If the covariates are collinear, then you can't properly estimate the
model. They are collinear if they explain the variance of each other.
I would first test for the relationship of your covariates before putting
them into the model as you don't want the covariates to be collinear.
Best Regards,
Freesurfers-
Question, when looking for interactions to justify using DOSS say i am
looking at cortical thickness and i have two discrete variables (G1 and G2)
and 4 continious variables (V1, V2, V3, V4). My thinking is that i should
look to see if there are interactions with the groups and each i