Key points are not available for this paper at this time.
This Viewpoint discusses experiments where the sample is exposed to a condition outside the researchers’ control, allowing for reasonable conclusions on the potential causal link between exposure and outcome.
Khullar et al. (Fri,) studied this question.