Dear Christopher,
I'm still not clear about what you said, sorry for this.
(1) I saw others used this command 'cluster -i tbss_tfce_corrp_tstat1.nii.gz -t 0.95 --mm' ,maybe I didn't express my problem clearly in previous email. I typed four command lines as tbss_tfce_corrp_fstat1.nii.gz -t 0.95 --mm,tbss_tfce_corrp_tstat1.nii.gz -t 0.95 --mm,tbss_tfce_corrp_tstat2.nii.gz -t 0.95 --mm and tbss_tfce_corrp_tstat3.nii.gz -t 0.95 --mm, respectively. I got no significant cluster in the first command and got same significant result in the command line 2 to 3. Regarding F-test is two-tailed and t-tests areone-tailed, I changed the threshold to 0.975 in first command line but still no significance. Is the above results mean that the three groups have no difference between any of them while individual groups have FA values different from 0?
(2) Yes, the scores are related to disease severity so controls didn't have this scores. An error message appeared 'error: efce requires a positive delta T input. I wonder that i should put the three groups in seperate EVs, and us contrast[0 0 0 1 0 0] for positive correlation and [0 0 0 -1 0 0] for negative correlation. Is this correct?
(3) I mistakenly added a s to tbss_tfce_tstat2, really sorry for my carelessness.
Thanks very much for your help and the time.
Best,
Qin
|