Contents<\/p>\n\n\n\n
In behavioral, social, health, and organizational research, it is rarely enough to know that an independent variable<\/a> (IV) is related to a dependent variable<\/a> (DV). Researchers also want to know how and why that relationship occurs, and under what conditions it gets stronger, weaker, or disappears altogether. Two statistical concepts answer these questions: mediators, which explain the mechanism or process underlying a relationship, and moderators, which explain the conditions or boundary limits of a relationship.<\/p>\n\n\n\n These two terms are among the most confused in quantitative research<\/a>, partly because they sound similar, partly because both involve a \u201cthird variable,\u201d and partly because both can be tested using regression-based methods. However, they answer fundamentally different questions, sit in different positions in a causal diagram, are tested with different statistical procedures, and require different interpretations.<\/p>\n\n\n\n This guide draws together the core ideas found across the classic methodological literature\u2014including Baron and Kenny\u2019s (1986) foundational paper, James and Brett\u2019s (1984) work on mediators versus moderators, MacKinnon and colleagues\u2019 work on mediation analysis, Hayes\u2019 modern approach to conditional process analysis, and various applied teaching resources\u2014to give a single, thorough explanation of mediation, moderation, the difference between them, how each is tested, and how they can be combined.<\/p>\n\n\n\n A mediator<\/strong> (or mediating variable, M) is a variable that lies between the independent variable and the dependent variable in a causal chain. It explains the process or mechanism by which the IV influences the DV. Symbolically, the relationship looks like:<\/p>\n\n\n\n X \u2192 M \u2192 Y<\/p>\n\n\n\n The independent variable (X) causes the mediator (M), and the mediator, in turn, causes the dependent variable (Y). The mediator is sometimes called an intervening variable because it intervenes between the cause and the effect.<\/p>\n\n\n\n stress (X) leads to illness (Y) because stress reduces immune function (M). Here, immune function is the mediator\u2014it is the \u201chow\u201d that connects stress to illness. Without the mediator, we would only know that stress and illness are correlated; with the mediator, we understand the biological pathway.<\/p>\n\n\n\n Mediators answer the question: \u201cBy what process or mechanism does X affect Y?\u201d<\/strong><\/p>\n\n\n\n A moderator<\/strong> (Z or W) is a variable that affects the strength or direction of the relationship between the independent and dependent variables. Rather than sitting on the causal pathway between X and Y, a moderator interacts with X to produce different effects on Y depending on the level of the moderator. This is statistically represented as an interaction effect:<\/p>\n\n\n\n Y = b1(X) + b2(Z) + b3(X\u00d7Z) + error<\/p>\n\n\n\n If the interaction term (X\u00d7Z) is statistically significant, the moderator is said to moderate the relationship between X and Y. A moderator essentially answers: \u201cFor whom, under what circumstances, or in what context does the relationship between X and Y hold, strengthen, weaken, or reverse?\u201d<\/strong><\/p>\n\n\n\n Social support (X) reduces stress (Y), but this effect is stronger for people who are more introverted (Z) than for those who are more extroverted. Personality type does not explain how social support reduces stress. It simply changes how strongly that relationship operates depending on who you are.<\/p>\n\n\n\n The most widely repeated heuristic across the literature is this:<\/p>\n\n\n\n Mediators explain \u201chow\u201d or \u201cwhy\u201d an effect occurs (the mechanism); moderators explain \u201cwhen,\u201d \u201cfor whom,\u201d or \u201cunder what conditions\u201d an effect occurs or is stronger\/weaker (the boundary conditions).<\/em><\/p>\n\n\n\n Mediation model:<\/p>\n\n\n\n a b Path a: effect of X on M<\/p>\n\n\n\n Path b: effect of M on Y (controlling for X)<\/p>\n\n\n\n Path c: the total\/direct effect of X on Y<\/p>\n\n\n\n Moderation model:<\/p>\n\n\n\n Z Here Z is drawn as a variable that influences the arrow itself, the relationship between X and Y. Z is not a step on a path from X to Y. In regression terms, moderation is modeled by adding a multiplicative interaction term (X \u00d7 Z) as a predictor of Y.<\/p>\n\n\n\n A particularly useful analogy (popularized in introductory statistics teaching) imagines a light switch (X) and a light bulb (Y):<\/p>\n\n\n\nBasic Definitions<\/a><\/h2>\n\n\n\n
What Is a Mediator?<\/h3>\n\n\n\n
Example:<\/h4>\n\n\n\n
What Is a Moderator?<\/h3>\n\n\n\n
Example:<\/h4>\n\n\n\n
The Core Distinction in One Sentence<\/h3>\n\n\n\n
Visualizing the Difference Between Mediators and Moderators<\/a><\/h2>\n\n\n\n
Path Diagrams<\/h3>\n\n\n\n
X ——-> M ——-> Y
\\ ^
\\___________________\/
c (direct effect)<\/p>\n\n\n\n
|
v
X ——–> Y<\/p>\n\n\n\nThe Light Switch Analogy<\/h3>\n\n\n\n
Different Statistical Signatures<\/h3>\n\n\n\n