A/B tests are the golden standard of causal inference because they allow us to make valid causal statements under minimal assumptions, thanks to randomization. In fact, by randomly assigning a treatment (a drug, ad, product, …), we are able to compare the outcome of i...