@besttd.bsky.social
MD, resident in clinical virology with a passion for science.
Dilatant in stats, code, history of science.
Fell into a rabbit hole on SO in 2019 thanks to excellent replies by @f2harrell.bsky.social and @wkhuber.bsky.social that I'm still digging into.
Dilatant in stats, code, history of science.
Fell into a rabbit hole on SO in 2019 thanks to excellent replies by @f2harrell.bsky.social and @wkhuber.bsky.social that I'm still digging into.
There's a maths museum primarily aimed at children in my hometown in Germany. It's also got some stats exhibits.
Every other school field day & even some birthdays(!) led me there ...
en.wikipedia.org/wiki/Mathema...
Every other school field day & even some birthdays(!) led me there ...
en.wikipedia.org/wiki/Mathema...
February 10, 2026 at 10:15 AM
There's a maths museum primarily aimed at children in my hometown in Germany. It's also got some stats exhibits.
Every other school field day & even some birthdays(!) led me there ...
en.wikipedia.org/wiki/Mathema...
Every other school field day & even some birthdays(!) led me there ...
en.wikipedia.org/wiki/Mathema...
Love it! I've been frequently showing it but usually got some meh-response. After having been trained on the other one ...
Now that it has a proper name and is published in a journal it MUST be true!
/s
Now that it has a proper name and is published in a journal it MUST be true!
/s
December 29, 2025 at 1:11 PM
Love it! I've been frequently showing it but usually got some meh-response. After having been trained on the other one ...
Now that it has a proper name and is published in a journal it MUST be true!
/s
Now that it has a proper name and is published in a journal it MUST be true!
/s
> AIsome (2077 EN)
December 17, 2025 at 2:35 PM
> AIsome (2077 EN)
It's sad we have to let you go.
But we'll always keep your reaction video to opening a spreadsheet in good memory.
Good luck on the other side!⛵️
May many prespecified t-tests await you there.
But we'll always keep your reaction video to opening a spreadsheet in good memory.
Good luck on the other side!⛵️
May many prespecified t-tests await you there.
December 15, 2025 at 1:32 PM
It's sad we have to let you go.
But we'll always keep your reaction video to opening a spreadsheet in good memory.
Good luck on the other side!⛵️
May many prespecified t-tests await you there.
But we'll always keep your reaction video to opening a spreadsheet in good memory.
Good luck on the other side!⛵️
May many prespecified t-tests await you there.
I second this 🙏
December 14, 2025 at 10:19 PM
I second this 🙏
Isn't that the best? If a system is well designed, many things fall into place without extra steps😅
December 14, 2025 at 9:00 PM
Isn't that the best? If a system is well designed, many things fall into place without extra steps😅
agreed. As I become less puristic, I recently found myself appreciating it sometimes as a tool.
December 14, 2025 at 4:08 PM
agreed. As I become less puristic, I recently found myself appreciating it sometimes as a tool.
Interesting! I assume it's a bit pointed to simplify.
Running with the "Job changes" section, I can see some arguments emerge against the (still) dominating narrative in (central?) Europe, that one has to go the US for the sake of having been there.
Running with the "Job changes" section, I can see some arguments emerge against the (still) dominating narrative in (central?) Europe, that one has to go the US for the sake of having been there.
December 14, 2025 at 1:31 PM
Interesting! I assume it's a bit pointed to simplify.
Running with the "Job changes" section, I can see some arguments emerge against the (still) dominating narrative in (central?) Europe, that one has to go the US for the sake of having been there.
Running with the "Job changes" section, I can see some arguments emerge against the (still) dominating narrative in (central?) Europe, that one has to go the US for the sake of having been there.
Different book(s) but same story every time I try to read "Advanced R" or "Casella and Berger" cover to cover for fun on the weekend.
December 14, 2025 at 1:11 PM
Different book(s) but same story every time I try to read "Advanced R" or "Casella and Berger" cover to cover for fun on the weekend.
Sometimes psychology and urology are closer than anatomy might make you believe ...
December 14, 2025 at 1:01 PM
Sometimes psychology and urology are closer than anatomy might make you believe ...
sorry for the long delay, I was off the grid for a while.
Thank you very much for the pointer! :)
Thank you very much for the pointer! :)
December 14, 2025 at 12:55 PM
sorry for the long delay, I was off the grid for a while.
Thank you very much for the pointer! :)
Thank you very much for the pointer! :)
But somehow the ideas of Kuhn, and the more extreme Feyerabend, still haven't entered into mainstream virology (& medicine/science) so NHST and a simplified version of Popper are the default.
Supplement 2/2
Supplement 2/2
November 30, 2025 at 1:31 PM
But somehow the ideas of Kuhn, and the more extreme Feyerabend, still haven't entered into mainstream virology (& medicine/science) so NHST and a simplified version of Popper are the default.
Supplement 2/2
Supplement 2/2
... and then finally Kuhn comes along and says "the whole Popperian idea is way too simple and assumes methodological unity where there is none".
Supplement 1/2
Supplement 1/2
November 30, 2025 at 1:31 PM
... and then finally Kuhn comes along and says "the whole Popperian idea is way too simple and assumes methodological unity where there is none".
Supplement 1/2
Supplement 1/2
The coincidence of Popper and NHST seems so obvious that their relation appears almost trivial, yet I haven't read this before.
Maybe @rmcelreath.bsky.social, could you help keep my train of thought from derailing?
4/4
Maybe @rmcelreath.bsky.social, could you help keep my train of thought from derailing?
4/4
November 30, 2025 at 1:29 PM
The coincidence of Popper and NHST seems so obvious that their relation appears almost trivial, yet I haven't read this before.
Maybe @rmcelreath.bsky.social, could you help keep my train of thought from derailing?
4/4
Maybe @rmcelreath.bsky.social, could you help keep my train of thought from derailing?
4/4
Or am I seeing parallels where there are none? Because within frequentism in general one can also encourage estimation over testing (and that has been done for a long time in stats books).
3/4
3/4
November 30, 2025 at 1:29 PM
Or am I seeing parallels where there are none? Because within frequentism in general one can also encourage estimation over testing (and that has been done for a long time in stats books).
3/4
3/4
And then this is contrasted by the Bayesian approach which lends itself more easily to confirmation (or at least the encouragement of point & interval-estimation).
2/4
2/4
November 30, 2025 at 1:29 PM
And then this is contrasted by the Bayesian approach which lends itself more easily to confirmation (or at least the encouragement of point & interval-estimation).
2/4
2/4
Not Christmas related but in the Christmas edition and surely a modern classic.
I highly recommend the figure in the Introduction:
pubmed.ncbi.nlm.nih.gov/22174322/
I highly recommend the figure in the Introduction:
pubmed.ncbi.nlm.nih.gov/22174322/
Orthopaedic surgeons: as strong as an ox and almost twice as clever? Multicentre prospective comparative study - PubMed
Male orthopaedic surgeons have greater intelligence and grip strength than their male anaesthetic colleagues, who should find new ways to make fun of their orthopaedic friends.
pubmed.ncbi.nlm.nih.gov
November 27, 2025 at 7:41 PM
Not Christmas related but in the Christmas edition and surely a modern classic.
I highly recommend the figure in the Introduction:
pubmed.ncbi.nlm.nih.gov/22174322/
I highly recommend the figure in the Introduction:
pubmed.ncbi.nlm.nih.gov/22174322/
I think @f2harrell.bsky.social would suggest the ANCOVA and usually advice especially against change scores: www.fharrell.com/post/errmed/...
Statistical Errors in the Medical Literature – Statistical Thinking
This article catalogs several types of statistical problems that occur frequently in the medical journal articles.
www.fharrell.com
November 27, 2025 at 7:20 PM
I think @f2harrell.bsky.social would suggest the ANCOVA and usually advice especially against change scores: www.fharrell.com/post/errmed/...