Hi FSL experts,
I am running some analyses using FSL's FEAT. I have been running third level group analyses looking at the difference between groups covarying for age, alcohol, and another group variable which is nominal and has been dummy coded.
My problem is, when I pull out the z values at the subject level (level 2 analysis), these group differences when covarying for the above variables in SPSS do not match up to the group differences that result from my level 3 analysis. Is there any reason as to why these results would not match up? My thoughts are that FSL is doing some behind the scenes work that SPSS doesn't do.
Any thoughts or comments would be greatly appreciated.
Danielle
|