Randomized Controlled Trial

A randomized controlled trial (RCT) assigns each unit to treatment or control by chance, like a coin flip, and then compares outcomes between the two groups. Randomization (隨機分派) makes the groups statistically alike on everything — both measured and unmeasured — except the treatment itself.

This is the gold standard for causal inference because it breaks the link between treatment and any confounder: any difference in average outcome can be attributed to the treatment. Its limits are practical and ethical — you cannot randomize people into smoking or poverty — which is why observational and quasi-experimental methods exist.