Internal and external validity reconsidered: Eliminating extraneous factors, understanding support factors

We salute the sociological science manifesto by Aksoy, Manzo, Birkelund, and Raub (2026). As experimental sociologists, we woud like to draw attention on one point of their argument, where they warn about the “external validity constraints” faced by experimental designs. What is external validity and why do we want it? After clarifying the logically asymmetric ideas of internal and external validity, we propose a renewed understanding of the latter. In this revised conception in tune with mechanism-based explanations and middle-range theorizing, the assessment of the external validity of an experiment starts with the recognition that every causal link necessarily depends for its operation on a host of enabling or support factors that must be set at the “right” level. In this light, the quest for a general “main effect” that operates independently from everything else reveals itself to be hopeless. The metric of external validity ceases to be the (necessarily limited) generality of a causal link to become the depth of understanding of its enabling conditions.