Michael Plank

A person in Washington State who kept backyard poultry has died from #H5N5 #flu. The person was the first known infection with this subtype of flu globally & the second recorded death in the US from an #H5 flu virus. Health authorities say there's no evidence the person spread the virus to others.

Very interesting perspective! I often have the opposite thought: "Experimenting like a modeller"
Perhaps the two could co-exist for better workflows. Extensive exploration on simulated data should be the norm before running experiments to make sure the design is capable of the intended inference.

We should not be in a situation where experts are having to debunk cherry-picked claims about vaccines on the CDC website, but here we are.

I wrote more previously about the evidence base around MMR and vaccination, and why people need to make *testable* hypotheses, rather than slipperly claims that endlessly evolve after being falisified... kucharski.substack.com/p/a-wild-vac...

More data outlining how COVID vaccines reduce the risk of hospitalizations & death, especially among those aged 65 and older.

This aligns nicely with recommendations from Canada's National Advisory Committee on Immunization. 🇨🇦

Link: tinyurl.com/yuwmmtt4 by Hansen et al.

Statisticians are increasingly posed with thought-provoking and even paradoxical questions, challenging our qualifications for entering the statistical paradises created by Big Data. By developing measures for data quality, this article suggests a framework to address such a question: “Which one should I trust more: a 1% survey with 60% response rate or a self-reported administrative dataset covering 80% of the population?” A 5-element Eulerformula-like identity shows that for any dataset of size n, probabilistic or not, the difference between the sample average Xn and the population average XN is the product of three terms: (1) a data quality measure, ρR,X, the correlation between Xj and the response/recording indicator Rj ; (2) a data quantity measure, √(N − n)/n, where N is the population size; and (3) a problem difficulty measure, σX, the standard deviation of X. This decomposition provides multiple insights: (I) Probabilistic sampling ensures high data quality by controlling ρR,X at the level of N −1/2; (II) When we lose this control, the impact of N is no longer canceled by ρR,X, leading to a Law of Large Populations (LLP), that is, our estimation error, relative to the benchmarking rate 1/√n, increases with √N; and (III) the “bigness” of such Big Data (for population inferences) should be measured by the relative size f = n/N , not the absolute size n; (IV) When combining data sources for population inferences, those relatively tiny but higher quality ones should be given far more weights than suggested by their sizes.

Does everyone know about this one already? Meng 2018. “Statistical Paradises and Paradoxes in Big Data ...” Annals of Applied Statistics doi.org/10.1214/18-A...

Covid Inquiry hit by yet another top-level resignation Inquiry boss quits only months into the job - the second executive director to resign this year

Yes, yes ... of course we can have complete confidence in the report findings of this inquiry set up for nakedly partisan purposes, given terms of reference designed to avoid criticism of a currently-governing party, and from which multiple professionals have run a mile. Why on earth would you ask?

Global Society for Infectious Disease Dynamics (GSIDD) | infectious disease dynamics GSIDD serves the global community of researchers, practitioners, and educators in infectious disease dynamics.

1/ 📰 Fantastic to see the official website for the Global Society for Infectious Disease Dynamics (GSIDD) is now live! #GSIDD #IDsky #EpiSky 🧪

🗺️ GSIDD serves the global community of researchers, practitioners and educators in infectious disease dynamics.

🔗: www.gsidd.org

Confidence interval discussion time! The perfect opportunity to repost this blog post answering the question you haven’t dared to ask: www.the100.ci/2024/12/05/w...

Substantial overview of HPAI H5 cases in humans, 1997-2005. Huge effort in data curation and cleaning, bring together a really clear and comprehensive story.
👉 academic.oup.com/nsr/advance-...

For all those men who, on International Women's Day decry "oh but when is it ever the men's day".... it's today mate. November 19.

Happy International Men's Day.

Why are there so many ways to compute a frequentist interval for a proportion? Because the sample space is discrete, coverage can be garbage without adjustment. The Agresti-Coull interval ("score" below) is really a Bayes interval with a weak prior, has good freq coverage! >

Yes indeed

Woah is that jump from 2023 -> 24 the scrapping of the clean car discount? 😮

"Government to raise cost and damage of driving".

First they weakened the fuel efficiency standards, now they're gutting the penalties for not meeting them. Basically a voluntary standard now. Penalty rates Aus: A$100/gCO2; EU: E95/gCO2; NZ: NZ$15/gCO2.
www.rnz.co.nz/news/politic...

I saw this and it made me think maybe the spike is reflecting increase in domestic violence, which I think was seen in other countries too (but in the US tends to involve firearms more than elsewhere)

bsky.app/profile/mega...

The importance of excessive mortality assumptions...

Left hand plot: 2017-19 average used as baseline. Conclusion: Sweden is 'best' with negative excess mortality.

Right hand plot: 2017-19 linear trend extrapolated as baseline. Conclusion: Sweden is 'worst' with huge positive excess mortality.

Ecology and spread of the North American H5N1 epizootic - Nature The panzootic of highly pathogenic H5N1 since 2021 was driven by around nine introductions into the Atlantic and Pacific flyways, followed by rapid dissemination through wild migratory birds, primaril...

Our lab's paper describing the North American H5N1 epizootic is out now in Nature! So thrilled to have this out, and congratulations to @lambod50.bsky.social for all the fantastic work on this: www.nature.com/articles/s41...

Epidemic Momentum Infectious disease outbreaks have precipitated a profusion of mathematical models. We introduce a unifying concept of "epidemic momentum" -- prevalence weighted by the capacity to infect in the future...

Nice preprint by David Earn and Todd Parsons (H/T @jmccaw.bsky.social) showing how the SIR model can be related to a more general class of renewal process models that don't assume exponentially distributed infectious period.
arxiv.org/abs/2511.01939

Musing: Universities that subsist on selling expensive degrees to international students via fully online delivery will be wrecked when LLMs undercut the trust that students actually earned those degrees. Institutional reputation will suffer where in-person instruction/assessment isn't the norm

enables some of the analytical machinery that works on the simple SIR case to be applied to other models. E.g. we know intuitively that SIR & SEIR share certain properties such as final epidemic size. This work formalises that relationship & extends it to a more general class of models. Neat!

Key concept is what they call 'epidemic momentum', which is essentially a generalisation of prevalence, weighted by future transmission potential. This quantity can be calculated directly from an incidence time series and

Epidemic Momentum Infectious disease outbreaks have precipitated a profusion of mathematical models. We introduce a unifying concept of "epidemic momentum" -- prevalence weighted by the capacity to infect in the future...

Nice preprint by David Earn and Todd Parsons (H/T @jmccaw.bsky.social) showing how the SIR model can be related to a more general class of renewal process models that don't assume exponentially distributed infectious period.
arxiv.org/abs/2511.01939

Canada has lost its measles elimination status...a sad milestone.

Now is the time to double down on lowering barriers to routine childhood immunizations and ensuring Canadians have access to credible, science-based information.

Link: tinyurl.com/c8nfxvk3

Canada loses measles elimination status — as does the entire Americas region Canada has formally lost its measles elimination status, the country’s public health agency announced, triggering the loss of that status throughout all of the Americas.

As expected, Canada has lost its measles-free status because of the long-running outbreak there. With it goes the #measles elimination status of the entire zone of the Americas, the only division of the #WHO to ever have achieved measles elimination. www.statnews.com/2025/11/10/m...

Liked this: "In conclusion, thinking about modelling as experimentation does not cage creativity; it structures it, giving us better tools to ask sharper questions, present clearer answers and build cumulative science."

Nice. Another one for the reading list!

"Universities teach us to ask questions without tidy answers. They train us to look for truths we didn’t know existed and to challenge assumptions we didn’t realise we held. At their best, they remind us that intellect isn’t about having opinions, it’s about earning them."

This, from OUSA's President, is very good and warrants reading:

"For decades, governments of every stripe have treated universities as businesses to be managed rather than institutions of learning and curiosity. ... Students are rebranded as customers, research, degrees and graduates as outputs."

Modelling Like an Experimentalist Dahlin et al. (2024) apply experimental thinking to a model of mosquito-borne disease transmissions.

"Validate With Simulated Truth: A first habit is to test whether an analytical pipeline can recover known conditions."

Very good advice below. So much COVID nonsense (e.g. 'immunological dark matter') basically came down to a non-identifiable model that hadn't been properly tested.