Dear FSL team
I have some questions regarding the design matrix
I'm working with 60 subjects divided in 3 groups. Two groups are matched but the third no. I have 2 confounders in the model (age and total brain volume). The aim is to detect differences in FA (or non-FA parameters) among groups, adjusting by confounders.
The matrix:
the first column is group 1, second column is group 2 and the third column is group 3. I added 2 extra-columns with age and total brain volume demeaned. However, I obtained an error about the design:
Design matrix use different groups (for different variances) but this do not contain "separable" EVs for the different groups.
My questions:
- Do I need to split the age and TBV (raw values) per group? It means, to have 6 columns with confounders?
- What does this error mean?
Kind regards,
Nelly
|