Many replications do not causal inferences make: the need for critical replications to test competing explanations of nonrandomized studies.

Many replications do not causal inferences make: the need for critical replications to test competing explanations of nonrandomized studies.