Against my better instincts, I have written some notes on how human probability judgements work and what you should expect from surveys that ask people to guess what proportion of the population is transgender. I hope never to speak of this matter again

this is going to end badly isn't it?

i swear to god i am about *this* close to writing a blog post explaining the basics of probability judgement and why people need to stop resharing that fucking yougov survey. it’s driving me batshit crazy seeing how their misleading write up spreads like wildfire on this horrible website

changing the topic. soothing myself by rewatching Foundation season 2 and um. Hober Mallow. um.

It’s taken me a few hours to work out why I am so angry at the fact that this particular claim goes viral over and over again. It is not true. This is not what people believe. But why am I, a trans woman with specific relevant expertise, so anger at the quoted post? A thread.

breaking news: when people are forced to estimate percentages for something they know absolutely nothing about other than “i’ve heard of it but it’s probably small”, the average response is about 20%. film at 11

sometimes i think it is a curse that i have a social media account and also have the misfortune to have expertise in statistics as well as the measurement of human behaviour. i want to scream, every single day

ah. that’s why this crossed my timeline. no, they don’t. the survey is badly designed and the numbers mean nothing. signed, a trans woman who actually worked in that fucking field.

apropos of nothing, i was just struck by the fact that there are two qualitatively different versions of the “capitalism is unfair” claim: one version asserts that it’s *corrupt* - the wrong people become billionaires. the other version asserts that it is *immoral* - there should be no billionaires

sigh. conferences going on in multiple parts of the world, in different disciplines. everyone sharing happy photos, telling stories. thinking it would be nice to travel, to participate. but it’s unsafe, and it’s hard not to feel bitter when i remember that other people aren’t restricted in this way

gauge. is it guage? or gague? gouge? gayge? who knows. it could be anything. ridiculous word

dear future danielle: no, the problem isn’t with lsoda. the problem is *never* lsoda. i don’t care what the error message says. the problem is that there’s an upstream bug in your model specification. no need to thank me, just try not to waste two hours looking in the wrong place next time ok?

sigh. of course. www.kenklippenstein.com/p/fbi-readie...

i miss the days, long past now, when i felt some sense of self-worth. it was never a strength, but over the last 10 years that what little self-confidence i had in my skills and abilities has completely evaporated. i can’t really remember what that felt like anymore, just that it wasn’t like this :/

for once i have managed to write a short blog post on a lazy sunday arvo, picking up on a suggestion from @coolbutuseless.fosstodon.org.ap.brid.gy for generating procedural palettes in #rstats. it works better than i expected blog.djnavarro.net/posts/2025-0...

I like this post because it explains distributional regression so clearly. It's not magic folks, just more regressions.

It’s been a little overshadowed by the slightly unexpected mental health issues that hit me very hard yesterday afternoon, but for what it’s worth here’s the GAMLSS regression post I’ve been working on for the past month #rstats blog.djnavarro.net/posts/2025-0...

May we all write as spitefully and beautifully as @djnavarro.net writes about splines here 🙏 READ THIS NOW

At the risk of stretching the imagination beyond the limits of human endurance, the reader is asked to pretend that the author is interested in splines #rstats

i have finally found a need to use rlang::env_clone() in the wild. please kill me

(wailing Homer Simpson voice) there’s so much I don’t know about P-splines!!!!

I’m especially proud of this article I wrote about Gaussian Processes for the Recast blog! 🥳 GPs are super interesting, but it’s not easy to wrap your head around them at first 🤔 This is a medium level (more intuition than math) introduction to GPs for time series. getrecast.com/gaussian-pro...

I didn’t start as an AI hater. I was wary but not opposed if it could be done with care. But after seeing how irresponsibly and unethically tech companies built and used LLMs my views have hardened. It didn’t have to be like this, but it *is* like this now. AI is bad, and I hate it.

It’s a surprisingly common stance I see among cis progressives on Bluesky. Trans people are expected to support politicians who despise us and advocate for our exclusion from society, on “lesser evil” grounds. I don’t see any reason why trans people should be expected to acquiesce to this

Ruin a book by adding Tom Clancy: Tom Clancy’s Modern Applied Statistics With S

This thread is hard reading for me. I’ve often found myself cast in the role of “the respectable trans woman”, the nice non-threatening one that cis people can use as an example. There’s a pressure to do exactly what Evan describes, to become an advocate for assimilation. It is deeply unpleasant.

rewatching Deadwood for the first time since the start of the pandemic, and fuck me the smallpox episodes hit so much harder now

it’s very surreal to have the experience of being trapped in bed most of the time and yet the cause of it isn’t depression, it’s just exhaustion (specifically, mild case of long covid). it’s a *completely* different feeling

she smol

i hate code i hate statistics i hate science …and i especially hate *data*

new #rstats art series posted

slightly damaged

I wrote a thing about substituting LLMs for survey respondents notstatschat.rbind.io/2025/08/15/i...

tech in 2025 source: mastodon.social/@nclm/115021...

eating a book today to teach the other books a lesson

oh

TIL that the thing i was taught in year 12 was not in fact the Riemann integral it is the Darboux integral and i feel betrayed

This is a pretty good read, albeit a depressing one. It’s remarkable to me that there is an actual war on science going on right now, and the only thing that Richard Dawkins et al want to talk about is how it’s unscientific to allow people like me to use the bathroom in peace.

Super excited to curate the @weare.rladies.org account this week. Follow me over there for data viz inspo, some things I think are magic, as well as chat about burnout, navigating career change and showing up for the #rstats community. Thanks for the nudge @pipinghotdata.com !

New paper with @richarddmorey.bsky.social now out in JASA, where we critically examine p-curve. Below is Richard’s excellent summary of the many poor statistical properties of p-curve (with link to paper). I wanted to add some conceptual issues that we also tackle in the paper.

There are more than a few accounts on here that are definitely cis people and whose actions probably make it harder to treat cis people with respect and dignity. JD Vance comes to mind?

my brain: you know, i’ve been feeling better lately, why don’t we go for a walk up to the shops? my body, *not* in fact recovered from long covid: the fuck did you just say bitch????

it’s saturday morning. i am bored, i am curious to find out if i can set my own laptop on fire, so i’m refitting an old gamlss regression model of mine using a data set with 10x as many rows as the original. what could possibly go wrong 😬

Just a standard reminder that statistics is a stand-alone discipline because it’s really hard and it is very easy to get catastrophically wrong. (This example is p-curves, beloved of the reproducibility/scientific fraud warriors, absolutely indefensible in practice)

i’m mindful of the fact that i’m no longer an academic and should probably stay out of this discussion but — this is what frustrates me the most. tools and procedures are being pushed very hard by “reform advocates”, and those tools don’t work. they shouldn’t have been published in the first place

Paper drop, for anyone interested in #metascience, #statistics, or #metaanalysis! @clintin.bsky.social and I show in a new paper in JASA that the P-curve, a popular forensic meta-analysis method, has deeply undesirable statistical properties. www.tandfonline.com/doi/full/10.... 1/?

I saw this elsewhere on bluesky and idk why but I just started crying. This is where we are at: governments around the world are passing Jim Crow style segregation laws for people like me, and everyone I know just ignores it. The apathy is somehow worse than the hate

good night all

somehow it never ceases to amaze me how differently i am treated post-transition. for the most part, its worse. but not entirely: men smile at me in conversation now, and i like that