Randomized controlled trials (RCTs) are regarded as the gold standard for causal inference in medicine and social science, but their external validity is often questioned.