Assessing the role of evidence of mechanisms in causal extrapolation



Published 19-12-2020
Saúl Pérez-González
Valeriano Iranzo


Extrapolation of causal claims from study populations to other populations of interest is a problematic issue. The standard approach in experimental research, which prioritises randomized controlled trials and statistical evidence, is not devoid of difficulties. Granted that, it has been defended that evidence of mechanisms is indispensable for causal extrapolation. We argue, contrarily, that this sort of evidence is not indispensable. Nonetheless, we also think that occasionally it may be helpful. In order to clarify its relevance, we introduce a distinction between a positive and a negative role of evidence of mechanisms. Our conclusion is that the former is highly questionable, but the latter may be a trustworthy resource for causal extrapolation.

How to Cite

Pérez-González, S., & Iranzo, V. (2020). Assessing the role of evidence of mechanisms in causal extrapolation. THEORIA, 36(2), 211–228.
Abstract 444 | PDF Downloads 232 XML Downloads 38



Extrapolation, evidence of mechanisms, statistical evidence, mechanism, causality, evidence, external validity, randomized controlled trial.