Researching reproducibility and creating the best and most comprehensive toolkit for "AI peer review" and forensic metscience. Run by @moreisdifferent.
Boston, MA
Joined February 2021
- Tweets131
- Following67
- Followers108
- Likes468
The Metascience Observatory retweeted
Today I am sharing a new preprint and accompanying blog post. Hundreds of published papers include impossible western blots showing protein expression in cell lines that have deleted or mutated genes. These clumsy mistakes are widespread and repeated by LLMs.
The Metascience Observatory retweeted
Cell Press investigating editor-in-chief accused of soliciting bribes from authors.
scmp.com/news/china/science/…
The Metascience Observatory retweeted
Dr. Willy Chertman of @HHSgov joins the Science: A New Golden Age Summit to discuss metascience and the process of improving scientific policy in the federal government.
Full remarks:
The Metascience Observatory retweeted
This is a big one - Opus 5.5's explainer for the 2026 Nobel Prize in Physiology or Medicine 👀
I had it do a deep dive into the history of optogenetics. The video covers the contributions of Boyden, Pan, Deisseroth, Nagel, Hegemann, Bamberg, Miesenböck, Zhang, and 15 others..
The Metascience Observatory retweeted
My full fireside conversation with @patrickc at the Science: A New Golden Age Summit.
We discussed building new institutions for discovery, accelerating scientific progress, and how America can move faster on the hardest scientific challenges.
The Metascience Observatory retweeted
Very exciting to see what @WillyInTheWorld has been up to @ HHS. Operation Trailblazer is a new initiative to make the US the #1 place in the world for clinical trials. Right now many trials are being moved overseas due to lower cost and better government support (faster, easier)
We can double the pace of scientific discovery.
Jay Bhattacharya is backing young investigators, accelerating rare disease research, and harnessing SI to transform biomedical discovery.
Willy Chertman is cutting research bureaucracy and modernizing clinical trials through Operation Trialblazer.
A new golden age of science.
@NIHDirector_Jay @DrJBhattacharya @willyintheworld @HHSGov
The Metascience Observatory retweeted
I had Claude create an explainer for the 2026 Nobel Prize in Chemistry 😍
The Metascience Observatory retweeted
I worry that academic fraud might be far more widespread than anyone can tell. It took 24 years to catch this case of likely fraud, and it was discovered only because the original data were available *and* because it was so obvious that certain rows had been copied elsewhere.
Imagine what people could get away with if only they were a little more clever in creating a few fake rows of data . . .
An influential paper by Ariely & Wertenbroch (2002) reported two studies with tampered data.
Post 1 of 2.
datacolada.org/138
The Metascience Observatory retweeted
Part 2 just dropped. Again, while there were warning signs in the paper, the smoking gun was in the raw data. How much fraud are we missing because raw data has been kept under lock and key? Should @NSF require that scientists to release old data as a prerequsite for future $$?
Colada[139] Artificial Deadlines Part 2
Evidence of fraud in the 25 year old gradesheet behind the key finding in an influential paper on deadlines and procrastination (Ariely & Wertenbroch 2002)
datacolada.org/139
The Metascience Observatory retweeted
Interesting effort to try to incentivize data sharing in science.
"NIH announces winners of S-index Challenge to advance data sharing across biomedical research"
nei.nih.gov/research-and-tra…
The winner was scholardata.io What are people's thoughts?
@NIH @NatEyeInstitute
The Metascience Observatory retweeted
For fun we're running the @MetascienceObs's forensic metascience agent on a number of papers on Cerebrolysin, a fake "peptide" drug. We're seeing contradictions between abstract text and data presented in the paper! Missing patients also. 3 PubPeer comments so far.. (link below)
The Metascience Observatory retweeted
A tool used to rephrase plagiarised text changed "final solution" to "mass killing of an ethnic group" in a chemistry paper. Now retracted thanks to a PubPeer user who spotted this.
The Metascience Observatory retweeted
I work at The Metascience Observatory (@MetascienceObs) on forensic metascience.
Our founder Dan Elton (@moreisdifferent) just published what we've learned so far.
One finding stands out, and it's not the one people will quote. 🧵
The Metascience Observatory retweeted
The Metascience Observatory (@MetascienceObs) has assembled a database w/ results from 8,307 replication experiments published in the scientific literature. We recently published a report on correlates of reproducibility. The results might surprise you. Link to full report ↓↓
The Metascience Observatory retweeted
In the @MetascienceObs's dataset of 5,832 papers containing a replication, works that failed to replicate are cited slightly more than works that replicated.
The Metascience Observatory retweeted
We're looking to grow community engagement at The Metascience Observatory as we scale up our work. If you'd like to join our Discord, DM @moreisdifferent. You can also sign up for our email newsletter here (it goes out every three months): metascienceobservatory.eo.pa…