What is an interaction effect?

What is an interaction effect?

• A. Significant effect where the effect of one factor is different across levels of another factor
• B. A main effect averaged over levels of other factors
• C. A random error term
• D. A post-hoc comparison result

Answer: (A)

Interaction occurs when the effect of one factor on the outcome changes depending on the level of another factor. In a two-factor design, this means the influence of one variable isn’t the same across all categories of the other variable—lines in a graph might bend or cross, signaling that the factors don’t work independently. The statement described captures this idea: a significant effect where the effect of one factor differs across levels of another factor. That is exactly what an interaction is. By contrast, a main effect is the average impact of one factor across all levels of the other factor, which ignores potential interactions. A random error term is simply noise in measurement, not a meaningful interaction. A post-hoc comparison result comes from follow-up tests after finding significant effects, not from defining what an interaction is.

Interaction occurs when the effect of one factor on the outcome changes depending on the level of another factor. In a two-factor design, this means the influence of one variable isn’t the same across all categories of the other variable—lines in a graph might bend or cross, signaling that the factors don’t work independently.

The statement described captures this idea: a significant effect where the effect of one factor differs across levels of another factor. That is exactly what an interaction is.

By contrast, a main effect is the average impact of one factor across all levels of the other factor, which ignores potential interactions. A random error term is simply noise in measurement, not a meaningful interaction. A post-hoc comparison result comes from follow-up tests after finding significant effects, not from defining what an interaction is.