Robin Blythe
@rbly.bsky.social
Assistant professor at Duke-NUS medical school. Mostly interested in health economics, biostats and clinical informatics.
Good thread. My concern is, CI is presented as a solution to bad stats when in reality it suffers from many of the same limitations. We can almost never be certain of most reqs in any case -- exogeneity; data integrity; missingness mechanisms; selection bias to name a few (1/2)
#statsky #econsky
#statsky #econsky
Some half-formed thoughts on causal inference bc I've been reading about it very casually. (fucking yes, managed to spell them both right first go)
I get similar feelings to when I was reading about how the replication crisis would be solved by pre-registration and so
I get similar feelings to when I was reading about how the replication crisis would be solved by pre-registration and so
January 4, 2026 at 10:30 PM
When you lack the primary data to fulfill the objectives on a grant you've been added to: do a simulation on the computer
#statsky
#statsky
When in doubt (in teaching and in research) do a simulation on the computer.
statmodeling.stat.columbia.edu/2026/01/04/w...
statmodeling.stat.columbia.edu/2026/01/04/w...
When in doubt (in teaching and in research) do a simulation on the computer. | Statistical Modeling, Causal Inference, and Social Science
statmodeling.stat.columbia.edu
January 4, 2026 at 9:26 PM
When you lack the primary data to fulfill the objectives on a grant you've been added to: do a simulation on the computer
#statsky
#statsky
Saloni's superb summary underscores how badly most media report scientific progress. Most of these advancements will reduce morbidity and mortality soon, globally! AI prediction...not so much. Why are our media institutions so bad at this? Are reporters bad? Are they overruled by bad editors?
New post!
There was a lot of innovation in medicine and biomedical research this year, and I've tried to summarize the biggest ones in this blogpost.
Medical breakthroughs in 2025. Plus a serious note at the end.
www.scientificdiscovery.dev/p/medical-br...
There was a lot of innovation in medicine and biomedical research this year, and I've tried to summarize the biggest ones in this blogpost.
Medical breakthroughs in 2025. Plus a serious note at the end.
www.scientificdiscovery.dev/p/medical-br...
Medical breakthroughs in 2025
... and a happy new year.
www.scientificdiscovery.dev
December 28, 2025 at 11:59 PM
Saloni's superb summary underscores how badly most media report scientific progress. Most of these advancements will reduce morbidity and mortality soon, globally! AI prediction...not so much. Why are our media institutions so bad at this? Are reporters bad? Are they overruled by bad editors?
Anyone ever been on a paper submission as a middle author where you don't believe the findings/the data are too weak and don't support the conclusions? Especially if it was a causal inference study. How did you handle it? Were you worried about offending colleagues? #statsky #episky #medsky
December 22, 2025 at 9:12 PM
Sadly many supervisors view grad students (in health at least) as a resource to boost publication count rather than as a mentee and future colleague. If you respect your students, your first priority is to support their learning, rather than their supporting your KPIs.
Anyway, I really feel like the whole environment, including the state of the published literature, is really colluding against students learning how to do proper research. That’s a shame; I always have the impression our students really do want to learn how to do things the right way.
December 20, 2025 at 8:30 PM
Sadly many supervisors view grad students (in health at least) as a resource to boost publication count rather than as a mentee and future colleague. If you respect your students, your first priority is to support their learning, rather than their supporting your KPIs.
In all seriousness, it's not til you start reading the utterly bullshit attributions in the papers that cite your work that you realise scientists legitimately cannot read above a grade 6 level
#statsky
#statsky
December 18, 2025 at 11:11 PM
In all seriousness, it's not til you start reading the utterly bullshit attributions in the papers that cite your work that you realise scientists legitimately cannot read above a grade 6 level
#statsky
#statsky
I'm new to APC models, & as usual @dingdingpeng.the100.ci has written an excellent paper on them re: the ID problem. I'm on a project using APC for predicting future effects of diff policies, but it seems like the same inference problems apply to prediction (1/3) journals.sagepub.com/doi/10.1177/...
Sage Journals: Discover world-class research
Subscription and open access journals from Sage, the world's leading independent academic publisher.
journals.sagepub.com
December 18, 2025 at 10:58 PM
I'm new to APC models, & as usual @dingdingpeng.the100.ci has written an excellent paper on them re: the ID problem. I'm on a project using APC for predicting future effects of diff policies, but it seems like the same inference problems apply to prediction (1/3) journals.sagepub.com/doi/10.1177/...
Reposted by Robin Blythe
Don't miss this unique opportunity to network with peers and be part of shaping the future of healthcare delivery.
📅Key Dates:
📌 Call for Abstracts: Opening Soon! February 2026
📌 Earlybird Registration: Opening Soon! March 2026
📌 Program Announced: Expected July 2026
#HSRAANZ26
📅Key Dates:
📌 Call for Abstracts: Opening Soon! February 2026
📌 Earlybird Registration: Opening Soon! March 2026
📌 Program Announced: Expected July 2026
#HSRAANZ26
December 16, 2025 at 11:01 PM
Don't miss this unique opportunity to network with peers and be part of shaping the future of healthcare delivery.
📅Key Dates:
📌 Call for Abstracts: Opening Soon! February 2026
📌 Earlybird Registration: Opening Soon! March 2026
📌 Program Announced: Expected July 2026
#HSRAANZ26
📅Key Dates:
📌 Call for Abstracts: Opening Soon! February 2026
📌 Earlybird Registration: Opening Soon! March 2026
📌 Program Announced: Expected July 2026
#HSRAANZ26
WHY do people ask you for a 5-7 point response scale, then during analysis just dichotomise the responses? Why make respondents deal with the cognitive burden of actually thinking through their beliefs when you're just going to squash it down anyway? I see it every day! It's stupid! #statsky #episky
December 10, 2025 at 2:01 AM
WHY do people ask you for a 5-7 point response scale, then during analysis just dichotomise the responses? Why make respondents deal with the cognitive burden of actually thinking through their beliefs when you're just going to squash it down anyway? I see it every day! It's stupid! #statsky #episky
Hot take: privacy as the biggest concern re: AI in medicine is a red herring. People put their lurid personal details into ChatGPT every 5 minutes. Making it all about privacy -> maintaining big funding for models that nearly always fail to cost-effectively improve patient outcomes. #medsky #statsky
December 8, 2025 at 5:44 AM
Hot take: privacy as the biggest concern re: AI in medicine is a red herring. People put their lurid personal details into ChatGPT every 5 minutes. Making it all about privacy -> maintaining big funding for models that nearly always fail to cost-effectively improve patient outcomes. #medsky #statsky
I am going to avoid QTing as the intention is not to dunk. However complaints that stats is "too hard" and there should be tools to automate it miss the mark. Stats is a profession that requires hard work, like most others. They ARE hard. That's why you partner with experts.
#statsky #rstats #medsky
#statsky #rstats #medsky
December 5, 2025 at 7:22 AM
Interested in perspectives on this. Let's say you have an inexperienced colleague you've just started working with. They start on a project and complete everything incredibly quickly. Not only that, but the analysis is incredibly nuanced. They aren't a statistician but seemingly know everything(1/3)
November 29, 2025 at 4:15 AM
Interested in perspectives on this. Let's say you have an inexperienced colleague you've just started working with. They start on a project and complete everything incredibly quickly. Not only that, but the analysis is incredibly nuanced. They aren't a statistician but seemingly know everything(1/3)
Reposted by Robin Blythe
Are peer reviewers influenced by their work being cited? Version of record at
@elife.bsky.social. Thorough and useful peer review - who needs and impact factor?!
Links to paper and code/data ⬇️
📄https://elifesciences.org/reviewed-preprints/108748
💻https://github.com/agbarnett/cited_reviewers
@elife.bsky.social. Thorough and useful peer review - who needs and impact factor?!
Links to paper and code/data ⬇️
📄https://elifesciences.org/reviewed-preprints/108748
💻https://github.com/agbarnett/cited_reviewers
elifesciences.org
November 29, 2025 at 2:24 AM
Are peer reviewers influenced by their work being cited? Version of record at
@elife.bsky.social. Thorough and useful peer review - who needs and impact factor?!
Links to paper and code/data ⬇️
📄https://elifesciences.org/reviewed-preprints/108748
💻https://github.com/agbarnett/cited_reviewers
@elife.bsky.social. Thorough and useful peer review - who needs and impact factor?!
Links to paper and code/data ⬇️
📄https://elifesciences.org/reviewed-preprints/108748
💻https://github.com/agbarnett/cited_reviewers
As a peer reviewer, if I see the authors have just paid lip service to the previous round of comments, I'd vastly prefer the editor to just reject the paper rather than sending it back for me to say the exact same thing twice. It's demeaning and a waste of everyone's time.
#academia #statsky
#academia #statsky
November 28, 2025 at 4:56 AM
Nice thread over on the DataMethods discussion forum (an invaluable resource for me, personally!) on differentiating between conditional and average treatment effects
discourse.datamethods.org/t/propensity...
#statsky #episky
discourse.datamethods.org/t/propensity...
#statsky #episky
Propensity score weights
Question for @ngreifer Considering ATE, ATT, ATO and ATM weights, it is commonly said that a weighted estimator (ATE/ATT/ATO/ATM) targets a marginal estimand — an average over a specified population...
discourse.datamethods.org
November 26, 2025 at 2:38 AM
Nice thread over on the DataMethods discussion forum (an invaluable resource for me, personally!) on differentiating between conditional and average treatment effects
discourse.datamethods.org/t/propensity...
#statsky #episky
discourse.datamethods.org/t/propensity...
#statsky #episky
Reposted by Robin Blythe
IHEA members - the largest ever Board election in history is now open. I’m standing for the role of President Elect. More about why is here healtheconomics.org/paula-lorgelly/ #votepaula
Paula Lorgelly – International Health Economics Association
healtheconomics.org
November 25, 2025 at 5:14 AM
IHEA members - the largest ever Board election in history is now open. I’m standing for the role of President Elect. More about why is here healtheconomics.org/paula-lorgelly/ #votepaula
Just boosting this solution if you want to do some brms model bootstrapping. It's a very simple model but the whole thing comes together in just a couple minutes and it's very easy to code up
#rstats #statsky #bayes
#rstats #statsky #bayes
Basically, fit the model with chains = 0, then create an update() function with recompile = FALSE and the sampling specs you want. I've usually had success with {{furrr}} for the parallelisation part. Can't take full credit for this one as I needed an LLM to show me how the recompile step worked.
November 21, 2025 at 4:45 AM
I'm pretty sure this joke has been made (probably by @zachweinersmith.bsky.social over on SMBC) but I'm starting to wonder whether my inability to prove anything empirically using mathematics and relying on simulation every time is an education issue, a me issue, or both
#statsky #rstats
#statsky #rstats
November 20, 2025 at 4:06 AM
I'm pretty sure this joke has been made (probably by @zachweinersmith.bsky.social over on SMBC) but I'm starting to wonder whether my inability to prove anything empirically using mathematics and relying on simulation every time is an education issue, a me issue, or both
#statsky #rstats
#statsky #rstats
Reposted by Robin Blythe
My sister is having a rough time with her Master's thesis, which was graded by a rude external reviewer. Is it normal for a thesis to go out for a letter grade? This seems particularly cruel and uncommon given a thesis is a 200+ page research document. How can you give it a grade?
#academicsky
#academicsky
November 17, 2025 at 1:35 PM
My sister is having a rough time with her Master's thesis, which was graded by a rude external reviewer. Is it normal for a thesis to go out for a letter grade? This seems particularly cruel and uncommon given a thesis is a 200+ page research document. How can you give it a grade?
#academicsky
#academicsky
This has me wondering whether a perfect replication study can actually be bad - wasting time and resources on something we strongly suspect is effective when we could be using those funds to feel out the edges of that knowledge or doing something novel.
#metascience #statsky #medsky
#metascience #statsky #medsky
Right! And assuming that (i) what we can learn from a study is limited to what's captured by a p-value, and (ii) a close enough replication of that p-value is the only way science can validate existing evidence.
November 16, 2025 at 3:16 AM
This has me wondering whether a perfect replication study can actually be bad - wasting time and resources on something we strongly suspect is effective when we could be using those funds to feel out the edges of that knowledge or doing something novel.
#metascience #statsky #medsky
#metascience #statsky #medsky
New paper: we validated the Clinical Frailty Scale for predicting 90-day mortality in frail older adults. We show that the scale's predictive utility is low, but argue that mortality risk shouldn't be what catalyses end-of-life care discussions anyway
doi.org/10.1136/bmjo...
#statsky #medsky #episky
doi.org/10.1136/bmjo...
#statsky #medsky #episky
Validation of the Clinical Frailty Scale for predicting 90-day mortality in hospitalised older adults screened as at risk of nearing the end of life in Queensland, Australia: a multisite observational...
Background The Clinical Frailty Scale is an ordinal scale from 1 (very fit) to 9 (terminally ill) commonly used to assess frailty in older patients. It is simple for clinicians to apply and can help i...
doi.org
November 14, 2025 at 1:12 AM
New paper: we validated the Clinical Frailty Scale for predicting 90-day mortality in frail older adults. We show that the scale's predictive utility is low, but argue that mortality risk shouldn't be what catalyses end-of-life care discussions anyway
doi.org/10.1136/bmjo...
#statsky #medsky #episky
doi.org/10.1136/bmjo...
#statsky #medsky #episky
It would be nice if this forces people to recognise that the point of a degree is learning how to think deeply/problem solve, rather than get a highly paid job or put some letters after your name. Not everyone needs to go to university!
Relying on LLMs to get your diploma and is similar to forging PhD research for boosting your political career in Germany
You think it creates shortcuts and notability for getting the top job. But as you are only interested in the CV line, you don't learn anything.
And one day, you are exposed.
You think it creates shortcuts and notability for getting the top job. But as you are only interested in the CV line, you don't learn anything.
And one day, you are exposed.
November 10, 2025 at 9:19 AM
It would be nice if this forces people to recognise that the point of a degree is learning how to think deeply/problem solve, rather than get a highly paid job or put some letters after your name. Not everyone needs to go to university!