Hi Rachel, are you saying that there does not seem to be an effect in the mean using the same contrast that was used to compute the clusters? How are you doing the test in the mean? Are you sure you have the same design and contrast matrices?
doug
On 06/06/2014 02:38 PM, Rachel Jonas wrote:
Hi Freesurfer users,
I ran a GLM vertex analysis testing an interaction between group and a task measure, and a few clusters popped up that survived correction. I extracted the mean thickness values at the clusters (using cache.th13.abs.y.ocn.dat) in order to plot these interactions in graph form. When I did this, the interactions look nonexistent, and there doesn't appear to be any relationships between my task measure and thickness in either group.
My FSGD file contains 4 groups (patients & controls with two different scanner sites), and my continuous variables are age and this task measure. I only have 15 patients and 15 controls, so my only thought so far is that my power is low, and these results are spurious.
My contrast file is: 0 0 0 0 0 0 0 0 0.5 0.5 -0.5 -0.5
Any insight would be greatly appreciated! Thank you!
Rachel
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer