iiinova — where your digital self keeps evolving

Reposted by Unfork Statistics

Joshua Weitz @joshuasweitz.bsky.social · 07/03/2026

The data is in: the NIH goalposts have shifted. What were once almost certain fundable scores have become coin flips and what used to be likely grants have become aspirational, leading to fewer awards. Another manifestation of how HHS policies have led to fewer awards and less science.

Graph of award probability of R35 and R01 from NIH factbook as a function of review rank percentile. As is apparent, 2025 is a significant departure, with lower award probabilities at all scores <40 and significant departures from norm, where even being in the top 10% is no longer a nearly certain indicator of success.

Data source: https://report.nih.gov/nihdatabook/report/302

18697420

Reposted by Unfork Statistics

Darren Dahly @statsepi.bsky.social · 25/02/2026

I tried to tell y'all.

69823

Reposted by Unfork Statistics

Peter Tennant @pwgtennant.bsky.social · 25/02/2026

This recent RCT of an "AI stethoscope" claims the technology "shows promise" for diagnosing cardiovascular conditions. It does not. It is a textbook example of the risks of conducting unprincipled 'per protocol analyses'. Once again, peer review at a major medical journal has failed. 🧵 1/

8436186

Reposted by Unfork Statistics

Ben Goldacre @bengoldacre.bsky.social · 23/02/2026

OpenSAFELY is open from today! Huge thanks to all who supported this vast collaboration: whole population GP data; in a productive platform; innovative privacy protections; unprecedented support from professions, privacy campaigners; &c Now it's over to users! www.bennett.ox.ac.uk/blog/2026/02...

bennett.ox.ac.uk

OpenSAFELY news: you can apply to do non-COVID research, from today! | Bennett Institute for Applied Data Science

We are delighted to announce that - from today - you can submit applications to the OpenSAFELY service for non-COVID-19 studies.

618692

Reposted by Unfork Statistics

American Statistical Association History of Statistics @hos-asa.bsky.social · 23/02/2026

Happy 🎂 Richard Price b #OTD 1723 (d 19 Apr 1791) Best known for popularizing the essays of Thomas Bayes, the basis of what is now called Bayes's Theorem. Price’s contributions were more substantial than he has been credited with. 1/4🧵

142

Reposted by Unfork Statistics

Dr Mel T @drmel.bsky.social · 10/02/2026

Can anyone tell me which crypto coin I can cash all this ‘Enhanced Reputation’ in at? Because I’m still carrying £15k worth of debt from altering my UK terrace so I can live downstairs when my MS gets to that stage… and I don’t have a job.

174

Reposted by Unfork Statistics

Alex Coad @coad.bsky.social · 29/01/2026

New submission format at SBE: “Replications as Registered Reports” link.springer.com/journal/1118... You can get "in-principle acceptance" before data collection even begins; final paper gets published regardless the results, if the study is conducted rigorously. #EconSky

12217

Unfork Statistics @btjo.bsky.social · 28/01/2026

Sadly the NPS is fundamentally flawed regardless of how nonsensical, or otherwise, it is to ask if customers are moved to do voluntary work for the marketing team. This is a decent summary of the many, many problems with it: archive.ph/I3uce

000

Unfork Statistics @btjo.bsky.social · 28/01/2026

More perils ... bsky.app/profile/did:...

010

Unfork Statistics @btjo.bsky.social · 28/01/2026

On the perils of open data. Archive link: archive.ph/ZavpU

archive.ph

110

Reposted by Unfork Statistics

Andrew Gelman et al. @statmodeling.bsky.social · 27/01/2026

Machine learning research is not serious research and therefore hallucinated references are not necessarily a big deal, agrees a prestigious group of machine learning researchers statmodeling.stat.columbia.edu/2026/01/26/m...

statmodeling.stat.columbia.edu

0168

Reposted by Unfork Statistics

QUOD Biobank @quod-biobank.bsky.social · 14/01/2026

We are always excited to see the QUOD Biobank grow, with samples from over 9,000 donors. We are grateful, as always, for the generosity of our donor families and to the teams across the UK who make collection possible, helping us to support research into improving transplantation outcomes.

011

Reposted by Unfork Statistics

Jeffrey Brainard @jeffreybrainard.bsky.social · 17/01/2026

#SpecialIssues have fueled the growth of some of the largest #OpenAccess publishers. Does a journal that allows a #GuestEditor to both plan a special issue and write many articles in it have a conflict of interest? #scicomm #peerreview @science.org www.science.org/content/arti...

science.org

Some guest editors pack special issues with their own articles

Thousands have penned more than one-third of a journal issue, raising conflict-of-interest concerns

22618

Reposted by Unfork Statistics

Julia M. Rohrer @dingdingpeng.the100.ci · 07/01/2026

Here's a suggestion for a New Year's resolution: If you see influential bad research, say something. One part of the whole replication crisis story is that a lot of psychological researchers privately knew that a lot of stuff was bad, but it wasn't discussed publicly.

1114243

Reposted by Unfork Statistics

Stelios Katsanevakis @skatsanevakis.bsky.social · 06/01/2026

Point of no return: academic #funding is at a tipping point. When the effort and money spent applying for grants equals or exceeds the funding awarded, the system stops serving science and starts draining it. Time to rethink how we fund #research. 🧪🌐🌍 www.nature.com/articles/d41...

nature.com

Point of no returns: researchers are crossing a threshold in the fight for funding

With so little money to go round, the costs of competing for grants can exceed what the grants are worth. When that happens, nobody wins.

14019

Reposted by Unfork Statistics

Robin Blythe @rbly.bsky.social · 04/01/2026

Good thread. My concern is, CI is presented as a solution to bad stats when in reality it suffers from many of the same limitations. We can almost never be certain of most reqs in any case -- exogeneity; data integrity; missingness mechanisms; selection bias to name a few (1/2) #statsky #econsky

141

Reposted by Unfork Statistics

Raywat Deonandan 🍁🇨🇦 @deonandan.bsky.social · 23/12/2025

WTF is this? I did not write or submit this paper. Are journals now run entirely by AIs who hallucinate entire publications? Knowledge Of Pregnant Women About The Benefits Of Pregnancy Books In Ottawa | International Journal of Health Engineering and Technology share.google/ArcgxhUDfSY9...

share.google

Knowledge Of Pregnant Women About The Benefits Of Pregnancy Books In Ottawa | International Journal of Health Engineering and Technology

135821

Reposted by Unfork Statistics

Joe Bak-Coleman @jbakcoleman.bsky.social · 20/12/2025

These claims about what AI can do truly need citations. Checking statistics isn’t routine work, nor is whether citations to the literature are sensible. Whether statistics are appropriate and render reasonable inferences is inherently a question of human judgement.

1210322

Reposted by Unfork Statistics

Ben Williamson @benpatrickwill.bsky.social · 19/12/2025

Finally, if you can get through the paywall this piece from earlier in the week illuminates how the experience above is just a tiny part of an industrial-scale, global, and sector-wide catastrophe of AI and citations to nonexistent papers in academic publishing bsky.app/profile/mile...

13823215

Reposted by Unfork Statistics

Floe Foxon @floefoxon.bsky.social · 18/12/2025

Does e-cigarette use cause stroke? ~2 years ago, Gal Cohen & I alerted MDPI to fatal flaws in a published paper which claimed nicotine vaping causes stroke. This morning, the journal finally retracted the flawed article: doi.org/10.3390/neur... ... #tobaccocontrol #harmreduction #statistics

112

Reposted by Unfork Statistics

Julia M. Rohrer @dingdingpeng.the100.ci · 17/12/2025

You remember that Nature Aging paper about how multilingualism protects against accelerated aging? Well…

916953

Reposted by Unfork Statistics

Lukas Röseler @aufdroeseler.bsky.social · 04/12/2025

I made an error that I found thanks to @lukaswallrich.bsky.social. The code and preprint have been corrected/updated. Effects on false positive rates are smaller and only for alternative and correlation method but I think the overall argument still works.

1146

Reposted by Unfork Statistics

Jeffrey B. Benyacar @drjeffreyb.bsky.social · 07/12/2025

They're eliminating the Statistic Department as the Admins apparently can't do basic math or data scrubbing. "Other errors they found included numerous data errors, such as misplaced decimal points and incorrectly typed data." #stats

021

Unfork Statistics @btjo.bsky.social · 04/12/2025

Of course they were funding it.

“In this case we are up against the seductive allure of claims we’d like to believe are true but aren’t – that tobacco is really not that harmful, at least not by itself, or that we can prevent and cure cancer with some kind of ‘mind work,’ and so on,” Buchanan told us. “This dubious research needs to be flagged with sufficient prominence and transparency wherever it appears to prevent researchers taking it in good faith.”

A 1991 book by Eysenck, for instance, claimed psychosocial factors like stress and personality type were six times as predictive of cancer and heart disease as smoking, cholesterol levels or blood pressure. Pelosi told us the book summarized Eysenck’s work with Grossarth-Maticek and a draft had been sent to tobacco giant Philip Morris International, which helped fund the two researchers’ work.

190

Unfork Statistics @btjo.bsky.social · 24/11/2025

[Pic] is such an important point. If you're never wrong, you can't learn anything. And if you're afraid to be wrong, you're easy prey for plausible-sounding misinformation. Education, no matter the subject, is about how to think. The facts/ideas/materials change, the skills required do not.

Students are afraid to fail, and AI presents itself as a savior. But what we learn from history is that progress requires failure. It requires reflection. Students are not just undermining their ability to learn, but to someday lead.

010

Reposted by Unfork Statistics

Nick Tsergas @nicktsergas.ca · 20/08/2025

The most important part of this article, to me, is the end. It challenges a false binary between “long COVID” and “everyone else.” Maybe it's a spectrum... 🧵 (1/5) www.bmj.com/content/390/...

bmj.com

Why scientists are rethinking the immune effects of SARS-CoV-2

“Immunity debt,” a theory to explain the global surge in non-covid infections since pandemic restrictions were lifted, is increasingly being challenged by emerging evidence. Nick Tsergas reports Myco...

24801335

Unfork Statistics @btjo.bsky.social · 22/11/2025

Domain knowledge is always critical for the appropriate application of statistical methods. Here, it would have prevented this embarrassing episode occurring at all.

To him, the research seemed off. Elkan used to work as a managing director at Goldman Sachs and as a scientist at Amazon, where he led machine-learning teams. The technology sounded too good to be true to him, given its supposed date of deployment. Also, 3M and Corning were the only U.S. materials companies Elkan knew of that employed a large number of scientists, and he questioned why, if they were inventing so many new materials, they would be keeping it secret.

020

Reposted by Unfork Statistics

E. David Klonsky, PhD @klonskylab.bsky.social · 04/11/2025

Not sure where this is from, but assuming it’s right, there are clear implications for countering the replication crisis and “false positive” psychology: Insufficient sample sizes are a primary culprit, and should be a primary target of reforms.

021

Unfork Statistics @btjo.bsky.social · 02/11/2025

Unfortunately, they can fake data just as easily, and do much more damage to the literature with it. Need checking of grants, ethical approval, etc to verify that the research actually took place.

110

Reposted by Unfork Statistics

Philipp Heimberger @heimbergecon.bsky.social · 31/10/2025

Publication bias is greater at the top five journals: "56% of statistically significant results [in leading econ journals] were selected to be statistically significant. Selection bias is greater at top5 journals, where 66% of significant results were selected to be significant."

1126

Reposted by Unfork Statistics

Jess Calarco @jessicacalarco.com · 01/11/2025

We have progressed from data collection to data analysis.

My 11-year-old sitting with her pile of Halloween candy, sorting it into a bar graph

971344674087

Unfork Statistics @btjo.bsky.social · 15/10/2025

That's against NHS policy (assuming you're age 50-71): www.gov.uk/government/p...

gov.uk

NHS population screening: information for trans and non-binary people

021

Reposted by Unfork Statistics

Rebecca Sear @rebeccasear.bsky.social · 13/10/2025

"Two major publishers have begun to automatically reject the vast majority of papers based on public health data sets, following revelations that unscrupulous actors use these data sets to churn out nonsense scientific papers." www.science.org/content/arti...

science.org

Journals and publishers crack down on research from open health data sets

PLOS, Frontiers, and others announce policies trying to stem the tide of suspect research

1218

Reposted by Unfork Statistics

Julia M. Rohrer @dingdingpeng.the100.ci · 01/10/2025

A lot of psych is already conducted with online convenience samples & ppl are probably excited about silicon samples bc it would allow them to crank out more studies for even less 💸 How about we reconsider the idea that sciencey science involves collecting own data. www.science.org/content/arti...

science.org

AI-generated ‘participants’ can lead social science experiments astray, study finds

Data produced by “silicon samples” depends on researchers’ exact choice of models, prompts, and settings

1423473

Reposted by Unfork Statistics

Dr. Jens Foell @jensfoell.de · 04/10/2025

People are running stats on LLM-generated participants and think they’re being social scientists when in fact they’re technically just playing a very strange video game. This is like saying you’re doing math research because you’re playing sudoku. www.science.org/content/arti...

science.org

AI-generated ‘participants’ can lead social science experiments astray, study finds

Data produced by “silicon samples” depends on researchers’ exact choice of models, prompts, and settings

1028991

Unfork Statistics @btjo.bsky.social · 01/10/2025

I walk my dog by the river. Properly managed sewage systems are sexy. top-of-the-poops.org/waterway/uni...

top-of-the-poops.org

Top of the Poops | Waterway | River North Petteril

Analysing sewage dumps by water companies

000

Unfork Statistics @btjo.bsky.social · 01/10/2025

This is a failure to explain herd immunity properly. Some parents hear only a demand that they vaccinate to protect others. But their (unvaccinated) kids are most likely to be harmed in an outbreak. It's their (unvaccinated) kids who pay the price of herd immunity breaking down. Spell it out.

"The impact of their vaccine choices on others is not necessarily part of their calculation. And why should it be? There are few other arenas in which parents are asked to consider the effects of their choices on other children."

000

Reposted by Unfork Statistics

Jake Scott, MD @jakescottmd.bsky.social · 13/09/2025

A US senate subcommittee was told this unpublished vaccine study was 'the most important ever.' Henry Ford Health now confirms why it was never published: it failed scientific standards. Here's what's actually wrong with the study: jakescottmd.substack.com/p/why-the-senates-most-important-vaccine

jakescottmd.substack.com

Why the Senate's 'Most Important Vaccine Study Ever' Was Never Published

The fatal flaws that made this study unpublishable

13320137

Reposted by Unfork Statistics

Ben Williamson @benpatrickwill.bsky.social · 07/09/2025

Academic authors, here's a peek into the black box of journal publishing from an journal editor if you can bear it:

18995470

Reposted by Unfork Statistics

Darren Dahly @statsepi.bsky.social · 06/09/2025

This can't be true because academic papers are all peer-reviewed by experienced experts who would be able to easily identify these shenanigans and stop them in their tracks.

34312

Reposted by Unfork Statistics

Achim Zeileis @zeileis.org · 05/09/2025

🚨 Relaunch of CRAN Task View: Clinical Trial Design, Monitoring, and Analysis #rstats Complete rewrite by an int'l team of experts in #clinicaltrials Ya Wang, Thomas Jaki @thomasjaki.bsky.social, Laura Pascasio Harris, Orla Doyle, Elias Laurin Meyer, Wilmar Igl cran.r-project.org/view=Clinica...

Screenshot of the CRAN Task View "Clinical Trial Design, Monitoring, and Analysis" at https://CRAN.R-project.org/view=ClinicalTrials

1156

Unfork Statistics @btjo.bsky.social · 30/08/2025

Nice example of the ecological fallacy, with a conclusion so absurd it's hard to understand how it got written down. #stats

A new WalletHub study shows how deeply car loan debt cuts into household budgets across the U.S.

The financial site reviewed more than 2,500 cities, measuring median auto-loan balances against median incomes. The results revealed a clear warning: in more than 160 cities, drivers spend half or more of their yearly income on car loans alone.

That burden comes as this particular debt keeps climbing nationwide.

Some of the hardest-hit cities are in Texas

Rio Grande City leads the list, where the median auto-loan balance is nearly equal to median income: $32,173 in debt versus $33,782 in earnings. That’s a 95% debt-to-income ratio.

Nearby towns like Donna, Mercedes, San Juan, and Brownsville all land in the danger zone as well. Uvalde, Eagle Pass, and Pharr also show residents committing well over two-thirds of their annual income to car notes.

020