Daniel Lakens
New paper by @tomhardwicke and colleagues looks at journal recommendations for statistical reporting guidelines. https://www.tandfonline.com/doi/full/10.1080/00031305.2022.2143897 Good to see recommendations to use p-values are still going strong - the often weak criticisms against them seem not to convince journal editors.
Statistical Guidance to Authors at Top-Ranked Journals across Scientific Disciplines

Scientific journals may counter the misuse, misreporting, and misinterpretation of statistics by providing guidance to authors. We described the nature and prevalence of statistical guidance at 15 ...

Taylor & Francis
Dec 1, 2022 at 8:58pmMastodon for Android
Show threadDavid Williamson ShafferDec 1, 2022

@lakens @tomhardwicke

Are there recommendations for reporting of effect sizes?

Show threadDaniel LakensDec 2, 2022
@dwshaffer @tomhardwicke yes, often as well, see the paper.
Show threadDavid Williamson ShafferDec 5, 2022

@lakens @tomhardwicke
Excellent, although one thing that has always bothered me, honestly, is that N, p, and d (or r, if you prefer) are basically determined by a single equation. So if you know two, you know the third.

This is actually consistent with Fischer's original interpretation of the p value (which he regarded as more or less parallel to effect size). So calling for not using p values is silly, and reporting p without d or r is redundant, but easier for the reader!

Show threadFelipe L Fontana VieiraDec 2, 2022
@lakens could you please elaborate on the “weak” part?
Show threadDaniel LakensDec 2, 2022
@felipelfv Sure, most criticism on NHST is weak - Mayo calls these 'howlers'. I have never seen a convincing criticism - have you?