Cathal Walsh
@cathalwalsh.bsky.social
Professor of Biostatistics, Trinity College Dublin. There’s more to counting than numbers. people.tcd.ie/walshc Born c. 335ppm. Small wheel cyclist. Biostatistician.
Reposted by Cathal Walsh
"..the first known human H5N5 infection has captured a good deal of attention, but the `other' flu story we've been following this week has been the rapid spread of a new seasonal H3N2 subcade (K) virus, which emerged in the Southern Hemisphere"
afludiary.blogspot.com/2025/11/cdc-...
afludiary.blogspot.com/2025/11/cdc-...
CDC FluView, ECDC CDT & Canada Announce Early H3N2 Subclade K Prevalence Numbers
Credit NIAID #18,947 Yesterday's announcement for the first known human H5N5 infection has captured a good deal of attention, but the `ot...
afludiary.blogspot.com
November 15, 2025 at 8:40 PM
"..the first known human H5N5 infection has captured a good deal of attention, but the `other' flu story we've been following this week has been the rapid spread of a new seasonal H3N2 subcade (K) virus, which emerged in the Southern Hemisphere"
afludiary.blogspot.com/2025/11/cdc-...
afludiary.blogspot.com/2025/11/cdc-...
Reposted by Cathal Walsh
If you're a clinician or researcher in Ireland feeling a little overwhelmed by statistics, join myself & @cathalwalsh.bsky.social on 11th October in SJH for a 6-credit CPD Essential Medical Statistics, where we'll demystify things together. Tickets here!
www.eventbrite.ie/e/essential-...
www.eventbrite.ie/e/essential-...
Essential Medical Statistics and Evidence Appraisal
Learn how to interpret and apply essential medical statistics and evidence in a practical and engaging way at our upcoming event!
www.eventbrite.ie
September 30, 2025 at 1:28 PM
If you're a clinician or researcher in Ireland feeling a little overwhelmed by statistics, join myself & @cathalwalsh.bsky.social on 11th October in SJH for a 6-credit CPD Essential Medical Statistics, where we'll demystify things together. Tickets here!
www.eventbrite.ie/e/essential-...
www.eventbrite.ie/e/essential-...
August 13, 2025 at 3:49 PM
This is outrageous but not without precedent
‘Stick to your principles and to the rules at all times. Never compromise. If you compromise in anything, you will not feel well with yourself, and it will create a liability that will come back to haunt you.’
rss.onlinelibrary.wiley.com/doi/full/10....
‘Stick to your principles and to the rules at all times. Never compromise. If you compromise in anything, you will not feel well with yourself, and it will create a liability that will come back to haunt you.’
rss.onlinelibrary.wiley.com/doi/full/10....
August 2, 2025 at 12:00 AM
This is outrageous but not without precedent
‘Stick to your principles and to the rules at all times. Never compromise. If you compromise in anything, you will not feel well with yourself, and it will create a liability that will come back to haunt you.’
rss.onlinelibrary.wiley.com/doi/full/10....
‘Stick to your principles and to the rules at all times. Never compromise. If you compromise in anything, you will not feel well with yourself, and it will create a liability that will come back to haunt you.’
rss.onlinelibrary.wiley.com/doi/full/10....
Reposted by Cathal Walsh
Leader Schumer: "What does a bad leader do when they get bad news? Shoot the messenger. That's just what happened with the head of the Bureau of Labor Statistics."
August 1, 2025 at 8:35 PM
Leader Schumer: "What does a bad leader do when they get bad news? Shoot the messenger. That's just what happened with the head of the Bureau of Labor Statistics."
Reposted by Cathal Walsh
It is very unfortunate that the US will no longer have any reliable federal statistics until Trump is gone.
Even the best private job statistics, ADP, tends to be far off from the more comprehensive Bureau of Labor Statistics numbers. These statistics cannot be privatized.
Even the best private job statistics, ADP, tends to be far off from the more comprehensive Bureau of Labor Statistics numbers. These statistics cannot be privatized.
And we're officially at the end of being able to trust the jobs numbers in this country ever again.
August 1, 2025 at 9:23 PM
It is very unfortunate that the US will no longer have any reliable federal statistics until Trump is gone.
Even the best private job statistics, ADP, tends to be far off from the more comprehensive Bureau of Labor Statistics numbers. These statistics cannot be privatized.
Even the best private job statistics, ADP, tends to be far off from the more comprehensive Bureau of Labor Statistics numbers. These statistics cannot be privatized.
Reposted by Cathal Walsh
🚨 Trump is demanding the firing of the Commissioner of Labor Statistics, Erika McEntarfer, because the jobs report was bad.
When you can’t rig the numbers, fire the messenger. Classic authoritarian move.
When you can’t rig the numbers, fire the messenger. Classic authoritarian move.
August 1, 2025 at 6:56 PM
🚨 Trump is demanding the firing of the Commissioner of Labor Statistics, Erika McEntarfer, because the jobs report was bad.
When you can’t rig the numbers, fire the messenger. Classic authoritarian move.
When you can’t rig the numbers, fire the messenger. Classic authoritarian move.
Reposted by Cathal Walsh
Donald Trump is firing the head of the Bureau of Labor Statistics right after they reported weak jobs numbers. If firing someone in charge of statistics because the objective statistics make you look bad isn't some authoritarian nonsense, I don't know what is.
Trump fires labor statistics boss hours after weak jobs report
The president implied that the BLS commissioner, long-time federal employee Erika McEntarfer, manipulated the data "for political purposes."
www.nbcnews.com
August 1, 2025 at 6:46 PM
Donald Trump is firing the head of the Bureau of Labor Statistics right after they reported weak jobs numbers. If firing someone in charge of statistics because the objective statistics make you look bad isn't some authoritarian nonsense, I don't know what is.
Reposted by Cathal Walsh
BREAKING: President Trump fires the head of the Bureau of Labor Statistics hours after weak jobs report.
Trump fires labor statistics boss hours after weak jobs report
The president implied that the BLS commissioner, long-time federal employee Erika McEntarfer, manipulated the data "for political purposes."
nbcnews.to
August 1, 2025 at 6:40 PM
BREAKING: President Trump fires the head of the Bureau of Labor Statistics hours after weak jobs report.
Reposted by Cathal Walsh
The head of the Bureau of Labor Statistics during Trump’s first admin.
August 1, 2025 at 9:00 PM
The head of the Bureau of Labor Statistics during Trump’s first admin.
Reposted by Cathal Walsh
How does the risk of death change as we age?
July 31, 2025 at 3:24 PM
How does the risk of death change as we age?
Reposted by Cathal Walsh
We often now hear the term “Bayesian borrowing” in trials. But what is Bayesian borrowing and what are the pros and cons around its use? 1/6
#MethodologyMonday #122
#MethodologyMonday #122
July 21, 2025 at 7:10 AM
We often now hear the term “Bayesian borrowing” in trials. But what is Bayesian borrowing and what are the pros and cons around its use? 1/6
#MethodologyMonday #122
#MethodologyMonday #122
Reposted by Cathal Walsh
We’ve all been in meetings like this
on.ft.com/3GS1Ykp
on.ft.com/3GS1Ykp
July 18, 2025 at 5:48 AM
We’ve all been in meetings like this
on.ft.com/3GS1Ykp
on.ft.com/3GS1Ykp
But, look, he is speaking publicly, and they are saying nothing to him. Do the authorities really know that this is the Christ?
July 5, 2025 at 10:01 PM
But, look, he is speaking publicly, and they are saying nothing to him. Do the authorities really know that this is the Christ?
Nice commentary @stephensenn.bsky.social - needs as much attention as the original article!
June 29, 2025 at 11:24 AM
Nice commentary @stephensenn.bsky.social - needs as much attention as the original article!
Reposted by Cathal Walsh
Reposted by Cathal Walsh
XKCD has a Tukey quote for Tukey's 110th birthday
xkcd.com/3104/
Many of his specific inventions have been made obsolete over time, but he was a major contributor to data analysis and visualisation being serious and respectable parts of statistics research
xkcd.com/3104/
Many of his specific inventions have been made obsolete over time, but he was a major contributor to data analysis and visualisation being serious and respectable parts of statistics research
Tukey
xkcd.com
June 19, 2025 at 7:04 AM
XKCD has a Tukey quote for Tukey's 110th birthday
xkcd.com/3104/
Many of his specific inventions have been made obsolete over time, but he was a major contributor to data analysis and visualisation being serious and respectable parts of statistics research
xkcd.com/3104/
Many of his specific inventions have been made obsolete over time, but he was a major contributor to data analysis and visualisation being serious and respectable parts of statistics research
Does anyone else have problems with CAPTCHAs? Is the back of a traffic light not a traffic light? We live in a world where Completely Automated Public Turing tests tell Computers and Humans Apart but don’t!
June 16, 2025 at 9:31 AM
Does anyone else have problems with CAPTCHAs? Is the back of a traffic light not a traffic light? We live in a world where Completely Automated Public Turing tests tell Computers and Humans Apart but don’t!
What’s the evidence base here anyone know? And what are the regulatory aspects? It’s a curious ad I’ve seen a few times in a number of places.
May 13, 2025 at 3:44 PM
What’s the evidence base here anyone know? And what are the regulatory aspects? It’s a curious ad I’ve seen a few times in a number of places.
Reposted by Cathal Walsh
I cannot fucking cope with this world of bullshitters.
May 10, 2025 at 9:32 PM
I cannot fucking cope with this world of bullshitters.
Reposted by Cathal Walsh
“Lots of good statistical methods make use of two models…
Perhaps, rather than trying to systematize all statistical learning into a single inferential framework, we would be better off embracing our twoishness” statmodeling.stat.columbia.edu/2011/06/20/t... via @seabbs.bsky.social
Perhaps, rather than trying to systematize all statistical learning into a single inferential framework, we would be better off embracing our twoishness” statmodeling.stat.columbia.edu/2011/06/20/t... via @seabbs.bsky.social
The pervasive twoishness of statistics; in particular, the “sampling distribution” and the “likelihood” are two different models, and that’s a good thing | Statistical Modeling, Causal Inference, and...
statmodeling.stat.columbia.edu
May 9, 2025 at 6:50 AM
“Lots of good statistical methods make use of two models…
Perhaps, rather than trying to systematize all statistical learning into a single inferential framework, we would be better off embracing our twoishness” statmodeling.stat.columbia.edu/2011/06/20/t... via @seabbs.bsky.social
Perhaps, rather than trying to systematize all statistical learning into a single inferential framework, we would be better off embracing our twoishness” statmodeling.stat.columbia.edu/2011/06/20/t... via @seabbs.bsky.social
Reposted by Cathal Walsh
It is not only ok for you to say when you don't understand something, it's critical. Experts know their limits, almost by definition. Novices get swept up in hype.
April 22, 2025 at 5:32 AM
It is not only ok for you to say when you don't understand something, it's critical. Experts know their limits, almost by definition. Novices get swept up in hype.