Hi All,
Do we need to use weights and adjust for all the covariates used in CEM again to
avoid possible residual confounding? There is no need for adjust for the covariates if the
CEM makes perfect/Multivariate L1->0 match?
Thanks!
Fang
Show replies by date