Jovo and Eric have causal combat which forces stricter match between demographic info across groups.
We could consider many problems in a batch effect context... E.g.
Rather than regressing out age/sex/etc., treat them as batches within a site
Determine the number of heterogeneous consortia required to robustly remove batch effects (i.e. how many sites are required to find a robust lowest common demoninator)
Jovo and Eric have causal combat which forces stricter match between demographic info across groups.
We could consider many problems in a batch effect context... E.g.