I'm conducting a TBSS analysis with 3 groups of subjects (an ANOVA) with 2 covariates (age and sex).
The contrast is something like:
-1.000000e+00 0.000000e+00 1.000000e+00 0.000000e+00 0.000000e+00 (F-test)
0.000000e+00 -1.000000e+00 1.000000e+00 0.000000e+00 0.000000e+00 (F-test)
1.000000e+00 -1.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00
-1.000000e+00 1.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00
1.000000e+00 0.000000e+00 -1.000000e+00 0.000000e+00 0.000000e+00
-1.000000e+00 0.000000e+00 1.000000e+00 0.000000e+00 0.000000e+00
0.000000e+00 1.000000e+00 -1.000000e+00 0.000000e+00 0.000000e+00
0.000000e+00 -1.000000e+00 1.000000e+00 0.000000e+00 0.000000e+00
From the analysis I obtain 1 file f-test.nii and 8 t-test files.
Two questions
1) The design is correct?
2) The 6 files (the 6 t-tests) have to be masked with the results of the F-test or they are the results of a post-hoc analysis? The results show large significant areas (with p<0.05) for some t-tests and smaller significant areas for the F-tests.
Andrea
|