A researcher runs a 2×2 between-subjects ANOVA on stress scores with factors: Caffeine (None vs High) and Study Method (Cramming vs Spaced). SPSS ‘Tests of Between-Subjects Effects’ shows:
Caffeine: F = 4.20, p = .044 Study Method: F = 0.31, p = .579 Caffeine × Study Method: F = 7.10, p = .009
Which interpretation is BEST?
Because the interaction is significant, we should ignore all main effects and immediately interpret the simple effects (how study method works within each caffeine level, or vice versa).
Caffeine has an effect on stress regardless of study method, because its main effect is significant.
Study method has no effect on stress in any condition, because its main effect is not significant.
The significant interaction means the effect of study method on stress depends on caffeine level (or equivalently, the effect of caffeine depends on study method), so we should probe simple effects even if a main effect is significant.