Unconfoundedness
Unconfoundedness¶
\begin{equation*}
\{Y_{i}(1), Y_{i}(0)\} \perp\!\!\!\!\perp D_{i} \mid X_{i}
\end{equation*}
Unconfoundedness is the assumption that, conditional on covariates \(X_{i}\), the assignment of the treatment variable \(D_i\) is independent of the potential outcomes \(\{Y_{i}(1), Y_{i}(0)\}\).
Matching¶
TBD
Propensity score methods¶
TBD
IPTW estimator¶
TBD