(Even weirder is the literature declaring ORs fine in certain circumstances in which they closely match RRs. Like come on, why not try to calculate the metric you actually prefer)

By all accounts, the founding father of (causal) Epidemiology was a Slow Scientist who was singularly focused on the pursuit of truth. The more I read about John Snow, the more convinced I am that he deserves his own museum! (Text from "The Medical Detective" by Sandra Hempel)

Every time you speak up - and every time you don't - there are people watching and learning about you and the world you help create.

"We are seeing readily available data assets, such as... the UK Biobank, NHANES, and many other databases, being weaponized to generate manuscripts with minimal scholarly value and utility"

Improvements to methods should lead to more robust evidence and more rapid advances in knowledge. In their Perspective, Marcus Munafò & @mendelrandom.bsky.social look at why this hasn't been the case and what can be done about it #AcademicSky 🧪

Every time you speak up - and every time you don't - there are people watching and learning about you and the world you help create.

By all accounts, the founding father of (causal) Epidemiology was a Slow Scientist who was singularly focused on the pursuit of truth. The more I read about John Snow, the more convinced I am that he deserves his own museum! (Text from "The Medical Detective" by Sandra Hempel)

The cure for ignorance is not information. It's humility and curiosity. Facts can be easily dismissed. What motivates people to gain insight is recognizing gaps in their understanding and wanting to find out more. The root of lifelong learning is knowing how little we know.

A variation: Scientists who claim they’re “not interested in causality” because they assume the term only applies to deterministic, law-like relationships that are unrealistic in their field. Instead, they’re interested in how “X drives Y”, the effects of X, the “extent to which X matters for Y”>

A masterpiece

Neyman is my hero

tipping my hat to @pwgtennant.bsky.social whose paper just stopped me doing something very stupid. academic.oup.com/ije/article/...

One of the biggest contributors to scientific scepticism among laypeople, IME, is these idiotic nutritional epi studies that come out with conflicting results every year and are uncritically picked up by credulous journos looking for clickbait. Get in the 🗑️. #episky #statsky

Fascinating talk by @oacarah.bsky.social on selection-augmented causal diagrams in fascinating @societyforepi.bsky.social mid-year meeting symposium on selection bias arranged by @haidonglu.bsky.social!. Blown away by the quality and clarity of thinking 🤯

Also: shifting scientific funding towards "government priority areas" is a surefire way to kill creativity and innovation. Generating scientific knowledge is not like ordering a takeaway. You can't just tick the 'science' that you want and expect anything useful.

The UK *is* losing a generation of scientists. I know lots of brilliant people who have left their jobs / the country because of the limited jobs & funding. Once lost, they cannot ever be replaced. The UK government is overseeing the death of UK academic science, and it doesn't seem to care.

Striking paper from researchers at Anthropic using a randomised control trial to look at the effects of AI use on skills acquisition. TL:DR ‘We found that using AI assistance led to a statistically significant decrease in mastery.’ www.anthropic.com/research/AI-...

To me it's another symptom of spending 98% of our time teaching people how to answer questions and 2% of our time teaching people how to ask questions.

This might be a good study to examine with my causal inference class 🤔

I tried to take a photo of a grasshopper on my windshield, but now it looks like a gigantic bug destroying the town.

When AI was added to a tool for sinus surgery: “Cerebrospinal fluid leaked from one patient’s nose. In another… a surgeon mistakenly punctured the base of a patient’s skull. In two other cases, patients suffered strokes after a major artery was accidentally injured” www.reuters.com/investigatio...

"All diseases are 100% genetic and 100% environmental" - Rothman and Greenland

The UK *is* losing a generation of scientists. I know lots of brilliant people who have left their jobs / the country because of the limited jobs & funding. Once lost, they cannot ever be replaced. The UK government is overseeing the death of UK academic science, and it doesn't seem to care.

Hundreds of early career researchers have warned the UK will lose a generation of scientists after the announcement of significant cuts to physics projects and research facilities. #AcademicSky #ECRs

Speaking of @societyforepi.bsky.social, the SER mid-year meeting is taking place (virtually) this week with an exciting 4-day programme of cutting-edged epidemiology! SER members should check out the programme and drop in on the fascinating sessions! epiresearch.org/mid-year-mee...

⚠️ Despite all the hype, chatbots still make terrible doctors. Out today is the largest user study of language models for medical self-diagnosis. We found that chatbots provide inaccurate and inconsistent answers, and that people are better off using online searches or their own judgment.

One of the biggest wastes of academic productivity is the variation in page/word limits across journals in the same/related fields.

"Metascience sandpit: scientometrics for research assessment" Looks like UKRI are leaning into the WORST possible vision for meta-science as some kind of scientific bean counting. Here with the awful intention of using AI to judge research quality in #REF2029 www.ukri.org/opportunity/...

I'm on the edge of the "R Statistics Community" in what looks like an @societyforepi.bsky.social gathering on causal inference with @lizstuart.bsky.social, @jeremylabrecque.bsky.social, @jlrohmann.bsky.social, @idiaz.bsky.social, @mattpfox.bsky.social, @miguelhernan.org, & @noahgreifer.bsky.social

Astonishing thread/blog on the rotten state of the @ucu.org.uk - a trade union of all things. I resent every pound of my membership that's spent on subverting democracy, bullying staff, and covering-up wrongdoing.

Ahead of the 10 February Certification Officer hearing, I’ve written about why Ewan McGaughey and I took this route. This is a regulatory process, not a court case. Ewan and I will represent ourselves. UCU has hired a King’s Counsel. vickyblakeucu.uk/202... /1

This kind of characterful human art is so truly beautiful

However, I do think UK epi programmes need to think about this more. Most PhD programmes in epidemiology don't even have formal training, so you need to rely on having had good Master's level training.

I think these are good competencies to aim for, and it is overall a good idea to think about what the core competencies should be, as long as the list doesn't contain things only attainable by training at elite (US) institutions...

Worth expanding to a multi-perspective view from employers, graduates, epi program directors & faculty, etc. We would want to hear from epidemiologists in government, industry, etc.

I wonder what proportion of epidemiologists worldwide would have these 'core' competencies'? Are the following really core: * Consistent estimators * "Super populations" * Synthetic control designs * Sentinel health events * FINER criteria * D-separation rules * Single world intervention graphs

Defining methodologic and other core competencies for PhD-level training in epidemiology academic.oup.com/aje/article/...

If you ask (and then answer) a scientific question the answer to which is unknown, then it is novel. Novelty is all or none, so an index is not required. How interesting the question is makes no difference.

I’m from the North East and when I lived in Chicago in the 1990’s I was regularly asked if I was Australian

An American guy in an airport in the US once said "you're from London, ain'tcha?". I thought about trying to explain where Barnsley is in relation to London, and all the differences. Instead I just said "yeah, near enough".

It is ridiculous, that they think the solution to simplistic quality assessments is adding more metrics. Here is my metascientist approach to see if someone is any good: read their work.

like the piper? like the piper who picked a peck of pickled peppers?

From now on, if anyone in the USA asks my name, I am William-George. Said in a faux upper-class English accent.

Welcome to Americaning 101. Today's class is about navigating breakfast. We'll start with coffee, briefly think about the nationality of muffins, the what to expect from a pancake, then finish with a deep dive into eggs and how to order them.

Happens within the US, too, for strong regional accents (South, New England, parts of NYC, for example). "Strong accent", of course, is a bit class-defined (though perhaps not to the degree in the UK). Took me a a few months in New England to get used to the local pronunciation and so on.

I was not prepared for the relentless commenting on my Northern English accent. Not: "Ooo, I love your British accent"! More: "Where is that accent from? It doesn't sound like normal British. Are you a bit Scottish?" Even: "Are you from Liverpool, you sound a bit like <celebrity from Liverpool>?"

Yeah, I immediately imagined the chain where this metric gets built into the Elsevier submission pipeline and either you pay and it tells you your novelty score and/or editors see it on submission and make decisions. And either way they train AI on it and sell that AI to help you write/edit papers.

This is the perfect way to move. Humans have it WRONG. #SealSky 🦭 1/4th speed

Same, I read their invitation up to "partnered with Elsevier". Great heuristic to detect bullshit.