Situates causal inferences in a broader, almost context that is enormously help understanding what causal graphs are trying to accomplish and how the complement existing approaches. Contextualizing methodology in such a way should be done far more often. Super helpful. Thanks.
