. Demeaned age is included as a 4th EV
because this is a young population and the groups are not precisely age
matched. I entered the 6 pair-wise t-test contrasts for potential future
post-hoc examination, and selected 2 of them for an F-test to answer the
question, "Are there any areas where FA is significantly different between
groups?" See the attached image for my model setup.

What I am confused about is this: When age is included as a covariate (as
opposed to when it is not), the individual FA_tfce_corrp_tstat maps are more
robust, but the overall F-test map (FA_tfce_corrp_fstat) drops almost to
zero. This opposite behavior surprised me and makes me think that I might be
doing something wrong in my model setup, or perhaps just not fully
understanding how F-tests work in FSL. Does this seem like normal behavior,
or does it look like I'm doing something wrong? Any help would be appreciated!

Randomise runs fine (no errors and all permutations finish). It was called as:
randomise_parallel -i all_FA_skeletonised -o FA -m mean_FA_skeleton_mask -d
design.mat -t design.con -f design.fts -n 1000 --T2 -V

Thanks!

John Colby

MD/PhD Student
Developmental Cognitive Neuroimaging Lab
David Geffen School of Medicine at UCLA