Estimating population average treatment effects from experiments with noncompliance

Kellie Ottoboni, Jason Poulos

Randomized control trials (RCTs) are the gold standard for estimating causal effects, but often use samples that are non-representative of the actual population of interest. We pro…