[Bug]: Bernoulli prior for Bayesian Regression treats some variables differently

[HeikoSchuett]

JASP Version

0.19.0

Commit ID

No response

JASP Module

Regression

What analysis are you seeing the problem on?

No response

What OS are you seeing the problem on?

macOS Silicon

Bug Description

I am running a Bayesian linear regression example with generated data and am getting strange behaviour for the Bernoulli model prior. Namely I am getting a higher prior inclusion probability for one of my variables, see screenshot below.

-------- Application Info --------
JASP Version: JASP 0.19 (Apple Silicon)
Build Branch: development
Build Date: Jul 13 2024 09:41:32 (Netherlands)
Last Commit: 8b929b98f058f23dd8c8cf3e891be5a77b2ebc0b

Expected Behaviour

use 0.2 inclusion probability for all variables. At least there should not be different priors for different variables.

Steps to Reproduce

just run the analysis

Log (if any)

data_lin_reg_bayes.jasp.zip

More Debug Information

-------- Basic Info --------
Operating System: macOS 15.1
Product Version: 15.1
Kernel Type: darwin
Kernel Version: 24.1.0
Architecture: arm64
Install Path: /Applications/JASP.app/Contents/MacOS
Platfotm Name: cocoa
System Local: en_LU

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.

[vandenman]

Thanks for reporting this! I had a look and it could be that this is an issue arises from the R package BAS rather than JASP. I've filled an issue at the GitHub repository where BAS is maintained (merliseclyde/BAS#87) and will keep you posted as we find out more.

[JohnnyDoorn]

I see that the issue has been fixed in BAS, so I think we can close this issue since it'll be in the upcoming JASP release (probably end of the month):