The only non-significant result, psychology journals will publish. We found no evidence of publication bias. LOL

When meta-scientists wanted to study whether science works, they put psychology in the control group.

The insanity of meta-analyses of pathological literatures. Four hundred individually “successful” studies should not culminate in uncertainty about whether the foundational effect exists at all.

Somebody needs to do the dirty work to keep a place clean and functioning. Call me #TheCustodian of psychological science. :)

There are still believers in priming in 2026. 😮 replicationindex.com/2026/08/01/g...

Nobody ever called psychology #TemuScience 😷

How powerful are bias tests? Turns out not very much, even with k = 300 studies. A better way to deal with publication bias is to estimate it with confidence intervals and focus on the upper bound (not the lower bound, lb > 0, there is bias, p < .05). arxiv.org/abs/2607.19626

Beyond Publication-Bias Detection: Estimating Bias under Uncertainty in Heterogeneous Literatures

Meta-analysts routinely test for publication bias, but a nonsignificant test is inconclusive because it may be a Type 2 error. In a large factorial simulation, I show that publication-bias tests have ...

arxiv.org

Most corrupt administration in American history!

"Hi Lindsay, damn it is hot here. I hope we get used to it in eternity."

leftylady.bsky.social@leftylady.bsky.social · 3w ago

Curious to hear Mitch McConnell's comments on the passing of Lindsey Graham.

#RMA #Metafor #MetaAnalysis by tradition, random effects models in meta-analysis assume a normal distribution of population effect sizes. This is a problem. The assumption is neither plausible nor necessary. Modern mixture models solve the problem. replicationindex.com/2026/07/12/w...

Xitter's community notes are working overtime calling out people trying to claim Mamdani's attempts to conserve power is socialist. This one might be my favorite.

I'm working on a new Metacheck module: Datacheck. Repocheck already lists all files on Github, Zenodo, OSF, Researchbox. Datacheck will download them and.... which automated checks would you like for datafiles? Can be general (e.g., flag outliers), or file specific (e.g., excel files). Let me know 👇

Testing the assumptions of step-function selection models with z-curve. replicationindex.com/2026/06/26/t...

Pathological Meta-Science: All pain, no gain.

New preprint: There is only one correct analysis. osf.io/preprints/ps... with @sajedehra.bsky.social and @mntunc.bsky.social We take a critical look at recent proposals for multiverse or many analyst procedures, and strongly argue that epistemic uncertainty should be reduced, not “embraced”. >

Statistical Power is nearly 100 years old, but the concept is still misunderstood, especially in applications that examine how much power researchers really had to get their significant results. open.lnu.se/index.php/me...

Why most meta-analysis are false. #MetaScience replicationindex.com/2026/06/18/t...

"You Cannot Step Into the Same River Twice", but you can "step into it" again and again. A snarky post on McShane's rhetorical simulation studies. replicationindex.com/2026/06/13/s...

A Cochrane system for checking whether any of its thousands of published reviews include retracted studies has already turned up dozens of reviews that do and will now get closer scrutiny to ensure their results and recommendations hold up.

New system for flagging retracted papers finds scores of them in Cochrane reviews

Cochrane has implemented a new system for checking whether any of its thousands of published reviews include retracted studies in their analyses, the organization announced today. The effort alread…

retractionwatch.com

Maybe AI is not really smart, intelligent, etc. what ever that means. Maybe we should give it an IQ test (LOL). However, it is much more rational than humans. Human intelligence is often wasted and compromised by self-deception willful ignorance, and cognitive biases that won KT a Nobel.

My PhD Sajedeh Rasti has published a World View piece in Nature today: "Don’t compete, collaborate: why collective funding applications are the future" www.nature.com/articles/d41... She discusses how she successfully brought together a team of meta-scientists to collaboratively apply for funding.

Don’t compete, collaborate: why collective funding applications are the future

Scientists with disparate expertise writing grants together can identify knowledge gaps and drive progress — but systems must change to incentivize them.

nature.com

Every math teacher when students say they will never need it in real life. Are you human?

What are standardized standard errors and why are they useful but ignored? replicationindex.com/2026/06/08/h...

Which kind of #MetaScientist are you? A. Basic / metaphysical - talk a lot about problems B. Applied MetaScientist - actually solve problems. Academia often rewards the talkers because they have a lot of time talking without doing. :)

Cancel culture much? Fuck oppression by the formerly oppressed. www.theguardian.com/us-news/2026...

‘We call it the P-word’: Chicago professor suspended after assignment mentions Palestinians

School of the Art Institute of Chicago professor put under investigation after a student complained about a case study

theguardian.com

Explicit Racism is killing Project Implicit and the foolish idea that racial biases are unconscious. replicationindex.com/2026/06/04/w...

#PathologicalScience The unbearable automaticity of experimental social psychology. 1. Run an underpowered study. 2. P-hack 2a. still not significant, discard. 2b. significant, tell a story. repeat, repeat, repeat.

If post-hoc power calculations are bad, why are post-hoc effective calculations good? replicationindex.com/2026/05/31/t...

Science is self-correcting, but it can take time. replicationindex.com/2026/05/28/w...