Dear SPMers,
I think this may be a basic problem but I really hope to confirm my idea….
I have 2 covariates (A, B) in a multiple linear regression model in SPM5. I’d like to know which region’s BOLD signals correlated with A without the effect from B, and vice versa. I entered these 2 covariates in SPM5 and after estimating the model, I thought contrast (1 0) revealed BOLD that correlated with A without the effect from B, and contrast (0 1) did in a reverse way.
The problem is: should I choose T or F test? What’s the difference between both tests?
Thank you very much!!
Carlos
|