No of samples Bayesian paired sample wilcoxon signed-rank

loctp

Hi all,

I'm sorry for this very basic question but I'm starting to do Bayesian analysis on non-normal data and I don't understand what the ‘No. samples’ corresponds to when I select the ‘Wilcoxon signed-rank’ option for a paired-samples Bayesian t-test in JASP?

Thanks in advance !

patc3

When you wonder about an analysis input, you can look up the documentation for this specific analysis by clicking the little "i" (info) button:

In this case the documentation indicates the following:

So the number of samples (default 1000) indicates the number of MCMC iterations. I assume the default is enough, though you need to monitor convergence. According to this same documentation, the output will provide you with the Rhat statistic to monitor convergence:

If Rhat is greater than 1 (say greater than 1.05), then you could try to increase the number of iterations to achieve convergence.

All this said, I agree 100% with you that it is not clear at all that this "No. samples" refers to the number of MCMC iterations.

loctp

Thank you so much for your feedback ! I was not aware of the information button haha