Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Supported by

Bayesian test assumptions

Hello everyone,
I'm really new to your program and Bayesian analysis and have been wondering: are assumptions pertaining to the data (i.e. normality, homogeneity of variances) the same for the Bayesian versions of t-tests, ANOVA etc.? If so, are they in some way more or less robust against violations of these assumptions? Thank you for your reply.


  • Yes and yes. Also, we are working to include nonparametric Bayesian tests.


  • Thanks for your reply E.J., that sounds like an awesome addition to your program (I really like using it already). Just to clarify: So you mean to say that Bayesian analysis is more robust than the frequentist counterpart? Could you (or anyone else) point me to a publication that demonstrates this for the analyses used in JASP? Thank you!

  • Oh, no, the degree of robustness has not been verified. But both classical and Bayesian tests rely on the same basic features of the data, so it stands to reason that they will be similarly affected.

  • E.J., in the meantime, what would be my options if my data violates both assumptions?

    Any time-frame for those non-parametric Bayesian tests?


  • Well the tests are ready, we need to finish up the paper and include them. It is a big of work, but when we have the paper submitted you are welcome to the code so you could run it before it's implemented in JASP

Sign In or Register to comment.