Hi, guys,
I have 3 conditions in an even-related design and I'd like to find out voxels that show a differential response to the 3 conditions (without necessarily predicting the exact relationship among conditions). I want to do this for single-subject analysis. It seems to me that I need ANOVA. I read the manual for 3dDeconvolve and seems this is the thing to do
$3dDeconvolve -input tseries-090223-172328.nii -polort 2 -float -num_stimts 3 \
-stim_file 1 'Stim_1D.txt[0]' -stim_label 1 'Low' -stim_minlag 1 0 -stim_maxlag 1 8 \
-stim_file 2 'Stim_1D.txt[1]' -stim_label 2 'Medium' -stim_minlag 2 0 -stim_maxlag 2 8 \
-stim_file 3 'Stim_1D.txt[2]' -stim_label 3 'High' -stim_minlag 3 0 -stim_maxlag 3 8 \
-fout -tout -vout -bout -rout\
-num_glt 3 \
-gltsym 'SYM: +High[0..8] -Medium[0..8]' -glt_label 1 'High-Medium' \
-gltsym 'SYM: +High[0..8] -Low[0..8]' -glt_label 2 'High-Low' \
-glt 2 all_cond.mat -glt_label 3 'three_levels'\
-iresp 1 IRF_Low \
-iresp 2 IRF_Medium \
-iresp 3 IRF_High \
-bucket anov_test
So I'm doing 2 t-tests first, and the 3rd glt is based on the all_cond.mat, which looks like
0 0 0 1 1 1 1 1 1 1 1 1 -1 -1 -1 -1 -1 -1 -1 -1 -1 0 0 0 0 0 0 0 0 0
0 0 0 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 -1 -1 -1 -1 -1 -1 -1 -1 -1
This is what I gathered from the 3dDeconvolve manual, and I do get a F-value for the 3rd glt. My question is, is this the right way to do a single-subject ANOVA test?
Thank you very much,
--ts