Hi,
I'm running a group analysis in randomise with a pretty large dataset. Some of the subjects have missing voxels and I was wondering if this can be modelled using voxelwise covariates? I've created a *single* covariate with voxel present / voxel absent values - is this correct or should I have a single voxelwise covariate for each individual with missing voxels (...RAM permitting...).
All the best,
Jonathan
|