Strategy to select most efficient RCT samples based on observational data

Wenqi Shi, Xi Lin

Randomized experiments can provide unbiased estimates of sample average treatment effects. However, estimates of population treatment effects can be biased when the experimental sa…