Reporting Bayesian Mann-Whitney U test

Martyna

Hello,

First, let me thank everyone who worked on developing JASP - I just started using it and honestly don't think I'll ever 'go back'.

Second, I am hoping to double check with more knowledgeable members on what exactly to report when performing nonparametric Bayesian estimations. I understand that from JASP output you need to clearly state what the hypotheses are, report BF 01 or BF10, error %, median and the credible interval. But when performing Mann-Whitney, when you don't have the error% - is it enough to simply report BF01, median and credible interval? Should you report the number of samples/ iterations?

Thank you for any help with this.

Martyna

EJ

Hi Martyna

Yes, you can report that. Or you can rerun the procedure several times and get an idea of the % that way.

And thanks for your kind words!

Cheers,

E.J.

Martyna

Thank you!