[Bug]:

[JohanACHARD]

JASP Version

0.18.3

Commit ID

No response

JASP Module

ANOVA

What analysis are you seeing the problem on?

Bayesian RM-ANOVA

What OS are you seeing the problem on?

Windows 11

Bug Description

I found myself again with big differences between Bayesian and frequentist with an RM-ANOVA.

But after a pause for reflection, when I turned JASP back on, the model converged totally differently... (version 0.18.3 JASP). I tried again after a pause, and the results diverged but remained stable this time (cf screenshots). I tried with other computer with JASP 0.16.4 (screen 4).

What do you think? A stability problem ?

Expected Behaviour

I expect same (or similar) result when I reopening and perfom Bayes RM-Anova

Git-hub link with data: https://github.com/JohanACHARD/jasp_RM_anova_difference_freq_bayesian/blob/f27e10fc869be6a0762b8d25d55b518a6b53ecbd/meanMT.jasp

Steps to Reproduce

No... now JASP compute similar result.... intriguing when I re try the analysis a new time

Log (if any)

https://github.com/JohanACHARD/jasp_RM_anova_difference_freq_bayesian/blob/f27e10fc869be6a0762b8d25d55b518a6b53ecbd/meanMT.jasp

mean_MT.csv

Final Checklist

• I have included a screenshot showcasing the issue, if possible.
• I have included a JASP file (zipped) or data file that causes the crash/bug, if applicable.
• I have accurately described the bug, and steps to reproduce it.

[JohnnyDoorn]

Hi @JohanACHARD ,

Thanks for reporting this. I discussed with @vandenman, and those first results do seem to be some anomaly in the underlying sampling procedure. These cases are hard to analyze, because they are not reproducible without the seed. The RM ANOVA has an option (under "Additional options") to specify the seed - so that you can reliably reproduce the exact same results. This could be useful in the future, for both your own result replications and for bug reporting. I am sorry that we cannot help you with this specific case, besides the advice to always run a sampling based analysis a couple of times to check if the results have converged.

Kind regards
Johnny

[tomtomme]

@JohanACHARD
Was Johnnys answer sufficient for your, so that we can close this one?

[JohanACHARD]

Hello team,
Thanks for your help, I tried to reproduce this problem using your recommendations but I couldn't do it.

Thanks, that's it for me!