In observational studies, propensity scores are commonly estimated by maximum likelihood but may fail to balance high-dimensional pretreatment covariates even after specification search. We introduce ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback