Open cyjtom92918 opened 4 years ago
Hi cyjtom92918,
The newest version of the package has updated how treatment effects are estimated. Firstly, the CATE
function has been changed and now uses some recent theory our group has developed for variance computations using our matching estimators. The implemented CATE estimator is asymptotically normal and so the means and variances you get out of the new CATE
function can be used to generate asymptotically valid confidence intervals. Average treatment effect estimates, which you can get out of the new summary
method, also have associated variances, though the estimators are not asymptotically normal. They will be updated with estimators that are asymptotically normal in a future release.
Hi, is there any feature in the package that we can calculate the confidence interval easily for ATE/ ATT/ CATE? Thank you!