Applications are now open for #MMED2024: @ICI3D's Clinic on Meaningful Modelling of Epidemiological Data. 17-29 June in South Africa, a 2‐week modelling clinic that emphasizes the use of data in understanding infectious disease dynamics. https://t.co/Teg7J9iWEl
@NateSilver538 Wait, what? Bluesky sucks because there are some annoying people with a lot of followers? I'm literally not following the argument the twitter is better here, and I would kind of like to before I commit to bluesky.
I have a bunch of links like this that all seem broken. They are https://t.co/3VAdtk35nB links, but seem to redirect correctly to https://t.co/a2rIXGkL0r. https://t.co/a2rIXGkL0r says they've never heard of them. Thoughts will be greatly appreciated.
#ICI3D is now accepting applications for our virtual-only Clinic on Dynamical Approaches to Infectious Disease Data (DAIDD), held 10-16 December. The application deadline is 20 October 2023 at 23:59 South African time. See https://t.co/udrlbhVZVp for information.
If you like any of my tweets, please consider coming over to mastodon. Not for any specific reason :-). Mastodon is a federation of instances. I'm on https://t.co/zmmoRamn8n, but there are other relevant ones as well.
We have a preprint with some simple ideas about modeling vaccination and the immune system up on @medrxivpreprint. See https://t.co/CkAMihrvlt. By @sang_woo_park with @mliwz1, @CJEMetcalf#BryanTGrenfell
@rlmcelreath To _evaluate_ the test itself, it makes more sense to sample (uniformly) from the Platonic range. If you do this for an exact test, you should get back to perfect histograms like the ones you show for the t test.
@rlmcelreath I find it useful to think about discrete-distribution P values in two steps. The "Platonic" P value for an observation is a _range_ between P(X<x) and P(X≤x). To _apply_ the P value, we conservatively choose the latter value (and double it for a two-tailed test).