How is Effect Modification Different from Confounding?

While both effect modification and confounding involve a third variable, their roles differ significantly. Confounding occurs when a third variable is associated with both the exposure and the outcome, potentially leading to a spurious association. Effect modification, on the other hand, describes a scenario where the third variable changes the strength or direction of the association between the exposure and the outcome. Unlike confounding, effect modification is not a bias and should be reported and interpreted, rather than controlled or adjusted for.