Moderation
Statistics & analysis
A relationship in which a third variable changes the strength or direction of the link between two other variables.
Moderation asks when, or for whom, an effect holds. A moderator W changes the strength or direction of the relationship between X and Y, and is tested by adding an interaction term (X multiplied by W) to the regression alongside both main effects.
Mean-centre or standardise the predictors before creating the product term. This reduces non-essential multicollinearity between the interaction and its components, and makes the main effects interpretable at the average value of the moderator. The coefficient on the interaction term is the evidence for moderation.
A significant interaction is the beginning of the analysis, not the end. Probe it with simple-slopes analysis, reporting the effect of X on Y at low, mean and high values of the moderator (commonly plus and minus one standard deviation), and plot it. The Johnson-Neyman technique identifies the exact range of the moderator over which the effect is significant.
Where it's used
- Testing whether an effect is stronger for one group than another
- Boundary-condition contributions to theory
Software used
Related concepts
Guides that use this concept
All guidesNot sure where to start?
Book a free 15-minute consult. We'll map your next three steps — no obligation.