effectsize
effectsize copied to clipboard
easystats
Improve CI coverage for non-parametric CLES and rank differences
Open a new issue / pr to discuss improving the CI covered for all of these methods - since they are all monotonic transformations of one another, it makes sense to me they can be improved together.
Originally posted by @mattansb in https://github.com/easystats/effectsize/issues/496#issuecomment-1256971597
Also related to #479
@arcaldwell49 feel free to contribute here if you'd like, there is not rush or pressure (:
The fisher z transformation applies to Pearson r and approximately to phi — it will undercover for any rank correlation or continuity correction like biserial, tetrachoric, or polychoric unless the n is adjusted downward
metafor::escalc has sampling error formulas for some of these I believe
Very good points @bwiernik . Also, I noticed an error in my code where I ran some simulations earlier this month.... The issue may be mute, but let me re-run the simulations to check.
