Hello EEG/SPM experts,
I was wondering if the following analysis was possible using SPM12:
I am conducting a spatiotemporal analysis using ERPs converted to smoothed images. I have two covariates and one factor within my design matrix, and I would ideally like to see where in scalp space and when in the epoch a covariate correlates with ERP activity while controlling for the second covariate.
To my understanding, the analogous fMRI analysis would achieve this during first-level analysis, by orthogonalizing one covariate against the other, akin to a step-wise or hierarchical regression analysis.
However, using SPM12 to conduct spatiotemporal analysis, I believe one converts ERPs to smoothed images and proceeds to second-level analysis initially at the participant level, and then at the group level.
Is it possible to conduct a multiple regression analysis using EEG data where a covariate is used to explain the residual variance after controlling for another covariate?
Any help would be greatly appreciated.
|