Three-Way Mixed ANOVA BWW (1 between × 2 within)
Coming soonRepeat / Within
This test is implemented and is currently going through StatMinds’ production verification: every statistic is independently checked against a trusted reference (scipy / R), locked with regression tests, and the screen is exercised across assumption-met/violated and significant/non-significant scenarios before it opens up.
See what’s live nowSplit-plot factorial ANOVA with one between-subjects factor (e.g., Treatment) and two within-subjects factors (e.g., Time × Condition).
Each participant is in ONE between-cell AND measured at every combination of the two within-factors. Tests SEVEN F-tests — three main effects, three two-way interactions, and the three-way interaction — with the correct split-plot error terms (subjects-within-groups for between effects, within-subject residuals for within-involving effects). Auto-applies Greenhouse-Geisser / Huynh-Feldt sphericity corrections per within-factor.
Worked example
One between-subjects factor and two within factors — how do they combine?
A three-way mixed ANOVA (between-within-within) tests one grouping factor and two repeated factors plus their interactions.
A significant three-way interaction, F(2, 76) = 4.1, p = .021, showed the two within-subject effects combined differently across groups.
A three-way mixed (BWW) ANOVA showed a significant three-way interaction, F(2, 76) = 4.1, p = .021.
When to use it
- Treatment × Time × Condition (longitudinal experimental)60 participants are randomised to (Drug A / Drug B / Placebo) (between).
- Group × Session × Phase (clinical / educational longitudinal)40 students in (Intervention vs Control) (between) complete tests at (Sessions: 1 / 2 / 3) × (Phases: Pre-task / Mid-task / Post-task) (within × within).
When NOT to — use instead
- All three factors between-subjectsEach subject in only one cell of all three factors → use a true three-way ANOVA. → Three-Way ANOVA (A \u00d7 B \u00d7 C)
- 2 between × 1 within (BBW)Two between-subjects factors → use the BBW variant of three-way mixed ANOVA. → Three-Way Mixed ANOVA BBW (2 between \u00d7 1 within)
- All three factors within-subjectsEach subject measured at every level of all three factors → use three-way RM-ANOVA. → Three-Way RM-ANOVA (3 within factors)
- Missing within-cells (some subjects miss measurements)BWW requires complete cases per subject. → Linear Mixed Effects (LMM) \u2014 clustered / repeated data (ICC, BLUPs)
Hypotheses
Parameter tested: set of cell means {μ_{ajk}} and their marginal / interaction decomposition
Assumptions (and what to do if they fail)
Check: Inspect per-cell boxplots; flag points beyond the whiskers.
If violated: Outliers can inflate within-cell variance and distort the F-tests.
Check: Shapiro-Wilk on residuals.
If violated: Severely non-normal residuals threaten F-test Type I error rate.
Check: Levene's p > .05 at every within-cell → homogeneity met.
If violated: Unequal between-group variances inflate F-test Type I error.
Check: Mauchly's test per within family; non-significant → sphericity met for that family.
If violated: Violated sphericity inflates the F-test's Type I error rate for the affected family.
Ready to run a Three-Way Mixed ANOVA BWW (1 between × 2 within) on your own data?
Guided setup, automatic assumption checks, effect sizes, figures and an APA write-up.
Run this test →