A student runs a moderation analysis in hierarchical regression to test whether exercise frequency (W) moderates the effect of caffeine intake (X) on anxiety score (Y). Step 1 includes X and W; Step 2 adds the interaction X×W. SPSS Model Summary shows: Step 1 R² = .18; Step 2 R² = .21; Change Statistics: ΔR² = .03, F-change(1, 196) = 7.20, p = .008. What is the most appropriate conclusion?
Exercise frequency moderates the caffeine–anxiety relationship because the change in R² after adding the interaction is significant.
Caffeine intake mediates the relationship between exercise and anxiety because R² increased.
The interaction should not be included because R² is still relatively small.
Moderation cannot be tested using hierarchical regression; only ANOVA can test interactions.