The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations.¶
Why this mattered¶
Baron and Kenny mattered because they gave social psychology a shared grammar for asking when an effect occurs versus how or why it occurs. Before this paper, “moderator” and “mediator” were often blurred in empirical writing; the article made the distinction operational. A moderator changes the strength or direction of a relation, while a mediator lies on the explanatory pathway between a predictor and an outcome. That distinction turned vague causal interpretation into a research strategy: investigators could specify whether a theory predicted boundary conditions, mechanisms, or both.
The paper also made mechanism-testing newly routine. Its stepwise regression procedures gave researchers a practical template for testing mediation and moderation with the statistical tools already common in the 1980s. This helped move large parts of social, personality, health, organizational, and communication research beyond simple demonstrations that “X affects Y” toward models asking whether the effect depends on context, individual differences, or intervening psychological processes. Its extraordinary citation record reflects not only influence but infrastructural status: it became a default methodological reference for theory-building across the behavioral sciences.
Its later importance is also visible in the work that revised it. Subsequent breakthroughs in mediation analysis, including bootstrapped indirect effects, causal mediation frameworks, structural equation modeling approaches, and modern conditional process analysis, often began by clarifying, extending, or correcting the Baron-Kenny template. Many later methodologists criticized the original stepwise criteria as underpowered or insufficient for causal identification, but that criticism itself depended on the conceptual architecture the paper popularized. The paradigm shift was therefore not that Baron and Kenny supplied the final method, but that they made mechanisms and contingencies central objects of empirical explanation.
Abstract¶
In this article, we attempt to distinguish between the properties of moderator and mediator variables at a number of levels. First, we seek to make theorists and researchers aware of the importance of not using the terms moderator and mediator interchangeably by carefully elaborating, both conceptually and strategically, the many ways in which moderators and mediators differ. We then go beyond this largely pedagogical function and delineate the conceptual and strategic implications of making use of such distinctions with regard to a wide range of phenomena, including control and stress, attitudes, and personality traits. We also provide a specific compendium of analytic procedures appropriate for making the most effective use of the moderator and mediator distinction, both separately and in terms of a broader causal system that includes both moderators and mediators.
Related¶
- cite → Convergent and discriminant validation by the multitrait-multimethod matrix. — Baron and Kenny cite the multitrait-multimethod matrix as a validity framework for distinguishing mediation and moderation constructs.
- enables → Psychological Safety and Learning Behavior in Work Teams — Baron and Kenny's moderator-mediator framework enabled Edmondson to test whether psychological safety mediated links between team context and learning behavior.
- cite ← Psychological Safety and Learning Behavior in Work Teams — Edmondson uses Baron and Kenny's moderator-mediator framework to test how psychological safety relates to team learning behavior and performance.
- enables ← Convergent and discriminant validation by the multitrait-multimethod matrix. — The multitrait-multimethod framework established discriminant validity logic that Baron and Kenny applied to distinguish moderator and mediator variables.