We’re doing an fMRI study (with approximately 10 subjects in each of 2 groups), and would like to include several covariates.  While I’ve often used multiple covariates in behavioral studies with large sample sizes, my impression is that some of the statistical “guidelines” for behavioral studies are not necessarily directly applicable to imaging studies where one is collecting hundreds of data points on each subject.  Are there any guidelines (or references) that discuss how many covariates one can include?

Any suggestions greatly appreciated!