Hi there,
I'm trying to assess whether a clinical group differs from a comparison group in terms of a difference in activity across runs, controlling for IQ and age. In other words, I expect that participants in the comparison group will show a greater decrement in activity in certain regions from run 1 to run 2 than will participants in the clinical group.
I set up my design matrix like this (this is a simplified version of the matrix with only one covariate and just two participants, one in the clinical sample and one in the comparison sample):
Group RunClin RunComp IQClin IQComp ID1 ID40
1 1 0 2.6 0 1 0
2 0 1 0 3.5 0 1
1 -1 0 2.6 0 1 0
2 0 -1 0 3.5 0 1
Unfortunately, FSL keeps telling me that this is rank deficient and throws errors saying "matrix is singular."
My guess is that this may be related to my attempt to include covariates in the analysis. Is it possible to run this type of analysis in FSL, and if so, what do I need to change about my set up?
Thanks in advance for your insight!
Best
Allison Jack
|