When one variable's effect on the outcome depends on the level of another variable.
An interaction effect occurs when the influence of one independent variable (e.g. treatment) on the outcome changes depending on the level of a second variable (e.g. age, sex, baseline severity). It is typically identified via a significant interaction term in ANOVA or regression models, and its presence means main effects alone can be misleading or must be interpreted with caution. Detecting true interaction effects usually requires larger samples than detecting main effects, so null findings in underpowered studies should be interpreted cautiously.
This guide was auto-drafted and is pending editorial review.