Mixed ANOVA

Mixed ANOVA is an analysis of variance that combines a between-subjects factor and a within-subjects (repeated-measures) factor to test their main effects and interaction on a continuous outcome.

혼합 분산분석(Mixed ANOVA)은 집단 간 요인과 반복측정(집단 내) 요인을 함께 포함하여 두 요인의 주효과와 상호작용이 연속형 결과변수에 미치는 영향을 검정하는 분산분석이다.

How do I compare groups on an outcome measured repeatedly over time or conditions?

반복 측정한 결과변수에서 집단 간 변화 양상이 다른지 어떻게 비교하나요?

A mixed (split-plot) ANOVA tests the effects of a between-subjects grouping factor, a within-subjects repeated factor, and their interaction on a continuous outcome. It fits pre-post or multi-wave designs where the key question is whether change over time differs between groups.

When to use Mixed ANOVA

• Treatment and control groups measured at pretest, posttest, and follow-up
• Each participant rated under several conditions, with participants split into groups
• Testing whether the time-by-group interaction is significant

When to use something else

• There are missing waves or unequally spaced measurements: use a linear mixed-effects model
• There is no between-subjects factor: use repeated-measures ANOVA

Assumptions and requirements

• Approximately normal residuals in each cell
• Homogeneity of variances across groups at each level of the within factor (Levene's test)
• Homogeneity of covariance matrices across groups (Box's M)
• Sphericity for the within factor (Mauchly's test, with Greenhouse-Geisser correction if violated)

What the analysis reports

• F tests for the between factor, within factor, and interaction
• Partial eta squared for each effect
• Sphericity test and corrected degrees of freedom
• Cell means and an interaction plot
• Simple-effects or post hoc comparisons

How to report it

A mixed ANOVA showed a significant [group] × [time] interaction, F([df1], [df2]) = [value], p = [value], ηp² = [value].

Related methods

• Repeated Measures ANOVA
• Linear Mixed-Effects Model
• Two-Way ANOVA

Run Mixed ANOVA on Quorant