JASP - No BF factor for Bayesian correlation

Izymil

Hi!

I conducted Bayesian correlation analysis in the latest version of JASP (0.16), but I cannot obtain the Bayes factor. I suppose it might be extremely high, but still, with scientific notation, it should be displayable somehow. The output of correlation in Jamovi works fine.

See attached file.

https://drive.google.com/file/d/1CMlzI-A8c09-l0_raQbkDNwMFXdMG41D/view?usp=sharing

Does anyone have any clue what might have caused it and how to fix this?

Cheers!,

Kamil I.

EJ

1. I did not know Jamovi offered a Bayesian Kendall's tau?! I know that at some point they copied the Bayesian analyses from JASP, but Kendall's tau is a recent addition -- can you confirm? If Jamovi does have it, it should fail for the same reason that JASP fails (because it is essentially our code)
2. I suspect that the correlation is so extreme that Kendall's tau is somehow not computable. But clearly there should be some sort of error message instead of an empty cell. I'll let Johnny know (he derived and implemented this analysis).

Cheers,

E.J.

Izymil

Hi EJ!

It has Kendall's and instead of an empty cell it shows infinity

JohnnyB

Hi Izymyl,

Thanks for getting back to us, showing the infinity seems like a good solution to avoid confusion.

Cheers,

Johnny