Shoutout to the heart-warming shoutout in this Acknowledgements section: "We also thank Carolyn Gilbert, who in no way helped this paper get written, but is wonderful so she gets mentioned anyway."

The world today, in a nutshell. h/t the other site

"Credit in research goes hand in hand with responsibility" www.nature.com/articles/d41... Hmm. How often have you seen a big Nature or Science paper with 20 authors, who all get to enjoy the credit, but if it gets retracted it turns out the first author was solely responsible?

If people expert in research methods publicise lists of research they have assessed as adequate and at least potentially significant, that might help focus attention on deserving work, and less on work that's very-difficult-to-assess-regardless-of-expertise. bsky.app/profile/matt...

"The rewards to 'discovering' a spectacular scientific finding [in psychology] are large; the rewards to debunking frauds or deflating exaggerated claims are small if not non-existent. If these are the rules of the game, we should not be surprised at the way the game is played."

"attempts to introduce [preprint commenting systems] like this have not, so far, been very successful..." tinyurl.com/3dtax8jz These words of caution on the naive design of preprint commenting systems (mostly) ring true ~9 years later. The one exception: www.alphaxiv.org

Shall we call traditional peer review (for example, secret, only involving a couple of reviewers who may not between them have all the required skills, knowledge and experience, and unpaid) initial peer review: IPR?

Even ice can't stand ICE.

This is the avant-garde preprint service (hosted on AT Protocol) that I've been dreaming of. If it's a small fraction of its vision down the road, I'd still be happy. Check out Chive (docs.chive.pub/getting-star...). - scholar owned - open/community peer review - semantic linking to terms

Do you like bookmarking things and losing them in the Bookmarks hole? Or not knowing what your Bluesky peeps are bookmarking No? Then try semble.so (built on ATProtocol) and use social bookmarking to learn from the academic hivemind.

BREAKING: The UK Government is working with Canada and Australia to launch a coordinated ban on X, per the Telegraph

We know the drivers of research waste in academia are ⚠️Pressure to maximize papers and PhD students ⚠️Endless demands on time due to poor management ⚠️Stakeholders don't insist on robust quality systems to underpin mission critical work Solutions that don't address these are pointless.

Useful browser extension to help assess issues in papers: - retraction check - PubPeer comments check - tortured phrases check - fake references check We need more tools in this spirit to help make sense of science.

Let's bring this guy's positive resistance into the new year.

Depth of contribution in addition to breadth of contribution should also be considered. Currently, CRediT doesn’t have a distinction between much work occurred in each contribution category.

First MEP report. Medical research often uses non-parametric tests. We looked, and noticed that small-sample rank-based tests (i.e. M-W, Wilcoxon) have significant granularity... like p-values! So, here's GRIM for U values -- GRIM-U. Enjoy. medicalevidenceproject.org/grim-u-obser...

To bring in the New Year, here's a proposal for external regulation of academic publishing, through a voluntary system of journal certification to the ISO 9001 quality management standard. 🧪 #ResearchIntegrity (1/2) www.nature.com/articles/d41...

The Tools for Thought workshop at #CHI2026 is now accepting submissions! (due Feb 12, 2026 AoE) We invite work that explores how to effectively design, evaluate, or apply *AI-based tools that augment and protect human cognition*. For more, see: ai-tools-for-thought.github.io/workshop/

NEW: How many books did you read this year? If it was more than 2, congrats: you read more than the median American. Read on for way more @today.yougov.com polling from me on America's book-reading habits in 2025. 1/

I know at least one university that needs to hear this: Please don't evaluate faculty on whether or not they are presenting at professional conferences if you expect faculty to pay out of pocket to attend professional conferences.

What if some great papers were published only as preprints? No journal, no peer review, just the credibility of the author. Would this strengthen Open Science or weaken trust? Join discussion on Peeref: www.peeref.com/notes/976848... #openscience #preprint

The RetractionRisk scanner integrates evidence from PubPeer comments, negative posts on X and Bluesky, as well as article status information, and applies algorithms to compute an overall retraction risk level for each article

make a Bond movie academic Quantum of Vacation

Joining Bluesky was just step 1. Now, time to execute the Master Plan and coordinate a scholar-centered open communication movement. Join ATProto Science to level up!

I'veh neber ussed a spell-chucker. All mistakes mine, huppy to report.

Ok, this is nuts. Once you see it you cannot unsee it. Do you see it? (OP @drgbuckingham.bsky.social )

I appreciate LLMs because at least now people are attempting to discern the Real from the Fake. Pre-LLMs, people didn't try much to dig deeper into things.

"Cuthbert Daniel claims to have invented the concept behind the variance inflation factor, but did not come up with the name." Shade thrown. 🤨 en.wikipedia.org/wiki/Varianc...

Seeing on Prolific subreddit that some researchers are putting white text against white b'ground to detect AI/bots (assuming only bots see/read the hidden text). BUT, many Prolific participants use dark mode, &/or highlight text, so they read the text & get dinged as if they're bots. Don't do this.

The Modular Peer Review Working Group we co-organized with @continuous.foundation kicks off in January, 2026! 🚀 Join the cohort shaping how we review modular research outputs like data, code, methods, and visuals. Sign up here: continuousfoundation.org/peer-form

From the people who brought you Dishonesty scholar is dishonest in their lab, here's Ethics scholar fakes references in ethics journal. Scholarly publishing is both comedy and tragedy.

Published in 2015, but still true. "Currently there is no mechanism for an author to be alerted to discussions of his or her article on social media." "The key goal... involve the entire community - authors,editors, and the target audience... Cf: @altmetric.com dx.doi.org/10.1080/0898...

If bluesky and Twitter have taught me anything, it's that every discipline is the worst at analyzing data, except for all the others.

Internships that request papers in top venues are truly my biggest gripe during application season. Am I going to play your dumb game or build something special myself? Hmm... I wonder...

Hello world! Follow this account to receive announcements, invitations for open peer review, and publications from Replication Research.

I've been using tools like Claude Code for AI-assisted coding lately, but always with an eye on what is being done. The results are genuinely impressive, but each tool requires special care. If you're using Google Antigravity, airgap it to avoid file deletions.

Has anyone successfully taught diverse statistical paradigms within a single course (Bayesian, frequentist, etc.)? Typically, students need to encounter them one-a-time. Is it too demanding to mix and match approaches in one course? #stats #Bayesian

#economics: I have plenty of acknowledgements, want to get them out of the way on page 1, but I refuse to use the #CreDiT taxonomy. Folks, there's a better way. Use the Tenzing app: rollercoaster.shinyapps.io/tenzing/ Better yet, modify CreDiT (high, medium, and low levels of contributions).

In 2014 I introduced a replication project in my grad research methods class. I taught this version of the class 4 times (no longer teach it). Some tallies: 9 published replication papers; 30 grad student authors; 19 *open* data sets; materials, syntax, etc also open (all on OSF). Check them out 👇

Has anyone successfully taught diverse statistical paradigms within a single course (Bayesian, frequentist, etc.)? Typically, students need to encounter them one-a-time. Is it too demanding to mix and match approaches in one course? #stats #Bayesian

Looking for a quick academic presence without much effort? Something like linktree for scholars? @lanyards.app by @renderg.host is intuitive, sleek, and user-centered. Connects with your bluesky profile.

🌏 A thought experiment for researchers: If you could click one button to share your (non-sensitive) notes from your articles in Zotero, Mendeley, PaperPile, etc. would you open those data up to the world? #metascience

Me: Paper titles are hard, require sophistication, and must conform to sensible standards. A/B test them, go back to the drawing board, attend a seance, consult an astrologer, run simulations on supercomputers, and meditate until I achieve Nirvana. Philosophers: "On [topic name here]"

Title and abstract: This is a "qualitative study." Results: Everything is quantitative. Am I missing something #medsky ? Wouldn't "content analysis" be more accurate?

It would, of course, be great if the underlying results were made public but it looks like meta in 2019 ran withdrawal experiments to examine polarization and well being ahead of forging collaborations with academics. Seems like more than enough to design bias the collaborations towards nulls 🧵

This Thanksgiving, I'll be thankful for the time to test the newer foundation models. Three to choose from...

#SciComment, what do we think of the benefits of using Batman in metros to nudge riders? Strong odds ratio, low p-value, but a priming claim in 2025 requires a bit more thought IMO. www.nature.com/articles/s44...

I'm tempted to add a full page of quotes to my next preprint because it's too hard to choose among the great options and because I really want to hammer home some ideas that I fear readers won't get. Is that too much? Are there precedents in the literature?

In a field where evidence suggests that a high proportion of papers have serious flaws, a database of "potentially important" papers which appear *not* to have serious flaws might be useful, as well as identifying those that do.

I'll come back to this idea because that complements something that's been brewing in my mind for some time. Some needed resolution in our citation practices where we explicitly distinguish between "load-bearing" work on which our inferences rely and others (kitchen sink citations). And the need to