Dear Gang and the AFNI experts,
This question concerns usage of 3dMVM, so I assume that it should be directed at Gang.
I'm interested in running ANCOVA, with 1 between-subjects factor (group, two levels) and 1 within-subjects factor (condition, three levels), and age as a covariate. I am interested in the group x condition interaction. I have two questions about this.
1. I am including age as a covariate because my groups differ in age, and I'd like to account in the model for linear effects of age, which are not of interest for our research question. I see on the 3dMVM help page (http://afni.nimh.nih.gov/pub/dist/doc/program_help/3dMVM.html) stated "However EXTREME caution should be taken when the groups differ significantly in the average value of the covariate." This confuses me, isn't the purpose of including a covariate often to account for variance in the model for relations/differences in a continuous variable of non-interest?
2. I've set up the 3dMVM commandline as follows:
3dMVM -prefix 3dmvm_output -jobs 12 -bsVars "group+age" -wsVars "condition" -qVars "age" @mvm_table.txt > 3dmvm_script_output.txt
First, please let me know if I'm using this command in an incorrect manner. My main question is: once using the AFNI viewer to the inspect 3dmvm_output, I see a subbrick that was automatically (i.e., i didn't have to specify a specific f-test) computed called "group:condition F". What do I set as the OLay and the Thr in order to have appropriate p-values computed on a voxel-wise basis in the viewer?
3. An unrelated question, is there an AFNI tool that can strip the cerebellum and brainstem from a high-res T1-weighted image?
Thank you for your help!
Matthew