SPSS command: SPSS BAYES REPEATED
Bayesian repeated-measures ANOVA in your browser
Bayesian repeated-measures ANOVA tests whether the mean of the same people changes across two or more conditions while separating stable differences between people from the condition effect.
Free · No account · Files stay on your device · Updated 18 October 2026
What it tells you
Crosstabs runs the SPSS BAYES REPEATED procedure on your device with the Nathoo–Masson BIC approximation or Rouder's mixed-design Bayes factor, the latter integrated by deterministic quadrature.
- Whether the within-subject factor changes the mean (BF10) or leaves it unchanged (BF01).
- The classical F, df1, df2 and significance, plus ΔBIC10 when the BIC method is used.
- A posterior mean and credible interval for the mean of each measure, with its observed standard deviation.
When to use it
Use it when
- The same participants measured under three or more conditions or time points, stored as one column per measure.
- Usability or within-subject experiments where you need evidence that versions perform alike.
- Pilot studies where a BIC-based Bayes factor gives a quick, prior-light summary.
Use something else when
- A between-subjects factor in addition to the repeated measures: use GLM Repeated Measures or linear mixed models.
- Subjects with missing time points you want to keep: the panel excludes incomplete subjects; consider linear mixed models.
- Ordinal ratings across conditions: use the Friedman test.
How to run it in Crosstabs
- 01
Open your data
Open a wide file with one row per subject and one column per condition, then switch to Statistics mode.
- 02
Choose the measures
In the Bayesian statistics group choose Bayesian repeated-measures ANOVA and tick at least two numeric variables under Repeated measures (at least two, in order).
- 03
Pick the Bayes factor
Choose BIC approximation (Nathoo & Masson) or Rouder's mixed design.
- 04
Run
Set Credible interval (%), declare any workspace weight (a frequency weight counts each subject that many times) and press Run Bayesian repeated-measures ANOVA. Subjects with any missing measure are excluded.
Worked example
Computed by the Crosstabs engine from the synthetic data below when this page was built — not typed in.
Data
16 synthetic participants rate workload (0–100) after using each of three interface versions, v1 to v3. A seeded per-person level is shared across the three ratings, so the measures are correlated within subjects.
| Method | N | Measures | F | df1 | df2 | Sig. | ΔBIC10 | BF10 | BF01 |
|---|---|---|---|---|---|---|---|---|---|
| BIC approximation | 16 | 3 | 8.618 | 2 | 30 | .001 | −8.928 | 86.847 | 0.012 |
| Rouder's mixed design | 16 | 3 | 8.618 | 2 | 30 | .001 | 24.490 | 0.041 |
The panel shows one method per run; both runs are listed here for comparison.
| Measure | Std. Deviation | Posterior mean | Posterior variance | 95% CrI lower | 95% CrI upper |
|---|---|---|---|---|---|
| v1 | 10.25 | 49.49 | 5.4332 | 44.93 | 54.06 |
| v2 | 9.01 | 46.44 | 5.4332 | 41.88 | 51.01 |
| v3 | 9.59 | 43.59 | 5.4332 | 39.02 | 48.16 |
Large-sample normal approximation (Nathoo & Masson 2016); the same posterior variance for every measure.
How to read it
The within-subject F(2, 30) = 8.62, p = .001. The BIC approximation gives very strong evidence for the alternative (BF10 = 86.847), and Rouder's mixed-design factor gives strong evidence for the alternative (BF10 = 24.490). The two methods use different priors, so their values differ; report which one you used. Mean workload is 49.5 for v1, 46.4 for v2 and 43.6 for v3.
Output tables explained
- Bayes Factor Repeated Measures ANOVA
- N subjects, number of measures, F, df1, df2 and significance, ΔBIC10 for the BIC method, BF10 and BF01. A footnote states which approximation or prior was used.
- Posterior Distribution Characterization for Means
- For each measure: observed SD, posterior mode, mean and variance, and the credible interval from a large-sample normal approximation; the posterior variance is the same for every measure.
Assumptions and notes
- Subjects are independent; the measures share the same scale and are listed in their natural order.
- Errors are roughly normal; the BIC method and the posterior use large-sample approximations, so treat results from very few subjects cautiously.
- No sphericity correction is applied to the F test, so read the classical significance with care when variances of the differences are unequal.
How the results are validated
The sums of squares and ΔBIC10 match the Python reference to 1e-11 (ΔBIC10 also with a frequency weight), and the posterior means and variance to 1e-6 and 1e-12. Rouder's factor, weighted and unweighted, matches an explicit-matrix quadrature to a relative 1e-6 and R BayesFactor 0.9.12.4.8 lmBF within four times BayesFactor's own Monte Carlo error. Validation evidence
Questions
BIC or Rouder: which Bayes factor should I use?
The BIC approximation is quick and implies a unit-information prior. Rouder's mixed design puts explicit g-priors on conditions and subjects, matching BayesFactor's lmBF with a random subject effect and both scales set to 1. The two can differ by a factor of several, so report the method.
Will Rouder's factor match SPSS?
SPSS estimates the integral with 30 000 Monte Carlo draws, so its value varies slightly from run to run. Crosstabs integrates the same expression by adaptive quadrature, giving a deterministic value that should lie within SPSS's sampling error.
How are frequency weights handled?
A frequency weight counts each subject that many times in both methods, the same as expanding the file. SPSS ignores frequency weights for Rouder's method, so weighted results differ there.
Does it give pairwise comparisons between conditions?
No. Compare the credible intervals of the measure means, or run Bayesian paired samples on the pairs you care about.
Related procedures
- Repeated Measures ANOVAGLM WSFACTOR
- Bayesian Paired SamplesBAYES RELATED
- Bayesian One-Way ANOVABAYES ANOVA
- Friedman TestNPAR TESTS /FRIEDMAN
- Linear Mixed ModelsMIXED
- Free SPSS alternative: full coverage list
Run Bayesian Repeated-Measures ANOVA on your own data
Open a .sav, CSV or Excel file. Everything runs in your browser; the file never leaves your device.
SPSS is a trademark of IBM Corp. Crosstabs is not affiliated with or endorsed by IBM.