Frank Harrell
@f2harrell.bsky.social
Professor of Biostatistics
Vanderbilt University School of Medicine
Expert Biostatistics Advisor
FDA Center for Drug Evaluation and Research
https://hbiostat.org https://fharrell.com
Vanderbilt University School of Medicine
Expert Biostatistics Advisor
FDA Center for Drug Evaluation and Research
https://hbiostat.org https://fharrell.com
A lot of crap research is being done with the microbiome. At the heart of much of the research is fundamentally unreliable data.
Perhaps this is why so much microbiome research is hard to believe.
www.gov.uk/government/n...
www.gov.uk/government/n...
MHRA-led study reveals major inconsistencies in global microbiome research
International collaboration establishes new quality standards to improve reliability of gut health studies – improving accuracy to provide better diagnosis and treatment.
www.gov.uk
November 16, 2025 at 12:46 PM
A lot of crap research is being done with the microbiome. At the heart of much of the research is fundamentally unreliable data.
Reposted by Frank Harrell
Weekend reads: Debunking ‘When Prophecy Fails’; ‘Godfather of AI’ first to reach 1 million citations; ‘Cake causes herpes?’
Weekend reads: Debunking ‘When Prophecy Fails’; ‘Godfather of AI’ first to reach 1 million citations; ‘Cake causes herpes?’
Author changes name, publishes 10 papers in journals that banned him Bug in Springer Nature metadata may be causing ‘significant, systemic’ citation inflation Math is back as Clarivate boosts integ…
retractionwatch.com
November 16, 2025 at 1:09 AM
Weekend reads: Debunking ‘When Prophecy Fails’; ‘Godfather of AI’ first to reach 1 million citations; ‘Cake causes herpes?’
Nice to see all these confidence intervals. I use Wilson as a default but should be using Bayes. Challenge: figure out why “exact” confidence intervals are not recommended.
For people trying to teach themselves more about statistics, go read about these different approaches and try to make sense of why they don't exactly agree. What are they doing differently? Use wikipedia. Look up new terms along the way.
My #rstats cheat code for today is the binom.confint function in the binom package that will spit out *12* different ways of calculating a CI for a proportion.
Also, this is why you use R for statistics...
(and of course the correct CI method is bayes 😎)
Also, this is why you use R for statistics...
(and of course the correct CI method is bayes 😎)
November 16, 2025 at 12:32 PM
Nice to see all these confidence intervals. I use Wilson as a default but should be using Bayes. Challenge: figure out why “exact” confidence intervals are not recommended.
Reposted by Frank Harrell
I wrote something about publication bias at statmodeling.stat.columbia.edu/2025/11/14/t...
November 14, 2025 at 9:17 PM
I wrote something about publication bias at statmodeling.stat.columbia.edu/2025/11/14/t...
Reposted by Frank Harrell
We did a thing. 😬
The link between the gut #microbiome and autism is not backed by science, researchers say.
Read the full opinion piece in @cp-neuron.bsky.social: spkl.io/63322AbxpA
@wiringthebrain.bsky.social, @statsepi.bsky.social, & @deevybee.bsky.social
Read the full opinion piece in @cp-neuron.bsky.social: spkl.io/63322AbxpA
@wiringthebrain.bsky.social, @statsepi.bsky.social, & @deevybee.bsky.social
November 13, 2025 at 5:59 PM
We did a thing. 😬
Reposted by Frank Harrell
What is the most profitable industry in the world, this side of the law? Not oil, not IT, not pharma.
It's *scientific publishing*.
We call this the Drain of Scientific Publishing.
Paper: arxiv.org/abs/2511.04820
Background: doi.org/10.1162/qss_...
Thread @markhanson.fediscience.org.ap.brid.gy 👇
It's *scientific publishing*.
We call this the Drain of Scientific Publishing.
Paper: arxiv.org/abs/2511.04820
Background: doi.org/10.1162/qss_...
Thread @markhanson.fediscience.org.ap.brid.gy 👇
November 12, 2025 at 10:31 AM
What is the most profitable industry in the world, this side of the law? Not oil, not IT, not pharma.
It's *scientific publishing*.
We call this the Drain of Scientific Publishing.
Paper: arxiv.org/abs/2511.04820
Background: doi.org/10.1162/qss_...
Thread @markhanson.fediscience.org.ap.brid.gy 👇
It's *scientific publishing*.
We call this the Drain of Scientific Publishing.
Paper: arxiv.org/abs/2511.04820
Background: doi.org/10.1162/qss_...
Thread @markhanson.fediscience.org.ap.brid.gy 👇
Reposted by Frank Harrell
Reposted by Frank Harrell
I had the opportunity to review the work experience and skills of numerous people who were laid off or fired from public health jobs by Donald Trump.
I want every taxpayer to know that the US has lost an immeasurable number of the smartest, most experienced, and hardworking people on the planet.
I want every taxpayer to know that the US has lost an immeasurable number of the smartest, most experienced, and hardworking people on the planet.
November 12, 2025 at 9:43 PM
I had the opportunity to review the work experience and skills of numerous people who were laid off or fired from public health jobs by Donald Trump.
I want every taxpayer to know that the US has lost an immeasurable number of the smartest, most experienced, and hardworking people on the planet.
I want every taxpayer to know that the US has lost an immeasurable number of the smartest, most experienced, and hardworking people on the planet.
Reposted by Frank Harrell
A new release of the mgcv #RStats 📦 is out on CRAN and Simon Wood (U Edinburgh) has added some significant new features despite the small bump in version number:
🌟 scasm() for estimating GAMs with shape constrained smooths. Can be used with any family & smoothness selection is via the EFS method
🌟 scasm() for estimating GAMs with shape constrained smooths. Can be used with any family & smoothness selection is via the EFS method
November 12, 2025 at 11:28 AM
A new release of the mgcv #RStats 📦 is out on CRAN and Simon Wood (U Edinburgh) has added some significant new features despite the small bump in version number:
🌟 scasm() for estimating GAMs with shape constrained smooths. Can be used with any family & smoothness selection is via the EFS method
🌟 scasm() for estimating GAMs with shape constrained smooths. Can be used with any family & smoothness selection is via the EFS method
Reposted by Frank Harrell
Here’s a gift link (no paywall) to the article, “Catholic bishops condemn ‘indiscriminate mass deportation’ in rare statement”:
wapo.st/448fDw6
wapo.st/448fDw6
November 13, 2025 at 5:03 AM
Here’s a gift link (no paywall) to the article, “Catholic bishops condemn ‘indiscriminate mass deportation’ in rare statement”:
wapo.st/448fDw6
wapo.st/448fDw6
Some of the best data detective work I’ve seen, and embarrassing for @bmj.com
This is one of the most remarkable academic debacles I've ever seen.
A large RCT got published in BMJ. There are currently 44 Pubpeer comments, mostly about the data, including...well. Read for yourself.
pubpeer.com/publications...
A large RCT got published in BMJ. There are currently 44 Pubpeer comments, mostly about the data, including...well. Read for yourself.
pubpeer.com/publications...
PubPeer - Prevention of acute myocardial infarction induced heart fail...
There are comments on PubPeer for publication: Prevention of acute myocardial infarction induced heart failure by intracoronary infusion of mesenchymal stem cells: phase 3 randomised clinical trial (P...
pubpeer.com
November 11, 2025 at 2:03 PM
Some of the best data detective work I’ve seen, and embarrassing for @bmj.com
Reposted by Frank Harrell
In the century leading up to 1975, nearly 6000 freighters went down in the Great Lakes.
The Edmund Fitzgerald was the last.
The last. In 50 years, not a single commercial freighter has been lost in the Great Lakes.
Why?
It's NOAA. Of course it's NOAA.
The Edmund Fitzgerald was the last.
The last. In 50 years, not a single commercial freighter has been lost in the Great Lakes.
Why?
It's NOAA. Of course it's NOAA.
November 11, 2025 at 1:50 AM
In the century leading up to 1975, nearly 6000 freighters went down in the Great Lakes.
The Edmund Fitzgerald was the last.
The last. In 50 years, not a single commercial freighter has been lost in the Great Lakes.
Why?
It's NOAA. Of course it's NOAA.
The Edmund Fitzgerald was the last.
The last. In 50 years, not a single commercial freighter has been lost in the Great Lakes.
Why?
It's NOAA. Of course it's NOAA.
Reposted by Frank Harrell
Ten Simple Rules for AI-Assisted Coding in Science arxiv.org/abs/2510.22254 - our latest, led by @ericwbridgeford.bsky.social
Ten Simple Rules for AI-Assisted Coding in Science
While AI coding tools have demonstrated potential to accelerate software development, their use in scientific computing raises critical questions about code quality and scientific validity. In this pa...
arxiv.org
October 28, 2025 at 1:12 PM
Ten Simple Rules for AI-Assisted Coding in Science arxiv.org/abs/2510.22254 - our latest, led by @ericwbridgeford.bsky.social
Reposted by Frank Harrell
What intro stats textbook to use?
statmodeling.stat.columbia.edu/2025/11/09/w...
statmodeling.stat.columbia.edu/2025/11/09/w...
What intro stats textbook to use? | Statistical Modeling, Causal Inference, and Social Science
statmodeling.stat.columbia.edu
November 9, 2025 at 4:20 PM
What intro stats textbook to use?
statmodeling.stat.columbia.edu/2025/11/09/w...
statmodeling.stat.columbia.edu/2025/11/09/w...
Reposted by Frank Harrell
Friends of Stats: The U. of Nebraska-Lincoln Department of Statistics needs your help. The university's Chancellor submitted a FINAL proposal today for budget cuts (budgetprocess.unl.edu/final-budget...) that eliminates the department. 1/ #LNK #Nebraska #Huskers @amstatnews.bsky.social
Final Budget Reduction Plan | Budget Process | Nebraska
budgetprocess.unl.edu
November 11, 2025 at 4:06 AM
Friends of Stats: The U. of Nebraska-Lincoln Department of Statistics needs your help. The university's Chancellor submitted a FINAL proposal today for budget cuts (budgetprocess.unl.edu/final-budget...) that eliminates the department. 1/ #LNK #Nebraska #Huskers @amstatnews.bsky.social
Reposted by Frank Harrell
I'm participating in an "Open Science Exploratory Roundtable" in the Max Planck Society later this week, so bumping my grumpy thoughts about open science again. I'm as curious as everyone else to see what a club of competitive narcissists who succeeded under the status quo (the MPG) can manage!
How can we reform science? I have some ideas. But I am not sure you’ll like them, because they don’t promise much. elevanth.org/blog/2025/07...
Which Kind of Science Reform
What hope is there for science reform, if we can't agree on what to reform? Right now, principles are more important than practices.
elevanth.org
November 10, 2025 at 8:38 AM
I'm participating in an "Open Science Exploratory Roundtable" in the Max Planck Society later this week, so bumping my grumpy thoughts about open science again. I'm as curious as everyone else to see what a club of competitive narcissists who succeeded under the status quo (the MPG) can manage!
I just posted a critique of this paper at discourse.datamethods.org/t/critique-o... where I hope others will add their thoughts #StatsSky #EpiSky #Statistics #rct
November 9, 2025 at 3:56 PM
I just posted a critique of this paper at discourse.datamethods.org/t/critique-o... where I hope others will add their thoughts #StatsSky #EpiSky #Statistics #rct
Reposted by Frank Harrell
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. ⬇️
November 9, 2025 at 9:23 AM
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. ⬇️
Reposted by Frank Harrell
This is the next blow that red states are taking to follow along with the MAGA plan of attack. By disaffiliating as a member station, APT will not show any national PBS programming and will ultimately control every aspect of that station. What does this mean?
wbhm.org/2025/no-more...
wbhm.org/2025/no-more...
No more Elmo? APT could cut ties with PBS
The board that oversees Alabama Public Television is considering disaffiliating from PBS, ending a 55-year relationship.
wbhm.org
November 8, 2025 at 12:03 AM
This is the next blow that red states are taking to follow along with the MAGA plan of attack. By disaffiliating as a member station, APT will not show any national PBS programming and will ultimately control every aspect of that station. What does this mean?
wbhm.org/2025/no-more...
wbhm.org/2025/no-more...
Reposted by Frank Harrell
LET ME BE PERFECTLY CLEAR: DOJ DID ALL OF THIS TO TRY AND STOP PAYMENTS TO PEOPLE UNDER AN ANTI-HUNGER PROGRAM.
Oh, this is wild.
DOJ went to SCOTUS *before* we got the First Circuit's administrative stay denial. (We know this because DOJ just filed a letter at SCOTUS about the admin stay denial.)
So, DOJ went to SCOTUS b/c the First Circuit did not meet its less-than-seven-hour "deadline" for a ruling.
DOJ went to SCOTUS *before* we got the First Circuit's administrative stay denial. (We know this because DOJ just filed a letter at SCOTUS about the admin stay denial.)
So, DOJ went to SCOTUS b/c the First Circuit did not meet its less-than-seven-hour "deadline" for a ruling.
November 8, 2025 at 12:06 AM
LET ME BE PERFECTLY CLEAR: DOJ DID ALL OF THIS TO TRY AND STOP PAYMENTS TO PEOPLE UNDER AN ANTI-HUNGER PROGRAM.
Reposted by Frank Harrell
Weekend reads: Journal retracts ADHD intervention papers; ‘Science and the crisis of trust’; letters to editors surge
Weekend reads: Journal retracts ADHD intervention papers; ‘Science and the crisis of trust’; letters to editors surge
Former Australian science agency ecology researcher loses two papers Dozens of board members resign from big-data journal after mass staff firings Review mill in Italy targeting ob-gyn journals, re…
retractionwatch.com
November 8, 2025 at 1:13 PM
Weekend reads: Journal retracts ADHD intervention papers; ‘Science and the crisis of trust’; letters to editors surge
Reposted by Frank Harrell
Probably the papers I share the most often are @f2harrell.bsky.social 1996 on how to effectively develop multivariate models and @benvancalster.bsky.social 2019 paper on Calibration. Both extremely well written and informative. What are you most shared stats paper? #statistics #statssky #academicsky
November 7, 2025 at 12:12 PM
Probably the papers I share the most often are @f2harrell.bsky.social 1996 on how to effectively develop multivariate models and @benvancalster.bsky.social 2019 paper on Calibration. Both extremely well written and informative. What are you most shared stats paper? #statistics #statssky #academicsky
Reposted by Frank Harrell
Only too true 🤢
Interested in helping ICE protect the Homeland?
Check out the Lincoln Project's useful video guide!
open.substack.com/pub/therickw...
Check out the Lincoln Project's useful video guide!
open.substack.com/pub/therickw...
Your Guide To Working for ICE
Interested in joining ICE to defend the Homeland?
open.substack.com
November 4, 2025 at 11:15 PM
Only too true 🤢
Wow I really like the constant alignment while showing extra information for “don’t know” category.
This is definitely an interesting idea for display of Likert scale data allowing for don't-know / didn't-answer
Just trying out a way of showing Don't knows/Didn't answers.
Calculating the percents of Trust level and Party, then multiplying the don't know by -1 and using geom_col()
#rstats
Calculating the percents of Trust level and Party, then multiplying the don't know by -1 and using geom_col()
#rstats
November 4, 2025 at 1:16 PM
Wow I really like the constant alignment while showing extra information for “don’t know” category.
Reposted by Frank Harrell
🎨 How to create elegant #dataviz with base #rstats and the #tinyplot pkg
📈 useR! 2025 presentation by Grant McDermott @gmcd.bsky.social
www.youtube.com/watch?v=MOB-...
📈 useR! 2025 presentation by Grant McDermott @gmcd.bsky.social
www.youtube.com/watch?v=MOB-...
tinyplot: convenient and customizable base R plots - Grant McDermott
YouTube video by useR! Conference
www.youtube.com
November 4, 2025 at 7:38 AM
🎨 How to create elegant #dataviz with base #rstats and the #tinyplot pkg
📈 useR! 2025 presentation by Grant McDermott @gmcd.bsky.social
www.youtube.com/watch?v=MOB-...
📈 useR! 2025 presentation by Grant McDermott @gmcd.bsky.social
www.youtube.com/watch?v=MOB-...