The random assignment plus masking are supposed to make it likely that the two groups have the same distribution of causal factors. It is controversial how confident these measures should make us that they do this. This issue bears on the trustworthiness of causal claims backed by RCTs. As we noted, trustworthiness is the central […]