Thanks Donald. I think I misspoke in my previous email; I have a group of subjects who each performed three tasks. There was a behavioral difference between each of these tasks on the group level, and I'd like to see if a similar effect exists in the neural data. So I performed a one-way ANOVA with 3 factors (tasks 1, 2 and 3), including each of my subjects' maps in each of the tasks. Because each subject's performance on task 2 and 3 may be linked to her previous performance on task 1, I decided to perform a within-subjects ANOVA.
I found significant clusters of activation using this method for the omnibus test, which tells me that there is a group effect of task. Now I'd like to see which tasks are contributing to this significant result.
Is my approach valid? If so, should I still contrast tasks within the ANOVA model in SPM to see which tasks differ from each other? Alternatively, I guess I could use the paired t-test option in SPM.
Thanks again for your help!
Arul
|