Hallucinated citations are polluting the scientific literature. What can be done? Tens of thousands of publications from 2025 might include invalid references generated by AI, a Nature analysis suggests.
That's tens of thousands of authors whose submissions can be desk-rejected from here on out, leaving some more room for the rest of us.
www.nature.com/articles/d41...
Om ni har svårt att komma ihåg grundlagarna, gör vårt Sex and the City-test!
Inslaget/debatten med forskaren bakom Finger-studien och kritikern Magnus Gisslén börjar 6 min 12 sek in.
Back by popular demand! Nu kan journalister söka min Fojokurs i vetenskapsrapportering. Under tre dagar i Stockholm går vi bland annat igenom hur man läser och tolkar studier, hur man undviker vanliga fallgropar och se igenom hajp. Sök senast 18 maj.
Peer review may be entering the age of verification
For decades, peer review has operated largely on trust. Editors trust reviewers to read papers carefully. Authors trust reviewers to evaluate work fairly. Publishers trust the community to follow agreed rules.
🧵
Reklam för röntgen och ortopedi precis i slutet av skidbacken
Det gäller att annonsera på rätt plats!
Many folk are surprised to discover thay Risk of Bias assessment tools tend not to interrogate the question “Did this study actually happen? And are its results trustworthy enough to believe?”
Jack’s Cochrane endorsed INSPECT-SR checks have done a lot to mainstream such Trustworthiness Assessment.
Very nice INSPECT-SR website containing the guidance and editable template, made by @ianhussey.mmmdata.io
inspect.sr
Hepatica in the sun.
Spring flowers.
It's a v weird retraction, but at least it's a retraction
retractionwatch.com/2026/03/31/b...
But why delegate responsibility for judging the integrity of the results to Iranian FDA? This just caused unnecessary delay, when the issues were blindingly obvious and easily verifiable. 1/3
We must go further: redesign institutions, test causal interventions, and have national funders establish dedicated meta-research panels: an investment this field has long deserved.
📄 Miske et al. doi.org/10.1038/s415...
📄 doi.org/10.1038/d415...
@ucoimbra.bsky.social @excelscior-era.bsky.social
Du väljer själv om du vill delta på plats eller digitalt!
Tips på spännande konferens!
Vill särskilt lyfta dessa två temasessioner där jag är med och bidrar!
”När forskning ger styrka inför säkra beslut”
”Från deepfakes till falska forskningsresultat, så möter vi AI-utmaningen.”
formas.se/kunskap-och-...
Recording of the recent INSPECT-SR Cochrane webinar here (requires Cochrane login): www.cochrane.org/events/inspe...
Graphic of a person with a magnifying glass near a large laptop displaying documents. Surrounding items include a plant, paperclips, lightbulb, and paper airplane icons.
A methodology paper introduces the "exit" meta-analysis concept to address the growing concern of research waste, which determines when a research question has been conclusively addressed, indicating no need for further studies. 👇
https://ow.ly/7NGr50VnRBn
#JBIMethodology
The elephant in the room: Peer review
Peer review is often described as the cornerstone of scholarly publishing.
It is meant to ensure quality, protect standards, and strengthen research before it enters the scientific record.
But there is an uncomfortable question that deserves more attention.
🧵
1/2 🎉 Exciting news for #EvidenceSynthesis! 19 BMC journals have incorporated the CEE Checklist for Assessing Validity of Systematic Reviews in the Environmental Sector into their submission guidelines through our collaboration with @springernature.com #CEEAtWork 👇
Oh nothing, just a peer reviewed paper my colleagues found...
doi.org/10.1016/j.ma...
"1 mL of the mass killing of an ethnic group was opposed to 20 mL of the skin sample and unprotected to light for 7 min."
Even AI knows what's wrong here, but @elsevierconnect.bsky.social doesn't.
@lajeunesselab.bsky.social has written the paper that meta-analysts were waiting for, and I can't recommend it more:
Converting and Constructing Effect Sizes With the Response Ratio doi.org/10.1111/ele....
Go check it out (and use it!)
Still time to join our #MetaAnalysis in R course with @asanchez-tojar.bsky.social & @shreyadimri.bsky.social (11–14 May, online)!
Learn systematic reviews, effect sizes, meta-analysis & meta-regression in R with hands-on exercises & real datasets.
www.physalia-courses.org/courses-work...
#RStats
Idag klockan 10.00 släpps den årliga klimatrapporten från regeringens expertgrupp Klimatpolitiska rådet. I år ska den innehålla framåtsyftande rekommendationer för kommande regering. Följ presentationen här ➡️ www.klimatpolitiskaradet.se/rapport-2026/
#medlibs - Version 3 of the Responsible use of AI in Evidence Synthesis (RAISE) is out! osf.io/fwaud/files/...
The Ctrl-Z Award: Recognizing the courage to correct or retract published work retractionwatch.com/ctrl-z-award/
Great discussion with journal editors as we start to develop INSPECT-JR (a version of INSPECT-SR for the editorial assessment context). Very different to the systematic review context.
Stylised image of a wolf holding papers in one paw, and with the other beginning to cover its face with a sheep mask
⚠️The number of predatory journals continues to rise.
This paper outlines some potential strategies for systematic review authors dealing with content they may encounter from predatory journals.👇
https://ow.ly/w3OE50YrC0P
#JBImethodology
This is hilarious.
Also, completely enraging.
A labrador with subtext in two versions one read ” I chased a squirrel” and the other one is a LinkedIn version starting with proud to announce that I efficiently executed a rapid response squirrel displacement strategy… and the goes on in long text.
Saw this in and 💯 love it!
Ett exempel på hur viktigt det är med utformning av svarsalternativ i enkäter!
Och vilket stort utrymme det finns för att bygga in bias….
Thanks!
I am looking for any published, peer‑reviewed scientific article that has been verified as written by an LLM for a presentation. Happy for input you might have!
