1/ The US Government has quietly removed a memorial to Black soldiers who died in World War II from the Netherlands American Cemetery in Margraten, South Limburg. The move follows a complaint from the right-wing Heritage Foundation to the American Battle Monuments Commission. ⬇️

They are not the same.

How about distinguishing positive and negative impact. IAT - huge citation impact, huge negative impact. 😀

Panorama made a stupid editing mistake. But compare this to the deliberate and systemic attempt, sustained across years, to ensure that BBC output aligns ever more closely to the demands of economic power. Against your straw of bias, I raise you a haystack.

Governor Gavin Newsom on Twitter a few moments ago Regarding Democrat Senators who caved Agree with you Governor 💯

Republicans were overwhelmingly getting blamed for the shutdown. There was no reason for the Democrats to cave to Trump on healthcare, except for the fact that they are wholly owned and operated by the health insurance lobby.

Amazon charge up to 40% rent to use their services. 80% of all profits are directly removed from the UK and into tax havens. Depletion of money in the UK economy. Sell a great invention—they will copy it and sell it for themselves. They steal & sell PRIVATE data. This is a PREDATORY COMPANY

So, none of these falsely suggest that the p-curve null-hypothesis (all significant results are false positives) is falsely rejected more than 5% of the time with alpha = .05?

A lot is going on at @forrt.bsky.social! #OER, #metascience, #socialjustice, and much more. If you want to get a glimpse of some of the projects, check out this beautiful booklet at forrt.org/booklet created by @irissmal.bsky.social

One of the jobs of #MetaScience is or should be to distinguish real science from pseudo-science; and the job of meta-scientists is to call out pseudo-science when they see it. Academic != scientist. replicationindex.com/2025/11/06/t...

some things are real: cognitive biases like confirmation and consensus bias, the problem is that they also influenced the work of the scientists who studied them. MetaBias.

A rumor is that Festinger asked his graduate students to do the study and Carlsmith was the lucky one who got the significant results, or was he?

There’s growing evidence that something was going seriously wrong in the classic early work on cognitive dissonance Latest revelation: The story in When Prophecy Fails seems to have been fabricated in the most egregious way But this is not the only one… onlinelibrary.wiley.com/doi/abs/10.1...

The not so open Uri Simonsohn. Have you ever wondered why Datacolada has no peer review and no comment section. This is worse than old-school journals. What a jerk. replicationindex.com/2025/11/05/w...

This is pretty stupid. The right response is to use AI to fight AI crap. Train a model to desk reject AI crap and publish stuff that passes peer review. The added benefit would be that it also rejects human-made crap.

@aoc.bsky.social says of Mamdani victory: "A lot of people...are willing to talk about party unity when it serves them, but not party unity when it serves everybody." "We have a future to fight for, and we're either going to do that together or you're going to be left behind."

Mamdani: “NYC will remain a city of immigrants, a city built by immigrants, powered by immigrants and as of tonight, LED by an immigrant!” 🔥 🔥 🔥 🔥 🔥 🔥 🔥 🔥 🔥 🔥

This is so stupid. Using AI is not p-hacking. This article is just a good example of questionable reasoning, if we can even call it that. Or it is really an implicit attitude test. If you like it, you don't like AI. replicationindex.com/2025/10/04/i... replicationindex.com/2025/10/04/a...

This is nonsense. Paradigms fight for survival no matter what the paradigm is. The only way to fight this is a true culture of falsification. We do not have that anywhere in the social sciences.

We already have reverse p-hacking. It is called Bayes Factors. 😀🤪😁

Another big endorsement for Andrew Cuomo. And it only cost $959 million in tax breaks.

How much does it cost to publish a rant in Nature?

There is a problem with the claim that "it is all bias". That is not true simply because there is clear evidence of heterogeneity. So, some results are true others are not, but we do not know which are which. But that is to complex, so we just say, all are false, which is also false.

Lot's of confusion inderd. Just repoft the effect size estimate. Large N means small confidence interval. P values are useless. Outdated for 100 years.

Will read But by definition 100% of significant results are selected to be significant. Phrasing 65% were p-hacked? How would we know?

Wtf. Trying to get a job in Florida?

Sanity prayer: I will use p-curve when it works and I will not use it when it does not work, but I need the wisdom to know the difference. Alternative: use z-curve, it works, and it provides more information. This is not a tough choice. replicationindex.com/2025/10/28/h...

#SSHRC, CIHR and Michael Smith Health Research BC have awarded more than 2.7 million in funding through the Research-on-Research Joint Initiative. 14 teams will be researching meta-research - the study of how research itself is funded, practiced, and shared. sshrc-crsh.canada.ca/en/news/2025...

Meta researchers uncovered hidden patterns in LLM traces that reveal when the model’s reasoning is on point. Curious how AI self‑checks its thoughts? Dive into the findings and see the future of trustworthy models. #AI #ML #Research 🔗 aidailypost.com/news/meta-re...

I missed that once in a blue moon call for funding of meta-science in Canada. Trying again in the regular funding competition.

Totally agree, especially as p-curve was not published in methods journals and with closed reviews. I mean who vetted these papers. Just another example that peer-reviewed means shit. But once shit is published with the stamp of approval, people uncritically eat shit.

reproducibility = share data, get same results with same analyses replicability = share materials, redo experiments, get same result with new sample

There is now a response by Uri on the @datacolada.bsky.social blog. I have my own views on this. replicationindex.com/2025/10/28/h...

I’ve developed a new method for detecting publication bias in heterogeneous effects, with initial tests showing it outperforms existing approaches. If you’d like to support a young independent scientist in advancing meta-science, please read, share, or review—your help makes a difference! #MetaSci

@datacolada.bsky.social did respond to the criticism of p-curve, but the response is not convincing. Simulation hacking will not save p-curve. It is just not made for heterogenous data. replicationindex.com/2025/10/28/h...

@datacolada.bsky.social did respond to the criticism of p-curve, but the response is not convincing. Simulation hacking will not save p-curve. It is just not made for heterogenous data. replicationindex.com/2025/10/28/h...

Federal Judges— or staff, but same-same— used "AI" to summarize & draft rulings, issued them w/o checking the work, leading to basic factual errors, & thus undermining the facticity & validity of the rulings entire. Gee. Who Could Have Foreseen. *stares directly into the camera like in the office*

This may only interest history of science nerds like @lakens.bsky.social. Comments are welcome. replicationindex.com/2025/10/26/t...

Publish errors or perish

Simulation studies have a conflict of interest problem. The same team: - develops a new method - designs a simulation study to evaluate it However, the new method has to show good performance to get published. We propose living synthetic benchmarks to address the issue (doi.org/10.48550/arX...).

Texan Sharp Shooter, but yes, funny.

and Nazi's labeled political opponents as having a personality disorder and send them to concentration camps. Forget left-wing bias when fascists rule.

OMG. I just found out that I am a Bayesian, but not the strict type. I am what is called an "empirical Bayesian" and z-curve is an empirical Bayesian method. No hypothetical long-run probabilities, just estimation without priors. Most of us probably are. So confusing.

If we only publish results of a certain kind, we can make the argument look good. We must publish both kinds of results. (Feynman, 1985)

Closed pre-publication per-review is dead. I only review when I either (a) want to write a positive review to get something published (race IAT criticism) or not (criticism of post-hoc power). Everybody else is the same way. Push your shit and let open post-pub. peer-review (or better AI) decide.

When I say, I am doing some forensic meta-science work and am auditing influential studies in our field. Two responses: "So cool, teach my lab how to do that" and "Are you going to audit my lab's papers?" 😄

The proof is in good predictions. In 2012, I showed that evidence for the glucose theory of willpower is even worse than evidence for erotic time travel (Bem, 2011). A decade later, even Baumeister is no longer pushing the glucose model. replicationindex.com/2025/10/22/t...

Multiverse studies seem doomed to perpetuate a cycle of "look it varies!" "but not that much" "does too!". Here's another exchange that I had missed. "Robustness is better assessed with a few thoughtful models than with billions of regressions" doi.org/10.1073/pnas...

Is Psychology a Science? A Meta-Psychological-Scientists Perspective replicationindex.com/2025/10/21/p...

The journal Psychological Methods Decides that Psychological Research is not underpowered. Doping also does not exist if we don't test for it. Way to go, psycho-science. replicationindex.com/2025/10/20/t...