Sometimes I'll post some paper showing a good design applied to good data, where the causal identification is excellent and theoretically sound and someone still posts "Correlation not causation"
17,04K