Weighting in PSM using R

Inverse probability weighting (IPW) is adopted to adjust for confounding in causality assessment in observational studies. It is grounded in the framework of potential outcomes and relies on the use of – you guessed it – propensity scores.

Once the propensity scores are estimated, each individual is assigned a weight. This is achieved by reweighting each data point using propensity scores, , which are the probabilities of receiving the treatment given the confounders. Treated individuals are reweighted by and untreated ones by . The ATE is then estimated by calculating the difference in expected outcomes between treated and untreated groups in this adjusted pseudo-population, essentially balancing out the influence of confounding variables to reveal the treatment’s true impact.

These weights are designed to create a synthetic sample in which the distribution of covariates is independent of treatment assignment, mimicking a randomized...