Hi,
I am running 3dMVM for a simple 1-way 3 group ANOVA and adding the post hoc ttests in the glt.
When I run the 3dMVM the fitt statpar for the ttest subbrik in the image header gives me degrees of freedom for the ANOVA, which differs from the fitt statpar if I run the same with 3dTtest++.
3dMVM
-- At sub-brick #4 'DBSvCon:t' datum type is float: -5.31884 to 4.19297
statcode = fitt; statpar = 100
vs
3dTtest++
-- At sub-brick #1 'pat-con_Tstat' datum type is float: -5.50756 to 5.5882
statcode = fitt; statpar = 38
the code for the 3dMVM is
3dMVM -prefix ${ResultsFolder}/${OutFile}.nii.gz \
-jobs 6 \
-mask ${MaskFile} \
-model "group" \
-num_glt 3 \
-gltLabel 1 DBSvR01 -gltCode 1 'group : 1*DBS -1*R01' \
-gltLabel 2 DBSvCon -gltCode 2 'group : 1*DBS -1*control' \
-gltLabel 3 R01vCon -gltCode 3 'group : 1*R01 -1*control' \
-cio \
-dataTable \
Subj group InputFile \
...
Is this what I should expect? or am I missing something?
The output image is the same just the thresholding is different.
thanks, Justin
Edited 1 time(s). Last edit at 06/19/2013 12:52PM by discoraj.