Hi Gang,
I have successfully be able to implement a 3 (condition (within-subject)) x 2 (sex) ANCOVA (covariate, age) with 3dMVM. My full model output includes F tests for the following:
Intercept
age
sex
age:sex
age:condition
sex:condition
age:sex:condition
But I am having a hard time wrapping my head around the -wsMVT and when to use it. It says this option will only give you a different result when your within-subject variable has more than two levels. Hence, I have a within-subject variable of condition with 3 levels I used the -wsMVT option. However, my output ANCOVA now has the above but also F tests for:
condition -wsMVT
age:condition -wsMVT
sex:condition -wsMVT
age:sex:condition -wsMVT
My understanding is that I would want to look at these as they would be the more appropriate estimates given the 3 levels in my within-subject variable. However, what is confusing is that I get results for the condition -wsMVT (e.g. what is different between the 3 levels of the conditions), but the remaining -wsMVT F results are completely blank without thresholding and in the view the autorange for each of these F sub bricks is 0 to %. I don't get any errors and this is what comes out in the command line:
++ 3drefit: AFNI version=AFNI_2011_12_21_1014 (Apr 11 2014) [64-bit]
++ Authored by: RW Cox
++ Processing AFNI dataset GO-NOGO135ANCOVA+tlrc
+ created 12 FDR curves in dataset header
++ 3drefit processed 1 datasets
And in the diary.txt:
[1] "Congratulations! You have got an output GO-NOGO135ANCOVA+tlrc"
Should these F results be completely blank or am I missing something?
Thanks for your time and help in advance.
-Megan
