Max
@oostermax.bsky.social
PhD student. Data scientist adopted by environmental epidemiologists. Interested in causal inf, modelling, meta-science, epi, policy etc. Sports enthusiast. Blog at maxoosterwegel.com
Reposted by Max
opencausal.org is online!
Open Causal is an open platform to share, discuss and reuse causal graphs. Join now as a beta user, help shaping the future of causal graphs!
Open Causal is an open platform to share, discuss and reuse causal graphs. Join now as a beta user, help shaping the future of causal graphs!
Open Causal (Beta)
opencausal.org
November 13, 2025 at 6:59 AM
opencausal.org is online!
Open Causal is an open platform to share, discuss and reuse causal graphs. Join now as a beta user, help shaping the future of causal graphs!
Open Causal is an open platform to share, discuss and reuse causal graphs. Join now as a beta user, help shaping the future of causal graphs!
Reposted by Max
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.
An investigation has identified more than 1,500 research articles produced by a network of Ukrainian companies that could be one of Europe’s largest paper mills
go.nature.com/3VxA1SD
go.nature.com/3VxA1SD
Europe’s largest paper mill? 1,500 research articles linked to Ukrainian network
A group of companies has flooded 380 journals with hundreds of suspect papers since 2017.
go.nature.com
September 6, 2025 at 7:00 AM
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.
Reposted by Max
First they came for the immigrants, and I did not speak out (because the message test performed at a low percentile rank compared to other messages in our testing bank for non college men who opt into online surveys through various consumer reward programs)
okay ryan, but does it raise or lower trump's approval rating among people who sign up to take surveys in a digital ad-testing production system? that is the only question that matters
August 25, 2025 at 1:39 AM
First they came for the immigrants, and I did not speak out (because the message test performed at a low percentile rank compared to other messages in our testing bank for non college men who opt into online surveys through various consumer reward programs)
Reposted by Max
My reason for sharing this is not that editors make wrong decisions, but that such long delays in publications also delay other people from building on work, or could lead to people just giving up and working on other projects instead. That seems like it would've been very unfortunate in this case
August 17, 2025 at 9:10 AM
My reason for sharing this is not that editors make wrong decisions, but that such long delays in publications also delay other people from building on work, or could lead to people just giving up and working on other projects instead. That seems like it would've been very unfortunate in this case
Reposted by Max
TIL the original paper describing CRISPR, by Francisco Mojica, was rejected by 4 journals and took 2 years to be published
![CRISPR as a microbial immune system
In 2003, Mojica wrote the first paper suggesting that CRISPR was an innate microbial immune system. The paper was rejected by a series of high-profile journals, including Nature, Proceedings of the National Academy of Sciences, Molecular Microbiology and Nucleic Acids Research, before finally being accepted by Journal of Molecular Evolution in February, 2005.[3][4]](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:73aicoale2mqwrap63rdzwep/bafkreibu3dburfn527r3sbu2ov6y7gebv6zopaznagnfytjtsn6y4k7cs4@jpeg)
August 17, 2025 at 4:00 AM
TIL the original paper describing CRISPR, by Francisco Mojica, was rejected by 4 journals and took 2 years to be published
Reposted by Max
You cannot be truly creative if it is not safe to fail.
Yet modern academic science requires constant short term success.
To survive, you must win grants & 'deliver' on those grants by publishing lots of papers.
That leads to a culture of low risk, iterative research oversold as 'groundbreaking.
Yet modern academic science requires constant short term success.
To survive, you must win grants & 'deliver' on those grants by publishing lots of papers.
That leads to a culture of low risk, iterative research oversold as 'groundbreaking.
Some say you either have "it" or you don't, and that creativity cannot be learned, but that really misses the point. Creativity is lost in the wrong environment. We need to recognize the "night science" aspect of our work and that we need to nurture it in our training. www.nature.com/articles/d41...
Can creativity in science be learnt? These researchers think so
Time pressure gets in the way of ideas. Developing ‘creative oases’ and small grants for risky ideas can encourage innovative thinking in science.
www.nature.com
August 14, 2025 at 4:48 AM
You cannot be truly creative if it is not safe to fail.
Yet modern academic science requires constant short term success.
To survive, you must win grants & 'deliver' on those grants by publishing lots of papers.
That leads to a culture of low risk, iterative research oversold as 'groundbreaking.
Yet modern academic science requires constant short term success.
To survive, you must win grants & 'deliver' on those grants by publishing lots of papers.
That leads to a culture of low risk, iterative research oversold as 'groundbreaking.
Reposted by Max
Reminder that all three books I've co-authored are freely available online for non-commercial use (and the fourth will be, too)
All three books I've co-authored are freely available online for non-commercial use:
- #Bayesian Data Analysis, 3rd ed (aka BDA3) at stat.columbia.edu/~gelman/book/
- #Regression and Other Stories at avehtari.github.io/ROS-Examples/
- Active Statistics at avehtari.github.io/ActiveStatis...
- #Bayesian Data Analysis, 3rd ed (aka BDA3) at stat.columbia.edu/~gelman/book/
- #Regression and Other Stories at avehtari.github.io/ROS-Examples/
- Active Statistics at avehtari.github.io/ActiveStatis...
August 11, 2025 at 5:44 PM
Reminder that all three books I've co-authored are freely available online for non-commercial use (and the fourth will be, too)
Reposted by Max
New blog post, on 52 pieces of science writing I've learned from over the last two years.
Thank you, from one reader, to all the authors! blog.jacobtrefethen.com/science-writ...
Thank you, from one reader, to all the authors! blog.jacobtrefethen.com/science-writ...
Science writing from the last two years that stuck with me
This list consists of writing that ticked two boxes: 1) did I think about the article again more than a week after reading it?, and 2) was it written for a popular audience? So, academic papers, inclu...
blog.jacobtrefethen.com
June 10, 2025 at 7:35 AM
New blog post, on 52 pieces of science writing I've learned from over the last two years.
Thank you, from one reader, to all the authors! blog.jacobtrefethen.com/science-writ...
Thank you, from one reader, to all the authors! blog.jacobtrefethen.com/science-writ...
Reposted by Max
Every month, I hear of at least one project where the primary data entry is lousy with errors, often because the staff collecting/entering the data got no training on data management.
I spend so much time teaching ppl how to sample 10k-D Hamiltonian manifolds. I should spend more time teaching RDM
I spend so much time teaching ppl how to sample 10k-D Hamiltonian manifolds. I should spend more time teaching RDM
February 26, 2025 at 7:52 AM
Every month, I hear of at least one project where the primary data entry is lousy with errors, often because the staff collecting/entering the data got no training on data management.
I spend so much time teaching ppl how to sample 10k-D Hamiltonian manifolds. I should spend more time teaching RDM
I spend so much time teaching ppl how to sample 10k-D Hamiltonian manifolds. I should spend more time teaching RDM
Reposted by Max
Currently reading Michael Nielsen's essay and I am completely hooked. Only 15% in and it'll take me a week for a first pass (day job), but I'm already finding concepts (like 'recipes for ruin') that I didn't have words to describe until now.
New essay: "How to be a wise optimist about science and technology?"
michaelnotebook.com/optimism/ind...
michaelnotebook.com/optimism/ind...
How to be a wise optimist about science and technology?
michaelnotebook.com
December 2, 2024 at 6:43 PM
Currently reading Michael Nielsen's essay and I am completely hooked. Only 15% in and it'll take me a week for a first pass (day job), but I'm already finding concepts (like 'recipes for ruin') that I didn't have words to describe until now.
Reposted by Max
Stop teaching students how to convince people their question is important unless you've already taught them how to find important questions.
November 30, 2024 at 12:16 PM
Stop teaching students how to convince people their question is important unless you've already taught them how to find important questions.
Reposted by Max
Remarkable
I had zero luck with grants. People really didn’t get why what I did was important and never figured out how to explain it to them.
December 1, 2024 at 8:50 AM
Remarkable
Reposted by Max
tl;dr The problem with science isn’t how we talk about to Aunt Flo and Uncle Earl at Thanksgiving - it’s that we’ve knowingly made its central purpose career advancement at the expense of advancing the public good.
Basic scientists continue to completely miss the mark in identifying the real threat to basic science. The erosion of funding for basic research does not stem from a loss of public support - the public already love discovery research.
This is a way to frame basic science when we're talking to family, friends, voters. How ridiculous is it to study fly eyes or worm vulvae or yeast life? This lead to understanding cancer genes and enabled personalized medicine. Then, thank everyone listening to us for funding NIH and NSF
December 1, 2024 at 4:13 PM
tl;dr The problem with science isn’t how we talk about to Aunt Flo and Uncle Earl at Thanksgiving - it’s that we’ve knowingly made its central purpose career advancement at the expense of advancing the public good.
Reposted by Max
People who get big grants but then don't produce quality work are performing *poorly*, not well. And people who don't have grants but do produce high quality research are performing *well*.
Yet we have an incentive structure that rewards the first of these people more than the second
Yet we have an incentive structure that rewards the first of these people more than the second
November 30, 2024 at 2:06 PM
People who get big grants but then don't produce quality work are performing *poorly*, not well. And people who don't have grants but do produce high quality research are performing *well*.
Yet we have an incentive structure that rewards the first of these people more than the second
Yet we have an incentive structure that rewards the first of these people more than the second
Reposted by Max
It is faintly horrific to me the way that obtaining research funding is treated as a goal and achievement in and of itself. Just getting the grant can be enough to supercharge a career.
This is horrific because grants are not the *goal*, they are the *means*. The goal is quality research
This is horrific because grants are not the *goal*, they are the *means*. The goal is quality research
November 30, 2024 at 2:05 PM
It is faintly horrific to me the way that obtaining research funding is treated as a goal and achievement in and of itself. Just getting the grant can be enough to supercharge a career.
This is horrific because grants are not the *goal*, they are the *means*. The goal is quality research
This is horrific because grants are not the *goal*, they are the *means*. The goal is quality research
Reposted by Max
Science is great but academia not so much. Earlier this week I revised my talk "Science as Amateur Software Development". The unprofessional way that academics curate & process data would benefit from professional habits of software engineers and chefs too 🧪 www.youtube.com/watch?v=8qzV...
Science as Amateur Software Development (2023 edition)
Software is both a cause of unreliable research and part of the solution. The bulk of scientific research relies upon specialized software for data managemen...
www.youtube.com
September 14, 2023 at 2:02 PM
Science is great but academia not so much. Earlier this week I revised my talk "Science as Amateur Software Development". The unprofessional way that academics curate & process data would benefit from professional habits of software engineers and chefs too 🧪 www.youtube.com/watch?v=8qzV...
Reposted by Max
This 100%. I don't see how you can be a substantive if you're not up on the statistical methods relevant to your field. You are, at best, and expert on the optimistic rhetoric other researchers use in their Abstracts and Discussion sections.
I don't understand how you can read and understand an evolving literature without keeping up with methodological developments.
What are we supposed to do, just read discussion sections and take people's word for it?
What are we supposed to do, just read discussion sections and take people's word for it?
December 1, 2024 at 12:49 AM
This 100%. I don't see how you can be a substantive if you're not up on the statistical methods relevant to your field. You are, at best, and expert on the optimistic rhetoric other researchers use in their Abstracts and Discussion sections.
Reposted by Max
If I’m up against the word limit here, the first words to go are those that convey uncertainty and nuance. Those words should be free!
November 21, 2024 at 10:39 AM
If I’m up against the word limit here, the first words to go are those that convey uncertainty and nuance. Those words should be free!
Reposted by Max
Something that's been bugging me for a while in bioinformatics data analysis is this overreliance on packages, workflows and what's been called "cargo cult science".
Can we have more conceptual thinking, more theory?
Asking for what we really want to achieve and what we need to do gets us there.
Can we have more conceptual thinking, more theory?
Asking for what we really want to achieve and what we need to do gets us there.
November 19, 2024 at 10:58 AM
Something that's been bugging me for a while in bioinformatics data analysis is this overreliance on packages, workflows and what's been called "cargo cult science".
Can we have more conceptual thinking, more theory?
Asking for what we really want to achieve and what we need to do gets us there.
Can we have more conceptual thinking, more theory?
Asking for what we really want to achieve and what we need to do gets us there.
Reposted by Max
I often hear that stating research questions as causal is "promising too much".
Is a research question a promise to give a fully certain answer? Or is it just a statement of what it is we want to know?
Is a research question a promise to give a fully certain answer? Or is it just a statement of what it is we want to know?
September 9, 2024 at 9:44 AM
I often hear that stating research questions as causal is "promising too much".
Is a research question a promise to give a fully certain answer? Or is it just a statement of what it is we want to know?
Is a research question a promise to give a fully certain answer? Or is it just a statement of what it is we want to know?
Reposted by Max
A Dictionary of Epidemiology:
new, 7th. edition (2024).
Call to submit contributions.
Deadline: 30 November 2023.
#EpiSky #Epidemiology
new, 7th. edition (2024).
Call to submit contributions.
Deadline: 30 November 2023.
#EpiSky #Epidemiology
September 27, 2023 at 5:10 AM
A Dictionary of Epidemiology:
new, 7th. edition (2024).
Call to submit contributions.
Deadline: 30 November 2023.
#EpiSky #Epidemiology
new, 7th. edition (2024).
Call to submit contributions.
Deadline: 30 November 2023.
#EpiSky #Epidemiology