This post came about as a follow-up to a comment @rmcelreath.bsky.social made on the sidelines of the great p-curve wars earlier this year. I wanted to see for myself just how non-uniform these things are and when.

I had a play in #rstats with the distribution of p-values under a true null hypothesis, for two-sample tests of equal proportions at different sample sizes. They're not uniformly distributed, but sometimes they are more non-uniform than other times. freerangestats.info/blog/2025/11...

"To obtain calm and quiet we had to come to New York, and there we got it in full measure. Life saunters along on an even keel. Nothing happens. Have we been mugged?" "No, sir." "Or shot by youths?" "No, sir." "No, sir," is right. We are tranquil. And I'll tell you why. There are no aunts here."

Day 4: my data #30DayMapChallenge Today I learned that #rstats 📦 `tidygeocoder` is super useful for pulling lat/long coordinates and that plotting a world map that centres the pacific is quite annoying. Shout out to @freerangestats.info for useful blog posts. jenrichmond.github.io/maps/2025-11...

Oh, yes! It ensures polygons (like Russia, NZ, Fiji, or Alaska’s Aleutian Islands) don’t get split or wrapped awkwardly around the map. 🌏

I've finally got around to curating a selection of films about the #climate crisis 🎥 There's lot of mediocre climate films out there, but for me these stand out head & shoulders above the rest 🎞️ They make excellent resources for classrooms, lecture halls, or community cinema's 🎬 Thread:🧵Plz RT

@bmj.com Please look at PubPeer comments on an article you published last week. pubpeer.com/publications... I think your research integrity dept shld act swiftly on this one, given clinical significance. I'm aware of even more evidence of problems so let me know if this is not sufficient.

I did actually look up the causes of death data late last year and blogged here: freerangestats.info/blog/2024/12.... 'Accidents, assaults and self-harm' is indeed the highest relative cause of death in USA. And the impact on life expectancy is amplified because victims are younger.

Studying sex ratios is just a lot harder than you think: effects are tiny and variation is large. statmodeling.stat.columbia.edu/2025/10/31/s...

#rstats people I'm sure I remember someone had recently developed a package or at least a hack for getting around the fact that geom_ribbon can't map the fill aesthetic to a variable (continuous or not), but now I can't find it, can anyone point me to it?

I just noticed there are 1,123 people in Australia who claim ancestral links to PItcairn (current population ~50). Anyone know if this is likely to be reasonably accurate ie plausible? From census 2021 tablebuilder.

#rstats/ #databs storytime 🧵 I have a good friend who I haven't talked to regularly for a time. While catching up with her, she lamented having to take stats courses for her PhD program (she's in archeology/anthro). I offered to help her study, and she put me off, saying she couldn't ask that. 1/n

I’m seeing some misinformation about pseudo-random number generator best practices going around the internets. Let’s talk about why the pseudo-random number generator seed you use shouldn’t actually have any impact on your results and, consequently, you can choose whatever seed you damn well please.

My department (Computer science) is recruiting a postdoc (18 months) on any topics related to trust and AI (could be a philosophical project or a more empirically minded one). Belgian postdoc salaries are competitive. To apply: jobs.unamur.be/emploi.2025-...

I virtually never rebase anything in Git. Beginning stuff should focus on 'it's like track changes for code' and the extreme basics of commit, push, pull request, merge.

Statisticians of Melbourne and environs! I will be giving a talk (the Belz Lecture) next Monday, 2pm, Melbourne Uni. The title is "Laws and Orders" The lecture is free but they want you to sign up. There's also an early career stats showcase with lunch that costs www.statsoc.org.au/event-6288035

Another great walkthru of an analysis that seems easy, but, well, most things in data analysis are only easy if you stop thinking about them after the first model (which is sadly the mode). Were I teaching data analysis -at any level- I'd certainly use these posts in class.

When an airline says a flight will have a slight delay 'due to operational reasons', what are the non-operational reasons that are ruled out by this explanation? Isn't any delay at this point by definition operational?

Countries with more economic inequality have higher homicide rates. But for any given country, the relationship isn't as strong. This poses a multilevel modelling problem that it's easy to trip over. I tripped, righted (I hope), and blogged about it with #rstats. freerangestats.info/blog/2025/10...

For years I thought the Greek and Roman obsession with public virtue was kind of a scam but now I understand why they valued it so highly, tyranny is indeed pretty bad

Teaching girls maths leads to jazz, my 1924 author in his 70s argues. (Excellent book BTW even if its judgements are sometimes a little questionable)

History is full of interesting snippets. This from a 1924 reminiscence of London in the 1850s and 1860s. I wonder what else made up Miss Holt's story.

Great to see @significancemag.bsky.social on bsky.

I was looking at some Scottish cause of death data yesterday and the extent to which transport accident deaths in young men have all but disappeared is pretty remarkable. Nice to find a good news story in this data for a change.

OK, an idea I am testing for thinking about increasing proportion of an under-represented group in a workforce. Could be women in STEM, people with disabilities anywhere, men in nursing, indigenous people in X. My thoughts are that there are perhaps four things that can be causing the problems. 1/n

Today's the 959th anniversary of the Battle of Hastings, so obviously I'm reposting my scenes from the Bayeux Tapestry in Playmobil.

Opening for an epidemiologist at my organisation, the Pacific Community (SPC). Based in Suva; not in my Statistics division but in the Public Health Division next door. Applications close 26 October Fiji time. careers.spc.int/job/epidemio...

How did qualification get two such opposite meanings? Like, you want your auditor to be qualified, but not your audit. And you can recommend Joe without qualification, because he has all the right qualifications.

A useful post on the bizarre citation cabal that is the IHME and the Global Burden of Disease study.

A fascinating blog post on the generational loss of reading & its consequences for human knowledge & society — 🔗 open.substack.com/pub/jmarriot...

Survey costs drive almost every data collection decision, but are woefully understudied. @aapor.bsky.social, @amstatnews.bsky.social, and @westatofficial.bsky.social are cosponsoring this important workshop on Survey Costs in February 2026! Submit your abstract today! aapor.org/aapor-2026-s...

This is a good quick read bringing these things together. It's easy to focus on #3 (prior information). But #4 (regularization), #5 (combining multiple sources of information) and #6 (latent data) are great practical motivations for Bayesian techniques too.

Probably a rabbit hole I'll regret, but what is going on in these guys' heads when they repeatedly talk about >100% discounts? My best guess is "reduce to the point that a 650% increase would be needed to bring it back to current price" ie to 1/7.5 of current. Is this a common kids' mistake?

Bajillionaires of the world, we don't need more employment growth nowcasts during the government shutdown. But we do need a very-large-sample monthly survey of US households to determine employment / unemployment status. macromostly.substack.com/p/we-dont-ne...

We're #hiring a demographer at Pacific Community-SPC to help Pacific Island countries and territories with their population statistics! Based in Fiji, we're seeking someone with a relevant Masters or PhD, 10+ years experience, and strong skills in #rstats or Stata. careers.spc.int/job/statisti...

2/ but I think social media is a good place to practice clear & concise communication of work you think is important (your own & others). And of course it has the chance to amplify the reach & understanding of that science much more than the readership of academic journals.

What movies are good, but badly tarnished by appalling sequels? Sequels that are not just 'not as good as the first movie', but bad enough to be basically a betrayal of the whole original concept. My list for starters: Jaws, Rocky, First Blood, The Matrix.

📣Calling all Pacific early career researchers📣 Want to gain experience working, living and conducting science at sea while exploring key issues affecting your region? 🚢🌊. Applications for the Floating University 2026 voyage. ⏰Applications open until 30 November. Apply here 👉 bit.ly/4nXODH6

Thread is worth reading.

I drew a map with #rstats of the locations involved in the Ameila Earhart / Fred Noonan disappearance in 1937. The sequence of events required for there to be anything helpful and new in US archives are pretty implausible. Most likely they fell in the sea, sadly. freerangestats.info/blog/2025/10...

The western Pacific in 1939 as seen by ChatGPT. American Samoa and the Marshalls appear to be west of Western Australia. Philippines has been mistaken for the Marianas and wrongly shown as Japanese-controlled. France seems in control of all 'Papua New Caledonia'.

In all the endless discussions of LLMs, there’s a point that is, on one level, obvious, but that I feel does not get sufficiently foregrounded: LLMs are transforming material that people have put up on the internet.

I think a Reacher + The Accountant remake is needed - a drifter statistician with an outrageous physique moving from one random hospital to the next, breaking up clinical research cartels and applying high levels of violence to restore appropriate rigour to studies in their early phases of set-up.

I restrained from a sarcastic reply-all of "I don't know if I'm available for this lunchtime 12:30 meeting as there is no indication which time zone you are talking about" so will vent here instead. It is offensive in the extreme to omit this when you have 'International' in your org's title.

OCCRP has identified the son of Palau's senate president among those arrested in last week's massive drug sweep in Fiji: www.occrp.org/en/news/son-...

One thing about this exchange on X is that they are talking past eachother. They both have their own weird idea about WWII ('the gods didn't consent', 'Churchill tricked Hitler into war' - he wasn't even in govt!) and think they are agreeing with eachother, but really these are different ideas.

Good article on the insane economics of the AI bubble. With rapidly obsolete physical assets and unit costs that dont scale, there is simply no way all the over-investment can generate anything like the revenue needed to make it pay. #econsky

The worst thing about notebooks isn’t that they’re an incentive to write horrendously unstructured code, but that they displace proper IDEs. Lots of data science tooling is built with a mindset of “data scientists only use Jupyter”. Plz I just want directories and plain-text files

Netflix has been targeting me with adverts for this show, which I had thought (almost certainly wrongly) was because they had inferred (how?) I was a statistician. But seems unlikely based on these reports.