Closed jeremylhour closed 4 years ago
Athey and Imbens (2006) have the same formula for the variance as us. But they made a change of variables, u = F_Z^{-1}(F_Y(t)) [Uploading cicsupp.pdf…]() to have more elegant formulas. Recall that to be consistent with their results we have:
Details of their implementation is in the enclosed document: https://athey.people.stanford.edu/sites/g/files/sbiybj5686/f/cicsupp.pdf
Write code for computing standard errors using the formula in the paper