I remember seeing a paper that ranked causal statistical models, published by some research body, by quality in terms of generalizability or some similar facet(s). This would be a "hierarchy of evidence" type of list. The models included difference in difference (ranked highly), instrumental variables, and other techniques of exploiting natural experiments, pseudo-randomization, etc.

Obviously, depending on context (assumptions, data availability, academic field), different models are applicable in different situations. With that caveat, the models were ranked, and I would like to find it again to review the argumentation used.

Although I am discussing a specific publication, I am interested in any literature related to general comparisons or rankings of causal statistical models.

  • $\begingroup$ Answers to this question may be relevant. $\endgroup$ – Lizzie Silver Oct 25 '15 at 19:10
  • $\begingroup$ The only way I could see some ranking work is based on the plausibility of identification assumptions. However, this is more context based than method based, so unless somebody is ranking methods based on a specific scientific context/subject, I find it very hard to believe the ranking would be meaningful. $\endgroup$ – Carlos Cinelli Aug 21 '17 at 21:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.