Open mmpeng9 opened 2 months ago
Hi @mmpeng9
Thanks for opening the issue and apologies about the late reply. Can you please share your experiments here? When doing so, can you please make sure that you are using the latest available fork of Pingouin (which includes https://github.com/raphaelvallat/pingouin/pull/424) and not the current stable release (0.5.4).
Thanks Raphael
Hi there,
I found that the computation of rank-biserial correlation (RBC) is problematic. This is related to https://github.com/raphaelvallat/pingouin/issues/417 and https://github.com/raphaelvallat/pingouin/pull/424.
According to the cited paper, there are three ways to compute RBC. It seems you adopted the third method based on Hans Wendt (1972): r=1 – (2U)/ (n1 * n2). From the paper, U is the smaller number between U1 and U2:
According to SciPy,
It seems that the returned
U
is not the smaller one inU1
andU2
. And, it will result in a RBC value that is negative (according to the paper, this should always be positive). This is also demonstrated in my experiments.Thanks!