Thomas Yeo

bttyeo.bsky.social

21 May 2026 04:32

@sbe.bsky.social @ndosenbach.bsky.social @kordinglab.bsky.social @danilobzdok.bsky.social 14/14

Thomas Yeo

bttyeo.bsky.social

21 May 2026 04:32

Thank you to all the co-authors who made this possible @tianfang.bsky.social @csabaorban.bsky.social @anlijuncn.bsky.social @sinamansourl.bsky.social @leonooi.bsky.social @yapeixie.bsky.social @lindenmp.bsky.social @elvisha.bsky.social @sidchop.bsky.social @andrewzalesky.bsky.social 13/N

Thomas Yeo

bttyeo.bsky.social

21 May 2026 04:32

Our meta-analysis also revealed broader reporting gaps: 54% of studies didn't clearly apply a statistical test — either none was done, or reporting was too vague. We recommend reporting: folds, repetitions, the specific test, and how fold-level statistics are aggregated. 12/N

Thomas Yeo

bttyeo.bsky.social

21 May 2026 04:31

So we don't claim 97% of studies contain false positives. Rather, 97% employed statistical tests that don't achieve the stated error control. Prior findings warrant reassessment, not wholesale dismissal. 11/N

Thomas Yeo

bttyeo.bsky.social

21 May 2026 04:31

Important caveat: an invalid test doesn't imply a false positive. Sufficiently large effects may remain significant under more conservative procedures, and any given model comparison may be just one component of a broader study whose conclusions hold. 10/N