PMAP 8521 • Example: Matching and IPW with R: 4: Matching

Demonstration of how to use the MatchIt package in R to close DAG backdoors, generate synthetic treatment and control groups, and estimate causal effects from observational data
Download the data and see the polished code at evalf20.classe...

[@gemmay2k974]

Exactly what I needed to help - brilliantly presented. Thank you!

[@zasimazharsiddiqui7528]

Very nice explanation. I have one question if I am working on two research aims one with the continuous dependent variable (linear regression) and the other with a binary outcome (logistic), but my primary independent variable (X) is binary and my "X" and covariates for both the research questions are same, only "Y" change for each aim. After calculating the propensity score or IPTW for the X~ covariates. Can I directly use my new dataset for both the linear and logistic models since the matching and weighting do not take the outcome variables into consideration? As my "X" is binary t/t vs. no t/t the PS and IPTW were calculated as the same method you demonstrated. Thank you very much!