non-parametric Bayes test
Hello everyone
I am new to this forum and JASP . I want to know if I can do non-parametric bayes test(alternative of wilcoxon test) in JASP? I have two ML models which produce 5 Mean Absolute Errors for both model each. I want to evaluate statistically if the two models have mean values not equivalent (if one model has better Mean Absolute Errors values than the others).
There are options in JASP about the bayes t-tests but I guess it require our data to be normally distributed.
Thanks
Neha
Comments
Hi Neha,
Yes, JASP offers Bayesian rank-based alternatives. They are under the regular t-test menus. The help file will point to the relevant references.
Cheers,
E.J.