I have a question of specifying the contrast when we do group analysis which has covariates of no interest in SPM5. For example, multiple regression with age as covariate of interest and brain volume as covariates of no interest. In SPM5 it has two places for us to put covariates,
1. Do I put group and age in the first place and put brain volume in the second place?
2. If I have 18 subjects in each group, how to I put group effect in the design?
3. How do I specify the contrast (or contrasts), if I want test group 1> group 2, or group 2> group 1?
4. I can do two sample t test, but the question is the same, how do I differentiate covariates of no interest (brain volume) and covariates of interest (Age) in the design matrix, and how do I specify the contrasts to test group 1> group 2, or group 2> group 1 ?
Thanks for any help in advance.
Free Windows Live software. Chat, search, share pics and more