Below is a post written by Keith A. Markus, Ph.D. He is currently a Professor in the Psychology Department at John Jay College of Criminal Justice. He serves on the graduate faculty of the CUNY Graduate Center in the Quantitative Psychology subprogram of the Educational Psychology Doctoral Program and in the Industrial and Organizational Psychology and Forensic Psychology subprograms within the Psychology Doctoral Program.
This article may hold interest for students, researchers, behavioral science methodologists, and philosophers of science or of causation. Rubin and Pearl have each offered theories of causal inference by which researchers can make inferences about causation based on some combination of theoretical assumptions and empirical observation. Causal inferences play a central role in both applied and basic research. Lewis and Pearl each offered accounts of counterfactual conditionals. These conditionals are If-Then constructions that also play an essential role in scientific inference. In addition to having a close connection to causation, counterfactual conditionals are central to policy decisions and to reasoning about the secondary effects of a change.
The article contrasts each pair of theories, probing and articulating their differences. Claims to the contrary notwithstanding, Rubin and Pearl offer distinct theories of causal inference and Lewis and Pearl offer distinct theories of counterfactual conditionals. The mistake in reasoning that leads to mistaken claims of equivalence runs as follows. First, spell out one theory in its native notation. Second, use the notation of the other theory to also express the first theory. Lastly, show that the former is equivalent to the latter. It does not follow from this exercise that the two theories are equivalent. Instead, this only shows that the two expressions of the first theory are equivalent. It fails to show that the two theories, in any notation, say the same thing about their subject matter. In the above cases, they clearly do not. Recognition of these differences underscores a need for more comparative research.
The article’s electronic supplement includes introductory material that should make the article accessible without assuming prior familiarity with the theories in question. A copy of the article will appear in Academic Works after the version of record has been published in print. However, an electronic copy is currently available from the publisher or by request (firstname.lastname@example.org).
Markus, K. (2021). Causal effects and counterfactual conditionals: Contrasting Rubin, Lewis and Pearl. Economics and Philosophy, 1-21. doi:10.1017/S0266267120000437
Rubin and Pearl offered approaches to causal effect estimation and Lewis and Pearl offered theories of counterfactual conditionals. Arguments offered by Pearl and his collaborators support a weak form of equivalence such that notation from the rival theory can be re-purposed to express Pearl’s theory in a way that is equivalent to Pearl’s theory expressed in its native notation. Nonetheless, the many fundamental differences between the theories rule out any stronger form of equivalence. A renewed emphasis on comparative research can help to guide applications, further develop each theory, and better understand their relative strengths and weaknesses.
No comments yet.