Why I cheered in the cinema when I saw the film Pressure - great lessons on uncertainty, contested advice, and forecasting methods davidspiegelhalter.substack.com/publish/post/221177660 www.youtube.com/watch?v=eYa8N3R2cZw
David Spiegelhalter
Sir David John Spiegelhalter (born 16 August 1953) is a British statistician and a Fellow of Churchill College, Cambridge. From 2007 to 2018 he was Winton Professor of the Public Understanding of Risk in the Statistical Laboratory at the University of Cambridge. Spiegelhalter is an ISI highly cited researcher. More on Wikipedia
-
David Spiegelhalter
@d_spiegel
-
David Spiegelhalter
@d_spiegel
Ahead of the Commons vote tomorrow, some very personal thoughts on the statistics of assisted dying davidspiegelhalter.substack.com/p/statistics-and-assisted… , largely based on the recent report from Dignity in Dying www.dignityindying.org.uk/why-we-need-change/why-legalisi…
-
David Spiegelhalter
@d_spiegel
and currently available on Kindle for 99p!!
David Spiegelhalter
@d_spiegel
I am beside myself with joy that The Art of Uncertainty is out in paperback today! Fits in any (largish) pocket, easy to read on the train or while jammed into a steaming underground carriage. The perfect present for the nerd in your life.
www.amazon.co.uk/Art-Uncertainty-Navigate-Ignorance-Pelic… -
David Spiegelhalter reposted
Sense about Science
@senseaboutsci
❗️Nominations are open❗️
The 2026 John Maddox Prize, supported by @royalsociety is now open for nominations!
Celebrate researchers who have championed evidence when the powerful (and sometimes the public) didn't want to listen.
Nominate or enter👇️
johnmaddoxprize.awardstage.com -
David Spiegelhalter
@d_spiegel
I am beside myself with joy that The Art of Uncertainty is out in paperback today! Fits in any (largish) pocket, easy to read on the train or while jammed into a steaming underground carriage. The perfect present for the nerd in your life.
www.amazon.co.uk/Art-Uncertainty-Navigate-Ignorance-Pelic… -
David Spiegelhalter
@d_spiegel
www.theguardian.com/science/2026/may/30/trial-of-multi-ca… This seems a predictable failure, as 2021 data showed the test only picked up 28% of Stage 1/2 cancers. How much did this study of 140,000 patients cost? And the list price for detecting one cancer using GRAIL is $200,000 - was it ever feasible for NHS?
-
David Spiegelhalter
@d_spiegel
Great winners of the Harding Prize for Trustworthy Communication in 2025. Tom Whipple at The Times, and IN hub tool from Rebecca Coniam-Gudka, Abby Russell and Charlotte Kelman at the University of Exeter. Congratulations!!! wintoncentre.github.io/harding-prize/ www.linkedin.com/posts/science-media-centre_free-or-paid-…
-
David Spiegelhalter
@d_spiegel
Kennedy supporting Trump's maths: a great example for teaching school students about percentage change. www.youtube.com/watch?v=jdceeoDmQzc
-
David Spiegelhalter
@d_spiegel
Delighted to see the Covid Inquiry used one of our Winton Centre graphics in their vaccine report (p 215). But they chose the 'medium' exposure (February), rather than the more relevant 'low' exposure (March). I think this is misleading.
David Spiegelhalter
@d_spiegel
Their choice meant they could say ‘For all age groups, the potential benefits outweighed the potential harms’, but this was not true in early April. The appropriate graphic clearly shows why under-30s were offered alternative vaccines.
-
David Spiegelhalter
@d_spiegel
Spring is here, so it's back to frittering away my remaining life by doing wooden jigsaws face downwards. This one was tricky, even with the grain to follow, as there were so few interlocking pieces
-
David Spiegelhalter
@d_spiegel
I was so pleased to be part of a great team writing a commentary on the draft FDA guidelines on Bayesian clinical trial designs. Should have been done 30 years ago, but better late than never
JAMA
@JAMA_current
💬 Perspective: Bayesian methodology supports efficient, flexible clinical trial designs; FDA guidance encourages prespecification and sensitivity analysis of priors.
@f2harrell @d_spiegel
jamanetwork.com/journals/jama/fullarticle/2847011?guestAc… -
David Spiegelhalter
@d_spiegel
My blog for OSR about the revised Code of Practice, and the importance of trustworthiness
Office for Statistics Regulation
@StatsRegulation
In our latest guest blog, leading British statistician, @d_spiegel explores why trustworthiness, not trust, should be at the heart of statistical communication 👇
osr.statisticsauthority.gov.uk/blog/dont-ask-for-trust-in… -
David Spiegelhalter
@d_spiegel
www.ucl.ac.uk/mathematical-physical-sciences/events/2026/… Looking forward to speaking with @jburnmurdoch at this event next Tuesday in London
-
David Spiegelhalter
@d_spiegel
OSR questioning the Covid Inquiry's claim that modelling 'established' 23,000 fewer deaths if mandatory lockdown imposed a week earlier on March 16th. In fact this estimate is not only very uncertain, but also assumes voluntary measures were brought forward a week, to March 9th
Office for Statistics Regulation
@StatsRegulation
We have today written to the Covid-19 inquiry regarding the presentation of statistical modelling in the Executive Summary of its Module 2 report👇 osr.statisticsauthority.gov.uk/correspondence/presentatio…
David Spiegelhalter
@d_spiegel
This number received huge media coverage, but was estimating something very different from what was claimed. Personally, I believe that much of the benefit would have come from starting 'advisory' lockdown on March 9th. But I suspect this will become a 'zombie' statistic...
-
David Spiegelhalter
@d_spiegel
I've had fun helping BBC Ideas produce this animation, which attempts to explain Bayesian ideas in 4 minutes. I hope you like it - I really like the animators style. www.bbc.co.uk/videos/c78ee072lnpo
David Spiegelhalter
@d_spiegel
Oh dear, realised that Bayes theorem isn't written correctly in the video - embarrassing that I missed this. Will be sorted soon. Hope it doesn't spoil your enjoyment.
David Spiegelhalter
@d_spiegel
error now corrected!
-
David Spiegelhalter
@d_spiegel
The Harding Prize for Trustworthy Communication is now open for entries until 15th February. We are looking for examples that inform people engagingly, without trying to persuade them what to think or do. Each Harding Prize is worth £3,141.59... wintoncentre.github.io/harding-prize/
-
David Spiegelhalter
@d_spiegel
www.bbc.co.uk/news/articles/cwywezx86nko I will be saving £250 a month in future, as Abi helps keep my locally advanced prostate cancer under control. I can afford it, but absurd it's taken so long for NHS to pay for this very effective drug. Hats off to campaigners and @BBCHughPym for coverage
-
David Spiegelhalter
@d_spiegel
BBC Ideas kindly gave me the TikTok treatment for this video www.bbc.co.uk/videos/cdxwvezy6g4o
-
David Spiegelhalter
@d_spiegel
my blog for MHRA on then role of regualation in the age of personalised medicine
www.gov.uk/government/news/sir-david-spiegelhalter-regula… -
David Spiegelhalter
@d_spiegel
some lessons learned from our contribution to the Infected Blood Inquiry rss.org.uk/policy-campaigns/policy-groups/statistics-unde…
-
David Spiegelhalter
@d_spiegel
These are excellent winners! The pros and cons of breeding pythons to eat. Can we believe some claims made about stem cell treatments for Parkinson's? "If it sounds too good to be true, it probably is". Balanced, trustworthy information, delivered in an engaging way
Sense about Science
@senseaboutsci
🤔Is snake meat a viable solution to saving our planet? Is stem cell therapy about to cure Parkinson's?
Our 2024 #hardingprize for Trustworthy Communication winners show that science reporting works best when cutting through the hype.
More information senseaboutscience.org/activities/2024-harding-prize-shows… -
David Spiegelhalter
@d_spiegel
osr.statisticsauthority.gov.uk/blog/demand-on-the-nhs-is-… excellent blog by OSR's Ed Humpherson on the importance of looking at the variability in health metrics rather than just their average - and the tools available for doing this
-
David Spiegelhalter
@d_spiegel
nextbigideaclub.com/magazine/make-smarter-decisions-unpre… Fun but challenging to compress a whole book (The Art of Uncertainty) into a 10-minute talk. Hope I got some main points over
-
David Spiegelhalter
@d_spiegel
I've been really enjoying the Past, Present, Future series on revolutions, and then I was asked to be in one myself! So here I am discussing the Bayesian revolution - 250 years old and still going strong www.ppfideas.com
-
David Spiegelhalter
@d_spiegel
A sad coincidence. 8.32 am om March 10, Health and Safety Executive send e-bulletin warning of risks of collision with fixed objects by distracted navigators www.hse.gov.uk/safetybulletins/risk-collision-offshore-in… An hour later at 9.48 am, cargo ship rams anchored tanker in North Sea www.bbc.co.uk/news/articles/c9q4129reqgo
-
David Spiegelhalter
@d_spiegel
In uncertain times....
W. W. Norton & Company
@wwnorton
Sparkling with wit and fascinating real-world examples, @d_spiegel's THE ART OF UNCERTAINTY is an essential guide to navigating uncertainty while also retaining the humility to admit what we don’t, or simply cannot, know.
In stores now.
wwnorton.com/books/the-art-of-uncertainty -
David Spiegelhalter
@d_spiegel
I had fun doing an interview for BBC Science Focus Instant Genius podcast www.youtube.com/watch?v=ownnsYAtWc8
-
David Spiegelhalter
@d_spiegel
nominations close on friday! Any more entries for this great competition?
Sense about Science
@senseaboutsci
The Harding Prize for trustworthy communication is awarded specifically to those who gather together information that helps someone make decisions without telling them what to do.
Nominate someone for the prize today (nominations close 28 February) : wintoncentre.github.io/harding-prize/ -
David Spiegelhalter
@d_spiegel
Just too proud of a certain Rosie Spiegelhalter (@spiegelton) who is exhibiting some of her embroidery at the top show by the Broderer's Company broderers-exhibition.co.uk/yinyang/
-
David Spiegelhalter
@d_spiegel
Plug alert! To any followers in the US, THE ART OF UNCERTAINTY will be released on March 4. Preorder now through Feb. 7 on bn.com to get 25% off. Use code PREORDER25 at checkout. #bnpreorder @BNBuzz (Must be a Rewards or Premium member)
www.barnesandnoble.com/w/the-art-of-uncertainty-david-spi… -
David Spiegelhalter
@d_spiegel
Of course this is @LucyGoBag !
David Spiegelhalter
@d_spiegel
This series on risk by Lucy Easthope seems rather good (and not just because I feature). Recommended listening www.bbc.co.uk/sounds/play/m0027l5n
-
David Spiegelhalter
@d_spiegel
This series on risk by Lucy Easthope seems rather good (and not just because I feature). Recommended listening www.bbc.co.uk/sounds/play/m0027l5n
-
David Spiegelhalter
@d_spiegel
excellent new population projection tool from ONS - put in your assumptions and watch the population pyramid stop looking like a pyramid! www.ons.gov.uk/peoplepopulationandcommunity/populationand…
-
David Spiegelhalter reposted
Royal Institution (now on Bluesky)
@Ri_Science
Tonight's livestream 💻
Life is the result of an unforeseen and unforeseeable sequence of occurrences. But is this chain of events random – or just complex?
Join @d_spiegel as he demonstrates how we can use statistics to navigate the uncertainty of life. www.rigb.org/whats-on/discourse-thinking-slowly-about-unc… -
David Spiegelhalter reposted
Office for Statistics Regulation
@StatsRegulation
Part of our series of blogs examining topics in the proposed refreshed Code of Practice for Statistics @d_spiegel from @Cambridge_Uni reflects on his experiences in the Infected Blood Inquiry and importance of transparency around statistical uncertainty 👇osr.statisticsauthority.gov.uk/blog/lessons-in-communicat…
-
David Spiegelhalter
@d_spiegel
Did a nice podcast with Abhishek from Mumbai podcasts.apple.com/in/podcast/the-point-blank-show/id2617…
-
David Spiegelhalter
@d_spiegel
I gave evidence to the Thirlwall Inquiry yesterday - transcript and witness statements here thirlwall.public-inquiry.uk/hearing/week-15-oral-evidence… Some coverage www.bbc.co.uk/news/articles/cpdxn4x5477o
-
David Spiegelhalter
@d_spiegel
Here is the link for entries for the Harding Prize wintoncentre.github.io/harding-prize/
-
David Spiegelhalter
@d_spiegel
Please nominate yourself or others for this Prize - vitally important to encourage trustworthy communication!
Sense about Science
@senseaboutsci
Nominations for the 2024 #hardingprize have opened today!
Celebrating those who tackle the communication of evidence well with two prizes - communication for lay audience + communication for expert audience
Awarded in partnership with Winton Centre @Cambridge_Uni + @SMC_London -
David Spiegelhalter
@d_spiegel
Really fed up with ITV News. I did a pre-recorded interview in which I expressed strong scepticism about the 'coffee only good in the morning' story. But they cut all that out, leaving me apparently supporting their credulous story. I really should not be so naive.
-
David Spiegelhalter reposted
Office for Statistics Regulation
@StatsRegulation
Our Director General Ed Humpherson and co author Simon Baugh head @UKgovcomms talk about growing public trust in statistics through collaborative communication and intelligent transparency in a blog released today on the @gov_analysis website👇
analysisfunction.civilservice.gov.uk/blog/growing-public-… -
David Spiegelhalter
@d_spiegel
My polemic in Nature on probability has attracted some attention www.nature.com/articles/d41586-024-04096-5 - you should be able to download from here
rdcu.be/d41qA -
David Spiegelhalter reposted
The Huddle Breakdown
@huddlebreakdown
☘️☘️☘️THE HUDDLE BREAKDOWN INTERVIEW EPISODE 2! ☘️☘️☘️
We are thrilled to welcome @d_spiegel to The Huddle Breakdown.
Join us as @Alan_Morrison67 and @jucojames have a fascinating conversation with one of the world's leading statisticians.
www.huddlebreakdown.com/p/the-huddle-breakdown-interview-… -
David Spiegelhalter reposted
Francisco J. (Paco) Jariego
@fjjariego
All of statistics and much of science depends on probability — an astonishing achievement, considering no one’s really sure what it is.
@d_spiegel @Nature
www.nature.com/articles/d41586-024-04096-5 -
David Spiegelhalter
@d_spiegel
why my spaniel is my role model www.theguardian.com/lifeandstyle/2024/dec/15/uncertainty-…
-
David Spiegelhalter reposted
Alexandra Freeman
@alex_freeman
The HoL committees hear evidence from experts all round the world on important topics and I think it’s vital that we can all hear this evidence too.
Delighted that this first move towards that is now here! Thank you to all those made it possible.
House of Lords
@UKHouseofLords
Unpacking The Evidence, a new podcast from the #houseoflords will be launching on Monday 16 December, looking at ‘Fixing Our Broken Food System.’
@HLFoodObesity chair Baroness Walmsley will take listeners on an in-depth exploration of what experts said during its recent inquiry. -
David Spiegelhalter
@d_spiegel
My contribution to Wired's series on prediction www.wired.com/story/embrace-uncertainty-forecasting-predi…
-
David Spiegelhalter
@d_spiegel
Last night I was one of the authors at Hatchards excellent book-signing evening. Predictably, I sat there by my pile of books, feeling a complete imposter, smiling at people as they walked past in search of Ben Macintyre, Michael Rosen, Robert Harris, and other real authors.
David Spiegelhalter
@d_spiegel
I was reduced to haranguing bemused passers-by like a market salesman, waving my book at them and saying my book was an ideal Christmas present for the nerd in their life. Thanks to those who talked to me and even bought a book! A great evening, so pleased to have been there.
-
David Spiegelhalter
@d_spiegel
if you are interested in the trustworthiness of statistics (and who isn't), then you may want to comment on these proposals!
Office for Statistics Regulation
@StatsRegulation
We are seeking feedback on our proposals for a new edition of the Code of Practice for Statistics via public consultation. Have your say! osr.statisticsauthority.gov.uk/consultation-on-proposed-c…
-
David Spiegelhalter reposted
Brian Clegg
@brianclegg
Interview: Probability and statistics expert Professor Sir David Spiegelhalter @d_spiegel tells us why his latest book doesn't have the P word in the title, why probability seems counter-intuitive, and why he's a dried-out husk. popsciencebooks.blogspot.com/2024/10/david-spiegelhalter-… #interview #probability
-
David Spiegelhalter reposted
The Economist
@TheEconomist
Much of probability and statistics can be counterintuitive, and the maths underlying it fearsome. But Sir David Spiegelhalter’s new book is not difficult to read or understand. For that reason alone, “The Art of Uncertainty” is a coup econ.st/4f2vWxu 👇
-
David Spiegelhalter
@d_spiegel
yet more shameless promotion of my book! But also such fun talking to Plus magazine
Plus Magazine
@plusmathsorg
In today's episode of our Maths on the Move #podcast we talk to @d_spiegel about double yolked eggs, why it's good to disagree, and why uncertainty is personal plus.maths.org/content/david-spiegelhalter-and-art-uncert… @FacultyMaths
-
David Spiegelhalter
@d_spiegel
Last week I did an interview for World at One on the fall in the rate of increase of life expectancy. I was dared to describe this as a 'negative second derivative'. So I did, and Ed Stourton appreciated it! www.bbc.co.uk/sounds/play/m0023pvj [30:42]
-
David Spiegelhalter reposted
Bobby Seagull MBE
@Bobby_Seagull
Would you want to know when you will die? ☠️ 1000 people were asked and the results were fascinating 📋
Link to read my FIRST ever book review ☺️ 📚
www.ft.com/products
Writing for @FinancialTimes about Professor @d_spiegel book "The Art of Uncertainty: How to navigate chance, ignorance, risk and luck".
#financialtimes #ft #statistics #probability #risk #finance #personalfinance #book #books #bookreview #booktok #bookstagram #bookrecommendations #money -
David Spiegelhalter
@d_spiegel
My relentless self-publicity for the new book continues with this nice interview with Tim Harford (sadly they did not use the bits when we were disagreeing!) . www.bbc.co.uk/programmes/p0jx6kvh. .
-
David Spiegelhalter reposted
Sylvie Delacroix
@SylvieDelacroix
Register for 07th Nov, 18:00: 'The Art of Uncertainty', to celebrate release of @d_spiegel's 'The Art of Uncertainty' + explore how our navigation of uncertainty is evolving in the era of data-driven technology. Panel with @Lzucca @RichardLehman1 @yulanhe
www.kcl.ac.uk/events/the-art-of-uncertainty-in-conversati… -
David Spiegelhalter reposted
Roger Highfield
@RogerHighfield
The big idea: should we be thinking about luck differently? We are all lucky, says @d_spiegel www.theguardian.com/books/2024/oct/07/the-big-idea-should…
-
David Spiegelhalter reposted
BRIGHTHINK
@BrighthinkOrg
CHANCE, CHAOS & LUCK with Dr Brian Klaas & Prof Sir David Spiegelhalter. Navigating uncertainty in a world that makes it inevitable. Mon 7th Oct, Brighton:
eventbrite.co.uk/e/chance-chaos-luck-with-brian-klaas-dav…
@d_spiegel @brianklaas @pelicanbooks @MichaelJBooks -
David Spiegelhalter
@d_spiegel
Had a wonderful interview with Chris Evans on Virgin Radio about my new book - he had done proper preparation, and asked great questions. And I could tell him that I had been a fan since Tiswas. www.youtube.com/watch?v=U2pjrVNxp4o
-
David Spiegelhalter reposted
Timandra Harkness
@TimandraHarknes
Wednesday! London! Come along
London Mathematical Society
@LondMathSoc
⚠️ Our next event in partnership w/ @WaterstonesTraf is TOMORROW, feat @d_spiegel in conversation w/ @TimandraHarknes about his new book 'The Art of Uncertainty' (@pelicanbooks)
Both authors signing books after, and LMS members get free entry!
➡️ www.waterstones.com/events/the-art-of-uncertainty-david-s… -
David Spiegelhalter
@d_spiegel
Really pleased to have been part of the team that produced this vital report.
The Royal Society
@royalsociety
We are happy to publish our new #mathematicalfutures report today, which calls for a radical reform of maths education from early years and throughout compulsory education, to enable people to thrive in a world where maths and data skills play an increasingly critical role. Read more:
royalsociety.org/news/2024/09/royal-society-calls-for-rad… -
David Spiegelhalter
@d_spiegel
more book-plugging! But this should be a fun event
-
David Spiegelhalter
@d_spiegel
Now available on www.bbc.co.uk/programmes/w3ct5xn6 - in which I rant on about the pointless obsession with the possible overall risks of low levels of alcohol (eg UK guidelines and below). If they are there at all (for which there is no good evidence), we know they are very small.
-
David Spiegelhalter reposted
Jemima Kelly
@jemimajoanna
On coincidences and the awful Bayesian sinking, by me, with thoughts from the great @d_spiegel
enterprise-sharing.ft.com/redeem/2eacf8fd-02e1-40ec-8e2d-… -
David Spiegelhalter reposted
Seb Walsh
@seb_walsh
🚨 New paper 🚨
How to maintain trustworthiness when doctors act as policy advocates
Long read in The BMJ @bmj_latest
www.bmj.com/content/386/bmj-2024-079929
1/ -
David Spiegelhalter reposted
Deborah Cohen
@deb_cohen
- “serious concerns” by independent National Screening Committee ignored
- class action lawsuit claiming the company exaggerated its effectiveness in order to increase share price
- evidence of lack of effectiveness ignored before deal done -
David Spiegelhalter
@d_spiegel
they're breeding!!!
-
David Spiegelhalter reposted
Office for Statistics Regulation
@StatsRegulation
Communicating uncertainty is a constant challenge for statisticians. This journal article by Ed Humpherson argues that we need to consider uncertainty not just from the technical perspective of producers, but also from the users’ point of view. osr.statisticsauthority.gov.uk/blog/ed-humpherson-reflect…
-
David Spiegelhalter
@d_spiegel
This should be fun
Waterstones Trafalgar Sq.
@WaterstonesTraf
Our 2nd event with the London Mathematical Society!
David Spiegelhalter in conversation with Timandra Harkness
🎟️ www.waterstones.com/events/the-art-of-uncertainty-david-s…
@d_spiegel @TimandraHarknes @LondMathSoc -
David Spiegelhalter
@d_spiegel
This is very interesting - a detailed and encouraging critique of the proposed Bayesian dynamic model for the population of England and Wales. Good to see external academic expertise being used like this
Office for Statistics Regulation
@StatsRegulation
We would like to thank Professor Arkadiusz Wiśniowski, an expert on Social Statistics and Demography from @uomsoss, for providing independent expertise and support for this project 👏(3/3) osr.statisticsauthority.gov.uk/publication/the-assessment…
-
David Spiegelhalter
@d_spiegel
Drained after 3 days in a booth recording the audio version of The Art of Uncertainty, with a sensitive mic that picked up every gut-gurgle. Note to self: if there is a next book, write it without maths, figures and tables. And short. www.amazon.co.uk/Art-Uncertainty-Living-Chance-Ignorance/…
-
David Spiegelhalter
@d_spiegel
My daughter’s real-time embroidery map of the election results
-
David Spiegelhalter reposted
Office for Statistics Regulation
@StatsRegulation
Some of our work promoting and safeguarding the use of statistics in public debate has been featured in this article 👇
BBC Politics
@BBCPolitics
It's been an election campaign featuring disputed claims - BBC Verify has fact-checked them
www.bbc.com/news/articles/cley43870l1o -
David Spiegelhalter reposted
Royal Statistical Society
@RoyalStatSoc
Passionate about stats & sharing knowledge?
📣You have until midday on Monday to apply for our next cohort of Statistical Ambassadors
From media interviews to public outreach, our ambassadors promote the importance of statistics in everyday life 🌟
rss.org.uk/news-publication/news-publications/2024/member… -
David Spiegelhalter
@d_spiegel
very nice summary of what to look out for, especially during this election campaign
Royal Statistical Society
@RoyalStatSoc
Understanding statistics is key to making sense of the world, but there's always a risk of misuse or misinterpretation
That's why we've created a guide to help you critically evaluate the figures you encounter as #ge24 approaches
Check it out here: rss.org.uk/RSS/media/File-library/Policy/2024/RSS-guide-t… -
David Spiegelhalter reposted
Paul Pharoah (R soul) 🇪🇺🇵🇬🇿🇦🇬🇧
@paulpharoah
Given the new version of PREDICT breast (breast.v3.predict.cam), I thought I'd plot usage since V1 was released in 2011. Over 3 million web sessions!!!
Many people involved over the years including
@d_spiegel @alex_freeman @ProfWishart @CarlosC24501217 -
David Spiegelhalter
@d_spiegel
Some rather complex modelling on a very important matter - brilliantly implemented by Ruth McCabe and Sarah Hayes.
Royal Society Publishing
@RSocPublishing
New from #rsos: Modelling #hepatitisc infection acquired from blood transfusions in the UK between 1970 and 1991 for the Infected Blood Inquiry. Read the full paper: royalsocietypublishing.org/doi/10.1098/rsos.231722 @d_spiegel @bloodinquiry
-
David Spiegelhalter
@d_spiegel
COI: I’m a non-exec on the UKSA Board that oversees OSR
Office for Statistics Regulation
@StatsRegulation
Today we have highlighted the importance of transparency in election claims: osr.statisticsauthority.gov.uk/news/osr-highlights-the-im…
-
David Spiegelhalter
@d_spiegel
Two great winners for the Harding Prize for Trustworthy Communication for 2023 hardingprize.online (and please ignore the deranged-looking photo below)
Sense about Science
@senseaboutsci
The Harding Prize for Trustworthy Communication recognises those who provide scientific evidence in a balanced and useful manner.
The prize is jointly awarded by Sense about Science, the Winton Centre for Risk and Effective Communication, and @SMC_London -
David Spiegelhalter
@d_spiegel
Still proud of this line I put into a much-cited paper about an even-more-cited paper; "After rather negative referees' reports had been received, the first author went into a massive sulk for a year" academic.oup.com/jrsssb/article/76/3/485/7076024
-
David Spiegelhalter
@d_spiegel
An explainer on how we obtained the numbers being quoted on the news today. It was very good working with the Inquiry, who recognised the need for careful statistical analysis.
Royal Statistical Society
@RoyalStatSoc
The Infected Blood Inquiry report tells a tragic story of families torn apart by illness, stigma and death
A number of RSS fellows helped with the statistics. Here some of them explain how they estimated the number of infections and deaths.
rss.org.uk/news-publication/news-publications/2024/genera… -
David Spiegelhalter
@d_spiegel
How to watch and stream Elvis - 2022 on Roku in the United Kingdom www.roku.com/en-gb/whats-on/movies/elvis?id=8cd469889d2d5…
-
David Spiegelhalter
@d_spiegel
This is the most wonderful news. Someone with the highest level of intelligence and integrity getting into Parliament. Just what we need.
Alexandra Freeman
@alex_freeman
I am unbelievably excited to be given this opportunity. I applied for it (and anyone can: lordsappointments.independent.gov.uk) because it’s a chance to work in public service, whilst not being at the mercy of party politics.
-
David Spiegelhalter
@d_spiegel
Calling all statisticians interested in communicating! This is a great scheme
Royal Statistical Society
@RoyalStatSoc
Passionate about stats & sharing knowledge?
Apply now to become one of our next cohort of Statistical Ambassadors
From media interviews to public outreach, our ambassadors promote the importance of statistics in everyday life 🌟
rss.org.uk/news-publication/news-publications/2024/member… -
David Spiegelhalter
@d_spiegel
Appropriate to be talking about Bayes at the Bayes Business School near Bayes' grave in Bunhill Fields
www.city.ac.uk/news-and-events/events/2024/may/book-launc… -
David Spiegelhalter
@d_spiegel
Tues 30th April: I will be talking about ‘Living with Chance, Luck and Coincidence’, at Perse School, Cambridge CB2 8QF. In aid of the excellent Sick Children’s Trust (I benefited from their charity).
Wine 19.00hr. Talk 19.30hr. £15.00 (including wine). sctcambridge@gmail.com -
David Spiegelhalter
@d_spiegel
publicstatistic.wordpress.com/2024/03/26/learning-from-a-… great post on the role of statisticians / data-scientists in scrutinising automated systems. How do they work? How good are they? How are they governed? (I think the Australian Robodebt fiasco should be better known)
-
David Spiegelhalter
@d_spiegel
Did a podcast with the wonderful Dr Roger Kneebone - including discussing performance and introversion rogerkneebone.libsyn.com/david-spiegelhalter-in-conversat…
-
David Spiegelhalter reposted
Alexandra Freeman
@alex_freeman
So pleased to see EasyRead versions of our latest decision aids for the NHS now out.
@lfinikarides and @luna9design in particular worked v closely with EasyRead users to ensure that they now have support to take part in decisions about their own health.
www.england.nhs.uk/personalisedcare/shared-decision-makin… -
David Spiegelhalter
@d_spiegel
Wonderful to see ex-colleague Ilan Goodman getting this well-deserved award!
-
David Spiegelhalter
@d_spiegel
good to see a proper model-based assessment of excess deaths blog.ons.gov.uk/2024/02/15/excess-deaths-a-new-methodolog…
-
David Spiegelhalter
@d_spiegel
Last chance to enter a trustworthy communicator into this most excellent competition ! D
David Spiegelhalter
@d_spiegel
Here's your chance to celebrate the best in communication! And each Harding Prize is worth £3,141.59 - geddit? We've had great past winners - now whose turn will it be? hardingprize.online
-
David Spiegelhalter reposted
David Spiegelhalter
@d_spiegel
Here's your chance to celebrate the best in communication! And each Harding Prize is worth £3,141.59 - geddit? We've had great past winners - now whose turn will it be? hardingprize.online
-
David Spiegelhalter
@d_spiegel
Oh dear, got a plug in the Guardian list of books to look out for in 2024 (September). I suppose this means I have to get on and finish it. www.theguardian.com/books/2023/dec/30/from-salman-rushdie…
-
David Spiegelhalter
@d_spiegel
some nice little puzzles for round the christmas table...
GCHQ
@GCHQ
The 2023 #gchqchristmaschallenge has arrived, and this year is our trickiest Challenge yet!
Have you got what it takes to complete our special festive puzzle mission?
Download it now!
gchq.gov.uk/xmas2023
🧵 1/4 -
David Spiegelhalter reposted
Sense about Science
@senseaboutsci
We have a new podcast! Join our director Tracey Brown as she reflects on the challenges of campaigning to put sound science at the heart of public life with guests @d_spiegel, @bengoldacre, @VivianT88 and Ed Humpherson of @StatsRegulation: shows.acast.com/657092f9e8850f0011a59919
-
David Spiegelhalter reposted
Royal Statistical Society
@RoyalStatSoc
Job alert: Professor of Geometry
Gresham College, London
rss.org.uk/jobs-careers/jobs-board/professor-of-geometry/ -
David Spiegelhalter
@d_spiegel
This has got some excellent discussion of communicating uncertainty about GDP estimates -see paras 11-18
Office for Statistics Regulation
@StatsRegulation
Today we have published our findings on @ONS's revisions of estimates of UK Gross Domestic Product (GDP): osr.statisticsauthority.gov.uk/news/osr-publishes-its-fin…
-
David Spiegelhalter
@d_spiegel
Here's your chance to celebrate the best in communication! And each Harding Prize is worth £3,141.59 - geddit? We've had great past winners - now whose turn will it be? hardingprize.online
-
David Spiegelhalter reposted
Alexandra Freeman
@alex_freeman
One of five new NHS decision aids released today - the result of much work by a lot of people (including @luna9design and led by @lfinikarides). We very much hope they help people with difficult health decisions to make.
Joanne Casey
@jo_casey1
@NHSEngland launches a suite of decision making aids today. Check out➡️Treatment of Varicose Veins and support your patient making the right decision for them @lfinikarides @Cherry_Linds @LegsMatter @SoTV_UK @AccelerateCIC @RoyColPod @footindiabetes
www.england.nhs.uk/wp-content/uploads/2023/11/PRN00250-ds… -
David Spiegelhalter reposted
Infinite Monkey Cage
@themonkeycage
Tune in at 7.15pm on @bbcradio4 for coincidence and luck! We were lucky (or were we?) to be joined by @d_spiegel @MarcusduSautoy @sophiedukebox
www.bbc.co.uk/programmes/m001slfs -
David Spiegelhalter
@d_spiegel
A great job for a good communicator
Gresham College
@GreshamCollege
We are recruiting for our next Professor of Geometry!
Open to those from any discipline within mathematical sciences, with a passion for communicating to diverse audiences. Find out more: www.gresham.ac.uk/professor-geometry
#recruiting #maths #mathsjobs #geometry #jobs #professorships -
David Spiegelhalter
@d_spiegel
In which I rant about trustworthiness....
ONS Focus
@ONSfocus
Facts and figures are coming at us fast every day, but ‘what’ or ‘who’ to trust?
In a NEW episode of Statistically Speaking, @d_spiegel, @StatsRegulation’s Ed Humpherson and Google data editor @smfrogers guide us across the minefields of misinformation ⬇️
traffic.libsyn.com/3470c01f-d2e6-4ea5-95b5-fb243d11309c/E… -
Post not available here, view on X
David Spiegelhalter
@d_spiegel
I am so pleased to see this! You and the whole team deserve the gratitude of the whole country
-
David Spiegelhalter
@d_spiegel
Great programme, in which I get to talk about how much sex young people had in 1973. www.bbc.co.uk/sounds/play/m001qw4s?partner=uk.co.bbc&orig…
-
David Spiegelhalter
@d_spiegel
If you have an interest in the future of statistics and data science education in schools, please read this discussion paper and send comments!
The Royal Society
@royalsociety
Our Mathematical Futures programme is considering what mathematical competencies will be needed in the future and how the education system must reform to meet those needs. A new discussion paper, 'A new approach to mathematics and data education', offers stakeholders an opportunity to engage in the discussion. Read more: royalsociety.org/topics-policy/projects/mathematical-futu…
-
David Spiegelhalter
@d_spiegel
World Cup birthday matches between squads: Chloe Kelly (Eng squad) and Alex Chidiac (Aus) both born 15th January. Jordan Nobbs (Eng) and Esther Gonzales (Spain) both born 8th December.
-
David Spiegelhalter
@d_spiegel
Each squad in 2023 Women’s World Cup has 23 players, so probability theory says we expect around half the squads to have two members with the same birthday (d/m). Neither England nor Australia have a matching pair, but out of 32 squads, 15 do! The goddess Fortuna is working well
David Spiegelhalter
@d_spiegel
Morocco and Nigeria’ squads have 3 pairs with matching birthdays, and for Nigeria, Glory Ogbunna and Christy Ocheibe were both born on Christmas Day. Data from www.fifa.com/fifaplus/en/tournaments/womens/womensworldcu…
-
David Spiegelhalter reposted
Alexandra Freeman
@alex_freeman
The new PREDICT algorithm for women with breast cancer coming very soon to the online web tool... several years in the making from lots of dedicated people (like Paul and Isabelle) and with help from patients and clinicians.
Paul Pharoah (R soul) 🇪🇺🇵🇬🇿🇦🇬🇧
@paulpharoah
Manuscript reporting work by PhD candidate Isabelle Grootes to update the PREDICT Breast prognostic model (www.breast.predict.nhs.uk) now on @medrxivpreprint
medrxiv.org/cgi/content/short/2023.07.18.23292777v1 -
David Spiegelhalter reposted
UCL Health of the Public
@UCLHealthPublic
What does the average British person think about sex? Catch up on our newest ep of #publichealthdisrupted as hosts @xandvt & @thewrittenro speak to @d_spiegel & Cath Mercer to find out more about The National Surveys of Sexual Attitudes & Lifestyles.🎧 👇 soundcloud.com/uclsound/episode-3-sex-by-numbers?in=uclso…
-
David Spiegelhalter
@d_spiegel
This is so sad. She was an extraordinary and wonderful person, and so generous. After her diagnosis, of course she starts a podcast on maths and cancer! I was so pleased to be included in an episode -I have just listened to it again, and can't believe she is no longer with us
Oxford Mathematics
@OxUniMaths
It is with huge sadness that we must announce that Vicky Neale, teacher, advocate, inspiration and much loved colleague, passed away yesterday after a long illness. Our condolences to her family, her friends and to all of you who have lost a champion for our wonderful subject.
David Spiegelhalter
@d_spiegel
This is the podcast oxford.shorthandstories.com/maths-plus-cancer/
-
David Spiegelhalter
@d_spiegel
Today we’re proud to announce the winners of the Harding Prizes for Trustworthy Communication in 2022, along with @SMC_London
and @senseaaboutsci.
The winners do the rare job of providing information without an agenda, and with uncertainties laid bare. hardingprize.online
David Spiegelhalter
@d_spiegel
Our two winners are: An interactive calculator for NHS waiting lists from
@TheIFS ifs.org.uk/calculators/what-could-happen-nhs-waiting-list…
A briefing on green steel for parliament from POST post.parliament.uk/research-briefings/post-pn-0672/
David Spiegelhalter
@d_spiegel
Congratulations also to the shortlisted entries:
- dynomight.net/stoves/
- www.vice.com/en/article/88qge4/the-new-study-on-serotonin…
- www.vice.com/en/article/88qge4/the-new-study-on-serotonin…
- x.com/jburnmurdoch/status/1562004612172873728
- hannahritchie.substack.com/p/lithium-electric-vehicles
- www.bgs.ac.uk/download/deep-sea-mining-evidence-review/
All fine examples of genuinely informing audiences about complex issues
John Burn-Murdoch
@jburnmurdoch
NEW: the collapse of emergency healthcare in England may be costing 500 lives every week, a close match for non-Covid excess deaths
Let’s look at how we reach that conclusion, by taking a deep-dive into non-Covid excess mortality and its possible causes
www.ft.com/content/f36c5daa-9c14-4a92-9136-19b26508b9d2 -
David Spiegelhalter
@d_spiegel
"Boffins want to ban the word boffins because they say it is stopping young boffins from becoming boffins". It's presumably showing my age and gender, but I have always quite liked being called a boffin - sounds more cuddly than frightening
BBC News (UK)
@BBCNews
Wednesday's Star: "Boffins: Don't call us boffins" #bbcpapers #tomorrowspaperstoday www.bbc.co.uk/news
-
David Spiegelhalter
@d_spiegel
I critiqued Andrew Bridgen's speech, trying to take it at face value without judging motivation - said reasonable to ask about benefits, harms and cost-effectiveness of vaccines. But let down by untrustworthy data and nonsense claims about excess deaths
www.bbc.co.uk/sounds/play/m001k837 -
David Spiegelhalter
@d_spiegel
Ministers have been repeatedly claiming that Labour left behind 500,000 legacy asylum cases in 2010. The correct number was .... 19,000. A substantial discrepancy. [COI: I am a non-exec for UKSA]
UK Statistics Authority
@UKStatsAuth
Our Chair considered several claims raised by @SKinnock relating to the number of asylum claims registered and not fully processed by the Home Office. (2/3) uksa.statisticsauthority.gov.uk/correspondence/response-f…
-
David Spiegelhalter
@d_spiegel
“Beware of a black beast that casts no reflection…”
Telegraph Global Health Security
@TelGlobalHealth
"The pressure did become stressful at times," said @d_spiegel. "As a boy I used to retreat to a cardboard box when I wanted some peace, so I constructed a substitute from clothes-lines and duvet-covers, a tiny ‘den’ from which even the dog was banned. Everyone should have one."
-
David Spiegelhalter
@d_spiegel
Honoured to be one of the subjects of Simon Townsley's brilliant pictures
www.telegraph.co.uk/global-health/science-and-disease/pan… -
David Spiegelhalter
@d_spiegel
So proud to have a (very) brief cameo in this work of utter genius from Bobby Fingers www.youtube.com/watch?v=QHi479xN_po
-
David Spiegelhalter reposted
Stian Westlake
@stianwestlake
The world would be a better place if governments had a more experimental mindset.
But there are barriers to overcome - not least, public buy-in. By me in the Observer
www.theguardian.com/books/2023/mar/06/the-big-idea-should… -
David Spiegelhalter
@d_spiegel
I'm lucky enough to have seen this already, and it's brilliant. Far too good to just let 9-13 year-olds read it
Tim Harford
@TimHarford
Oo, first glimpses inside The Truth Detective, which is out next week! Ollie Mann has done a wonderful job bringing my ideas to life.
For those asking, the book is aimed at 9-13 year olds, although I keep having to confiscate my copy from the grown-ups... -
David Spiegelhalter
@d_spiegel
Delighted to see this. Shows it can be worth complaining (as I did about this programme)
Ofcom
@Ofcom
Ofcom has found the Mark Steyn programme, which aired on GB News on 21 April 2022, in breach of our broadcasting rules.
Full statement linked below:
www.ofcom.org.uk/news-centre/2023/ofcom-finds-the-mark-st… -
David Spiegelhalter reposted
Michael Hubank
@generoom
Re-reading this excellent and entertaining article I realised it is a masterclass in British understatement. As an aide to understanding and to avoid misinterpretion by our non-British friends, I will translate some of the salient points below:
Anneke Lucassen
@annekeluc
Realistic expectations are key to realising the benefits of polygenic scores | The BMJ www.bmj.com/content/380/bmj-2022-073149
-
David Spiegelhalter
@d_spiegel
HM Treasury on the naughty step "We consider that the way the tweet presents this data to be misleading"
Office for Statistics Regulation
@StatsRegulation
We have written to @hmtreasury about a graph giving a misleading impression of the scale of deceleration in inflation.
osr.statisticsauthority.gov.uk/correspondence/ed-humphers…
It is important that data visualisations follow Government guidance to avoid undermining public confidence in the outputs of departments. -
David Spiegelhalter
@d_spiegel
this is a very useful review (certainly for me, lots here I didn't know)
Office for Statistics Regulation
@StatsRegulation
"Statistical Literacy – it’s all in the communication” - our new review published today explores what statistical literacy is and what it means for how statistics are communicated to the public osr.statisticsauthority.gov.uk/publication/statistical-li…
-
David Spiegelhalter reposted
Robert Cuffe
@robertcuffe
Big part of counting “excess” deaths is working out what you mean by “expected”.
@ONS - and others - looking for feedback on how best to define that in the years after the pandemic.
Sarah Caul
@Sarah__Caul
Today we released a new blog on what we are doing around excess mortality: blog.ons.gov.uk/2023/02/10/how-do-we-measure-expected-and… (1/3)
-
David Spiegelhalter
@d_spiegel
Warning: Shameless self-promotion. The Art of Statistics is the Kindle Monthly Deal for february - only £1.99!
www.amazon.co.uk/Learning-Data-Statistics-Pelican-Books/d… -
David Spiegelhalter
@d_spiegel
Heat is very dangerous. ONS has also published deaths by day of occurrence, which shows a big spike during the record hottest days of July 19th and 20th - there were over 1,100 excess deaths in England and Wales on those two days alone
www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa… -
David Spiegelhalter
@d_spiegel
Latest data from ONS: since start of the pandemic, over 160,000 deaths registered as caused by Covid in E+W. And 52 registered as caused by a Covid vaccine (one aged under 20)
www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa… -
David Spiegelhalter
@d_spiegel
Pleased this story included my comments that 2,000 extra heart attacks, over lifetimes, is surprisingly reassuring, given normal 100,000 admissions a year. So looks like this disruption in CV prevention is a negligible contribution to current excess deaths www.theguardian.com/society/2023/jan/19/thousands-at-risk…
-
David Spiegelhalter
@d_spiegel
Great to see the ‘RESIST 2 Counter-disinformation toolkit’ gcs.civilservice.gov.uk/publications/resist-2-counter-dis… recommends our PROVE framework for trustworthy communication (p 40) www.nature.com/articles/d41586-020-03189-1
David Spiegelhalter
@d_spiegel
And now we know that this approach, informing rather than persuading, is seen as more trustworthy by sceptics. This means that one-sided messages actively *reduce* trust in the very audience that you are trying to reach academic.oup.com/pnasnexus/article/1/5/pgac280/6881732
-
David Spiegelhalter reposted
Gemma Chandratillake
@GemmaChand
@acethecurious presenting @d_spiegel Winton Centre work done in collaboration with @East_Genomics and @Cambridge_ICE #genomicmedicine How to make genetic test reports actionable for clinicians and patients (3 publications to date from this work!) @NHSgms @genomicsedu #genechat
-
David Spiegelhalter
@d_spiegel
In 2021 I was in a fine team that concluded that “COVID-19 largely multiplies existing risks faced by patients” www.thelancet.com/journals/lanepe/article/PIIS2666-7762(2… Plausible that Covid survivors continue to have increased risks, which could contribute to current excess deaths, as well as acute care delays
-
David Spiegelhalter
@d_spiegel
This is tragic. Some of the best times in my life have been when wild-camping on Dartmoor
Dartmoor National Park
@dartmoornpa
We're disappointed with the result of the wild camping legal challenge. We'll consider our position before deciding if to appeal and also discuss with landowners.
We maintain wild camping is key part of open-air recreation, a way to enjoy Dartmoor.
www.dartmoor.gov.uk/enjoy-dartmoor/outdoor-activities/cam… -
David Spiegelhalter
@d_spiegel
this is an excellent guide on communicating health risks by my colleague @mc_climent - and available in Spanish too!!
Alexandra Freeman
@alex_freeman
How to communicate health risks: a nice guide by the incomparable @mc_climent for @sciencemedia_es
sciencemediacentre.es/en/how-communicate-risks-reported-s…
@SMC_London @d_spiegel -
David Spiegelhalter reposted
Timothy Gowers @wtgowers
@wtgowers
Simon Jenkins dusts off his anti-maths article and publishes it for about the tenth time. One of those times was six years ago and I was moved to reply. Here's the latest version of the article, and linked in the next tweet is my reply to the previous one.
www.theguardian.com/commentisfree/2023/jan/05/maths-schoo… -
David Spiegelhalter
@d_spiegel
Of course, this covers a very cold spell in 1st half of December.
David Spiegelhalter
@d_spiegel
Large jump to 2,064 excess non-Covid deaths registered in week Dec 17-23, up from 796 in previous week. Big increases in both deaths at home and in hospital. Apparently not due to delayed registrations
Office for National Statistics (ONS)
@ONS
14,530 deaths were registered in England and Wales in the week ending 23 December 2022 (Week 51).
This was 20.7% above the five-year average (2,493 excess deaths).
➡️ www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa… -
David Spiegelhalter
@d_spiegel
Large jump to 2,064 excess non-Covid deaths registered in week Dec 17-23, up from 796 in previous week. Big increases in both deaths at home and in hospital. Apparently not due to delayed registrations
Office for National Statistics (ONS)
@ONS
14,530 deaths were registered in England and Wales in the week ending 23 December 2022 (Week 51).
This was 20.7% above the five-year average (2,493 excess deaths).
➡️ www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa… -
David Spiegelhalter reposted
John Kerr
@SciComGuy
Scientists! Do not fear being balanced about risks and benefits and honest about uncertainties when communicating evidence. This doesn’t undermine trustworthiness (and may boost it for some audiences). -- Our new paper in @PNASNexus. (1/8) academic.oup.com/pnasnexus/advance-article/doi/10.1093/pn…
-
David Spiegelhalter
@d_spiegel
On this day in 1917 my grandfather Cecil Spiegelhalter was commissioned, from a Sergeant, to 2nd Lt 18th Battalion Lancashire Fusiliers, aged 35. A very dangerous rank: expected to first up the ladder to lead his men into attack. But he survived the war, by great good luck 1/n
Post not available here, view on X
David Spiegelhalter
@d_spiegel
So Cecil survived the war, risking his life but missing the worst actions through illness and, ironically, getting blown up at the right time. He married Gladys, and lived till 1961 when I knew him as an old man (with a WW1 pension for gas injury) 7/n
David Spiegelhalter
@d_spiegel
Each of us exists because of a string of unforeseeable events. He may not have thought himself lucky, but Cecil’s well-documented experiences show the repeated role of arbitrary fortune. I really should not be here. More details at spiegelbull.com?page_id=1310 8/n
-
David Spiegelhalter
@d_spiegel
This is such great news - so well deserved. And Pouria is of course right that the whole team has been amazing .
Pouria Hadjibagheri MBE
@Pouriaaa
I am truly humbled to have been awarded an MBE in the King's New Year honours.
This belongs to our entire team, without whose dedication none of it would have happened.
I am completely speechless by @janemerrick23's amazing article. 💐
inews.co.uk/news/politics/new-year-honours-uk-covid-hero-… -
David Spiegelhalter
@d_spiegel
Top Christmas present to do upside-down!
-
David Spiegelhalter
@d_spiegel
Personal tweet: I am so proud to be connected to this fine charity, who support such extraordinary work in communities in India and Nepal. Really worth supporting this Christmas!
Karuna
@karuna_trust_uk
'Our son is improving only because of this project. We are so poor we wouldn't have been able to afford their therapies otherwise.'
Watch Ornab's story here...
www.youtube.com/watch?v=YUQip5JWGqg -
David Spiegelhalter reposted
Sense about Science
@senseaboutsci
Have you seen clear and transparent evidence which is presented to inform but not persuade?
Submit an entry for the Harding Prize to help us find the best communications of evidence of 2022: hardingprize.online
@SMC_London @d_spiegel @alex_freeman @CochraneUK -
David Spiegelhalter
@d_spiegel
This wonderful story from my family makes me too proud for words - so inspiring
metro.co.uk/2022/12/11/i-hiked-the-three-peaks-with-my-fa… -
David Spiegelhalter
@d_spiegel
@IsabelOakeshott says Covid was “a disease with a low mortality rate among all but the very elderly”. Around 1% of men my age (66) who caught it in 1st wave died. This is about the risk of coronary bypass graft surgery, 1,000 parachute jumps, or 70,000 miles on a motorbike.
David Spiegelhalter
@d_spiegel
Either this is ‘low’ risk, I am ‘very elderly’, or she is wrong. Refs: 1% mortality:
www.nature.com/articles/s43856-022-00106-7 CABG survival of 99%
www.nicor.org.uk/adult-cardiac-surgery-surgery-audit/ -
David Spiegelhalter reposted
David Spiegelhalter
@d_spiegel
Entries now open for the Harding Prize for Trustworthy Communication hardingprize.online Have you seen something this year that gave you the information you needed to make a decision? Something you could trust, and which was trying to help rather than persuade you?
-
David Spiegelhalter
@d_spiegel
New report from CMOs etc on Covid www.gov.uk/government/publications/technical-report-on-th… Chapter 11 'Communications' says sensible things. "The media medics, and most specialist health and science correspondents, provided challenge, were well informed and generally relayed accurate technical messages clearly"
Post not available here, view on X
David Spiegelhalter
@d_spiegel
"The work of dedicated journalists and organisations like the Science Media Centre .. helped interpret research, put it in context, explain limitations and in doing so support informed debate while being clear where scientific consensus lay"
David Spiegelhalter
@d_spiegel
"Sometimes independent scientists had strong views on policy choices. Informed debate is important, but the blurring of science advice and policy opinion could cause confusion"
-
David Spiegelhalter
@d_spiegel
Something to pass the time before the football…
Cristina Peñasco
@chrispenasco
Last call ⏰! If you have not registered yet… do not miss the chance of listening to our fantastic guest @d_spiegel! Is communication a key puece in policy making?
-
David Spiegelhalter
@d_spiegel
Entries now open for the Harding Prize for Trustworthy Communication hardingprize.online Have you seen something this year that gave you the information you needed to make a decision? Something you could trust, and which was trying to help rather than persuade you?
David Spiegelhalter
@d_spiegel
Separate prizes, each worth £3,141.59, for communicating evidence to a general or specialist audience. Judging panel includes Helen Boaden, Helen Jamieson, @FraserNelson and Professor Onora O’Neill.
Closing date: 31st December 2022. @SMC_London @senseaboutsci
David Spiegelhalter
@d_spiegel
Harding Prize for 2021 was jointly won by ONS Covid Infection Survey and the Cochrane Review of Hydroxychloroquine for Covid-19. Helen Boaden said “Both our winners set the gold standard for clearly communicating accurate, trustworthy, transparent data without frills or spin.”
-
David Spiegelhalter
@d_spiegel
Finally finished the Coronation of George VI, a family and friends project. A Victory puzzle from the 1930s, around 700 pieces. Great whimsies, lots of challenging grey, and it’s complete! Now how to waste the rest of the winter?
-
David Spiegelhalter
@d_spiegel
I am honoured to have taken part in this excellent series by Vicky Neale
University of Oxford
@UniofOxford
🎧 | In the latest episode of our #mathspluscancer podcast, Dr @VickyMaths1729 talks to Prof Sir @d_spiegel about the role of statisticians in communicating risk to the public, and the human side of cancer.
-
David Spiegelhalter
@d_spiegel
Should have added: we start at 10:30
David Spiegelhalter
@d_spiegel
Tomorrow I will be giving evidence to the @bloodinquiry as part of the Statistics Expert Group - live on Youtube www.youtube.com/channel/UCmFkDoDeSsnYVZtNgo3150g. Our report is here: www.infectedbloodinquiry.org.uk/news/inquiry-publishes-re…
-
David Spiegelhalter
@d_spiegel
Tomorrow I will be giving evidence to the @bloodinquiry as part of the Statistics Expert Group - live on Youtube www.youtube.com/channel/UCmFkDoDeSsnYVZtNgo3150g. Our report is here: www.infectedbloodinquiry.org.uk/news/inquiry-publishes-re…
-
David Spiegelhalter
@d_spiegel
I do think Fran could have airbrushed out the bags under the eyes - but not really her (excellent) style
Keble College, Oxford
@KebleOxford
"At Keble I developed an enthusiasm for the ideas behind probability and statistics, honed from hours of discussion in the Lamb and Flag." David Spiegelhalter (Maths 1971) portraits.keble.net/2022/david-spiegelhalter 📷@FranMonks #portraitfriday
-
David Spiegelhalter
@d_spiegel
I've just read The Irrational Diary of Clara Valentine from Coralie Colmez. www.coraliecolmez.com/books I'm obviously not the intended audience (hardly being a YA), but I enjoyed the unusual mixture of mystery, maths, and ..... other stuff. A positive and fun read.
-
David Spiegelhalter
@d_spiegel
Delighted to be part of this event to celebrate Stephen. I will be talking about statisticians and public inquiries , featuring Bristol, Shipman, and Infected Blood
Data & Statistical Science for Health, LSHTM
@LSHTM_datastats
Join this symposium in-person or via livestream to celebrate the career of Professor Stephen Evans.
📆 19 October
🕧 17:00 BST
📍 LSHTM or livestream
🗣️ @statsfizz, @d_spiegel, Ian Douglas, Sebastian Schneeweiss
🔗www.lshtm.ac.uk/newsevents/events/symposium-celebrate-car… -
David Spiegelhalter
@d_spiegel
Thanks to my colleague @mc_climent for translating my discussion of coincidental Mexican earthquakes and the birthday problem www.bbc.co.uk/sounds/play/m001cq31
Maricarmen Climént
@mc_climent
🇲🇽Les comparto una explicación calrísima sobre la #probabilidad de tres #sismos con fecha del #19septiembre. La probabilidad es alrededor de 1 en 25: muchísimo más alta de lo que varias personas especularon.”Lo explicó hoy @d_spiegel, estadístico de @CamUni en @BBCMoreOrLess.🧵
-
David Spiegelhalter
@d_spiegel
I've finally done a blog on my experience of #thequeue. Worth every minute of 14.5 hours spiegelbull.com?page_id=1184
-
David Spiegelhalter reposted
Data & Statistical Science for Health, LSHTM
@LSHTM_datastats
Join this symposium in-person or via livestream to celebrate the career of Professor Stephen Evans.
📆 19 October
🕧 17:00 BST
📍 LSHTM or livestream
🗣️ @statsfizz, @d_spiegel, Ian Douglas, Sebastian Schneeweiss
🔗www.lshtm.ac.uk/newsevents/events/symposium-celebrate-car… -
David Spiegelhalter
@d_spiegel
3 Mexico earthquakes on same date since 1985, but tiny estimates of the chance are wrong
english.elpais.com/international/2022-09-20/whats-the-pro… Apparently 22 in Mexico over magnitude 7 in last 50 years 1/2 en.wikipedia.org/wiki/List_of_earthquakes_in_Mexico
David Spiegelhalter
@d_spiegel
If they occur at random, this is a ‘birthday problem’. Wolfram calculator says 1 in 90 that 3 of them will happen on same day. So surprising, but not amazing 2/2 www.wolframalpha.com/input?i=birthday+problem+calculator&…
-
David Spiegelhalter
@d_spiegel
OSR on The Metric Question. "We would be surprised to see the evidence drawn from this specific question being used in any kind of formal quantitative way"
Office for Statistics Regulation
@StatsRegulation
In our latest blog, Director General Ed Humpherson responds to concerns raised with OSR regarding the recent consultation by @beisgovuk on the Choice on units of measurements: marking and sales:
osr.statisticsauthority.gov.uk/weights-and-measures-how-c… -
David Spiegelhalter
@d_spiegel
Kind plug by @stephenkb - so good to see ‘algorithms’ being clearly discussed, instead of mystifying with the label ‘AI’
Stephen Bush
@stephenkb
The beauty of algorithms in public policy is, used correctly, they can allow you to make better decisions once you've made those political decisions. But need to be transparent. @d_spiegel has some good thoughts about how which are worth reading in full: hdsr.mitpress.mit.edu/pub/56lnenzj/release/3
-
David Spiegelhalter
@d_spiegel
Rather than a queue, The Queue is perhaps better thought of as a pilgrimage, in which the journey is as important as the goal. Yesterday reminded me of walking the Camino to Santiago – dogged continuity made hugely enjoyable by meeting wonderful people who look out for each other
David Spiegelhalter
@d_spiegel
My gang included a 7-year-old, uncomplainingly walking for 14½ hours before finally getting to do her practiced curtsey – in Brownie uniform – at 2am. And a 75-year-old who made friends with everyone, and got there by walking ahead, finding a bench and waiting for us to catch up
David Spiegelhalter
@d_spiegel
So pleased I did it. But grumpy that DCMS gave untrue information that entry to queue was stopped for 6 hours, which I fortunately ignored. Also journalists repeat the tracker’s claim that the route is 4.9 miles long, but this ignores the zig-zags; we walked around double that!
-
David Spiegelhalter
@d_spiegel
Isn't there an option missing? Truly awful survey question from Office for Product Safety and Standards at BEIS. But it does provide great material for teaching about manipulative phrasing www.gov.uk/government/consultations/choice-on-units-of-me…
-
David Spiegelhalter
@d_spiegel
brilliant and convincing thread from @jburnmurdoch - excellent forensic statistical work
John Burn-Murdoch
@jburnmurdoch
NEW: the collapse of emergency healthcare in England may be costing 500 lives every week, a close match for non-Covid excess deaths
Let’s look at how we reach that conclusion, by taking a deep-dive into non-Covid excess mortality and its possible causes
www.ft.com/content/f36c5daa-9c14-4a92-9136-19b26508b9d2 -
David Spiegelhalter reposted
Koenfucius 🔍
@koenfucius
Just had a look at the NHS decision support tool for patients with Wet Age-related Macular Degeneration—one of several. Really well made, clear language, and in particular the representation of the benefits and risks is excellent. www.england.nhs.uk/shared-decision-making/decision-suppor… HT @d_spiegel
-
David Spiegelhalter
@d_spiegel
I am so proud of the Winton Centre team for producing these amazing decision aids - just click and have a look at the great design and balanced information. And tell people to use them!
Alexandra Freeman
@alex_freeman
Please try these decision aids, to help patients and doctors make treatment decisions together
www.england.nhs.uk/shared-decision-making/decision-suppor…
and give feedback.
Alongside are our free e-learning courses:
portal.e-lfh.org.uk/Component/Details/543925
for GPs
portal.e-lfh.org.uk/Component/Details/604821
for ophthalmologists
... -
David Spiegelhalter
@d_spiegel
Instant movement (in both directions) between relative measures (risk, odds and hazard ratios) and changes in absolute risks. Essential to most risk stories! With added explanations.
Alexandra Freeman
@alex_freeman
Proudly presenting RealRisk Light, courtesy of @d_spiegel and @grumplet.
RealRisk (realrisk.wintoncentre.uk) is our online tool to help communicate the results reported in academic papers in a meaningful way. RealRisk Light is our new add-on with more maths and fewer words! -
David Spiegelhalter reposted
Paul Mainwood
@PaulMainwood
The time series of ONS's age-standardised mortality rates for deaths involving COVID has a lot in it.
Plotting on a linear scale has drawbacks - you see the mortality in early 2021 ... but then the scale needed squashes everything since then; you can't see what's going on. (1/x) -
David Spiegelhalter
@d_spiegel
Why there should have been more high-quality evaluations of public health measures
Alasdair Munro
@apsmunro
A recent article in BMJ EBM claimed we should accept mechanistic studies as a higher form of evidence for “complex systems”
This is wrong
Failing to generate better quality evidence on NPIs was undoubtedly one of the biggest mistakes of the pandemic
alasdairmunro.substack.com/p/the-pandemic-evidence-failur… -
David Spiegelhalter
@d_spiegel
Just like to thank everyone for such kind and supportive comments - I’m afraid I won’t be replying to them all!
David Spiegelhalter
@d_spiegel
not my usual topic.... www.theguardian.com/lifeandstyle/2022/jul/23/meeting-same…
-
David Spiegelhalter
@d_spiegel
not my usual topic.... www.theguardian.com/lifeandstyle/2022/jul/23/meeting-same…
-
David Spiegelhalter
@d_spiegel
He’s not the only one who gets annoyed at this….
Tom Chivers
@TomChivers
In which I get annoyed at the "There's no such thing as a safe level of drinking for the under-40s" paper that was doing the rounds the other day inews.co.uk/news/dont-worry-there-is-such-a-thing-as-safe…
-
David Spiegelhalter reposted
Pouria Hadjibagheri MBE
@Pouriaaa
I present to you, (most of) the team behind the UK Coronavirus Dashboard.
We worked together day, night, and midnight for over two years. We built something amazing, unique, and yeah, something cool!
We didn't set to be the best in world, nor did we do it to win awards... 1/3 -
David Spiegelhalter
@d_spiegel
Another day, another awful IHME study published in the Lancet www.theguardian.com/society/2022/jul/14/alcohol-is-never-… Remember all the confident 'all alcohol harmful' headlines in 2018? www.bbc.co.uk/news/health-45283401 Well now they have completely changed their minds. 1/3
David Spiegelhalter
@d_spiegel
They throw all studies into an impenetrable statistical model and try to ‘adjust’ for their poor quality - conclude J-shaped risk curve, with very little dip for young people, but for people my age the first 2 drinks each day are medicine. May be reasonable... 2/3
David Spiegelhalter
@d_spiegel
But they call any risk above non-drinkers ‘harmful’, and so say young people should not drink. By this logic, they should not drive, or even get out of bed. No absolute risks provided, which breaks Lancet guidelines and provides no basis for policy recommendations. Awful. 3/3
-
David Spiegelhalter
@d_spiegel
It was wonderful to work with Marc on the cluster of failures paper, and it changed my career
Kate Brown
@katebrown220
As a junior doctor I worked with Professor Marc de Leval (1941-2022): what to say about this amazing, intelligent man? 1/5, though there could be more:
He wasn’t afraid to explore failures for learning
Analysis of a cluster of surgical failures: JTCVS. 1994; 107: 914-924 -
David Spiegelhalter reposted
Paul Mainwood
@PaulMainwood
Through 2021 there was a choice of denominators to use for vaccine efficacy, take-ups and so on.
Awkwardly, ONS's mid-2020 denominators made most sense for all age-groups except the very oldest, where NIMS worked better. But that felt ad hoc.
Census 2021 - out today...
(1/6) -
David Spiegelhalter
@d_spiegel
sorry, poorly worded! … should be taught in schools, but not as part of traditional maths.
David Spiegelhalter
@d_spiegel
The President of the Royal Society, Prof Adrian Smith, argues that ‘data literacy’ should not be taught in schools as part of traditional maths royalsociety.org/blog/2022/06/envision-adrian-smith/
-
David Spiegelhalter
@d_spiegel
The President of the Royal Society, Prof Adrian Smith, argues that ‘data literacy’ should not be taught in schools as part of traditional maths royalsociety.org/blog/2022/06/envision-adrian-smith/
-
David Spiegelhalter reposted
Dr Susan Oliver (PhD)
@DrSusanOliver1
Antivaxxers have seized on a preprint by Peter Doshi to falsely claim that vaccines cause more harm than good. In fact, the preprint just uses P-Hacking and apples to oranges comparisons to fool the unaware. Cindy and I take you through their tactics:
youtu.be/drSAsfuMkuw -
David Spiegelhalter
@d_spiegel
Covid infection risk has been on average 2 micromorts for a healthy child / young person. That is very low en.wikipedia.org/wiki/Micromort
Alasdair Munro
@apsmunro
A new study using high quality data from the UK is helping us better understand the risk of death from #covid19 in children
The good news: not only is it low, it keeps getting lower
For a healthy child ~ 2 per million infections
alasdairmunro.substack.com/p/children-are-really-really-u… -
David Spiegelhalter
@d_spiegel
Following @FryRsquared excellent Making Sense of Cancer www.bbc.co.uk/iplayer/episode/m0017wzq/horizon-2022-makin… , note the sort of stats she was using are in Predict for breast and prostate cancer breast.predict.nhs.uk
David Spiegelhalter
@d_spiegel
Images show additional 10-year survivors due to chemotherapy for a specific type of woman with breast cancer, and rates of side-effects from alternative treatments for prostate cancer
-
David Spiegelhalter
@d_spiegel
How to communicate uncertainty? This is a great report on Winton Centre's John Kerr's work with POST
POST (UK Parliament)
@POST_UK
Uncertainty is a fundamental part of all research and requires those who produce research to communicate it to others. Want to know more about how we handle uncertainty in POST's research? Our latest article explains: post.parliament.uk/communicating-uncertainty-in-research-…
-
David Spiegelhalter reposted
Alasdair Munro
@apsmunro
Thank you @StuartJRitchie for the best piece I’ve read in a while
Science should be as apolitical as possible, and as scientists it behooves us to strive towards this
Some groups have abandoned this principle during the pandemic - and it shows
stuartritchie.substack.com/p/science-is-political -
David Spiegelhalter
@d_spiegel
I so enjoyed doing this podcast for Bluesci - great questions
Bluesci Podcast
@BlueSciPod
📣 NEW PODCAST EPISODE! 📣
This week, Mark and Georgia spoke to @d_spiegel about his work in statistics, Loop and Winter Wipeout?! 🎿
Listen now on Anchor: anchor.fm/bluesci-podcast/episodes/Society-and-statistics… -
David Spiegelhalter reposted
Office for National Statistics (ONS)
@ONS
How does #covid19 compare with flu as a cause of death?
While there are limitations in comparing mortality from both diseases, our latest article explores trends in mortality over the pandemic and how this compares with pre #coronavirus years ➡️ www.ons.gov.uk/peoplepopulationandcommunity/healthandsoci… -
David Spiegelhalter reposted
Oxford Mathematics
@OxUniMaths
The numbers don't speak for themselves. And it's not about mathematics.
In his lecture last week @d_spiegel gave his recipe for the successful communication of complex statistics. Watch online from 5pm BST this Thursday (19th) & any time afterwards.
www.youtube.com/watch?v=fKOmIHkq_i8 -
David Spiegelhalter
@d_spiegel
this is brilliant work from Winton Centre colleagues, empirical testing of formats for genetic reports. Hope the new versions can be widely implemented
Gabriel Recchia
@mesotronium
1/7 It’s important people understand results they get from medical testing– like genetic testing. We’ve been using user-centred design (Lauren Capacchione, @alex_freeman, and @acethecurious) to build better genetic reports than those now in use by the NHS: www.gimjournal.org/article/S1098-3600(22)00745-6/fulltext
-
David Spiegelhalter reposted
John Kerr
@SciComGuy
Thrilled to be presenting (virtually) in Berlin at the @Official_EMWA symposium ‘Communicating with the Public – what has the COVID-19 pandemic taught us?’ Talking about misinformation and mix-ups in the media. A few links/thoughts related to my talk… 1/9
-
David Spiegelhalter reposted
Maria Skoularidou (she/her)
@skoularidou
The keyword "trustworthy" is a direct link to @d_spiegel and Baroness O'Neill of Bengarve.
I'm very curious to see the programme and the accepted manuscripts of this conference -
David Spiegelhalter
@d_spiegel
My grandmother got the Royal Red Cross Medal in WW1 for her work in France under heavy bombing. Then she got married (rather sombre photo). I’ve just found this abrupt letter from the War Office telling her she now had to resign as a nurse. Not even any thanks.
-
David Spiegelhalter reposted
Stian Westlake
@stianwestlake
Our next Covid-19 Evidence session: did the UK have the right data and statistical resources in place to tackle the pandemic?
The event will be both in person or online - sign up now!
Royal Statistical Society
@RoyalStatSoc
📢Our next Covid evidence session is taking place next Tues on statistical resources
Hear from @adamjkucharski, @rozeggo, @RossjournoClark, @peterkwells, Ed Humpherson of @StatsRegulation, @ScottHeald72 & others
rss.org.uk/training-events/events/events-2022/rss-events/… online
rss.org.uk/training-events/events/events-2022/rss-events/… in-person -
David Spiegelhalter reposted
Office for Statistics Regulation
@StatsRegulation
It's incredibly powerful to see 96% of responses agreed it is important for there to be a body like @UKStatsAuth to speak out against the misuse of statistics.
This research offers valuable insights into public attitudes towards statistics and helps us further our own work 👇
UK Statistics Authority
@UKStatsAuth
We welcome this report from @NatCen and are pleased to see that public confidence in official statistics remains high. The National Statistician and Director General for Regulation @StatsRegulation have responded here: uksa.statisticsauthority.gov.uk/news/public-confidence-in…
-
David Spiegelhalter
@d_spiegel
In which my opinions about epidemiologists giving public health advice get reported... gregmantell.substack.com/p/leading-academics-stop-epidemi…
-
David Spiegelhalter
@d_spiegel
so looking forward to a live gig!
Oxford Mathematics
@OxUniMaths
In-person @OxUniMaths Public Lectures are back & we start in style on 12 May with @d_spiegel reflecting on the battle to convey complex statistics during the pandemic.
Register: www.maths.ox.ac.uk/node/60099
The lecture will also be premiered on our YouTube channel one week later. -
David Spiegelhalter
@d_spiegel
this really is extraordinary
Full Fact
@FullFact
As the Commons debate whether the Prime Minister should be investigated for misleading Parliament, a reminder that Boris Johnson has got the same employment claim wrong in the House nine times now.
fullfact.org/economy/boris-johnson-makes-false-employment… -
David Spiegelhalter
@d_spiegel
delighted to be part of this
International Science Council
@ISC
Hosted in collaboration with @IntStat, hear from leading statisticians on the relationship between #covid19 and the field of statistics 📊
@d_spiegel | @ellenpetersjdm | @KerrieMengersen | @craigjcalhoun
Registration ⬇️
council.science/events/pandemic-statistics/ -
David Spiegelhalter
@d_spiegel
Recently found this photo in my grandfather’s papers, labelled ‘Shanghai 1933’. I knew he was working in the hedonistic International Settlement, but was still rather taken aback. And then I realised that the girl in the bottom-left is my mother. Aged 9. One for the archives.
-
David Spiegelhalter reposted
Florence Nightingale Museum 💙
@florencemuseum
A hugely enjoyable day filming for our new ‘Nightingale & Statistics’ display with @d_spiegel & @StatsJen. Thanks for your support @TeamPhastar Good work Team Flo & @unusualexpo
-
David Spiegelhalter
@d_spiegel
great article by Fiona Fox www.theguardian.com/science/2022/apr/03/scientists-must-b… (and features our Harding Prize for Useful and Trustworthy Communication wintoncentre.maths.cam.ac.uk/news/winners-2021-harding-pr…
-
David Spiegelhalter reposted
Prof. Jennifer Visser-Rogers
@StatsJen
I’ve organised an event for the @RoyalStatSoc with @d_spiegel and @TomChivers where we’ll be discussing the communication of data and statistics through the COVID-19 pandemic ahead of the official inquiries…
Royal Statistical Society
@RoyalStatSoc
📢You can now register to attend the first of our Covid-19 evidence sessions - to discuss the important statistical issues of the pandemic ahead of the official inquires
This session is themed around communication and is organised by @StatsJen, @d_spiegel & @TomChivers (1/3) -
David Spiegelhalter
@d_spiegel
and here’s our paper on the "problem of simultaneously monitoring over 200000 indicators for excess mortality”
rss.onlinelibrary.wiley.com/doi/10.1111/j.1467-985X.2011.… -
David Spiegelhalter reposted
PACAC Committee
@CommonsPACAC
We have endorsed the appointment of Sir Robert Chote as Chair of the UK Statistics Authority. Read our report: publications.parliament.uk/pa/cm5802/cmselect/cmpubadm/11…
-
David Spiegelhalter reposted
Christopher Burr
@chrisdburr
Applications close *next week* for our course on public engagement of data science and AI 📊 at the @turinginst
In addition to a guest lecture from Professor @d_spiegel, participants will have their ideas brought to life by the excellent illustrator @lighthands
Apply below👇
Christopher Burr
@chrisdburr
Applications are now open to graduate students and early-career researchers on the second @turinginst Turing Commons course, 'Public Engagement with Data Science and AI'.
Limited spaces are available. Information and application link are here: www.eventsforce.net/turingevents/frontend/reg/thome.csp?p… -
David Spiegelhalter
@d_spiegel
this should be very good
John Burn-Murdoch
@jburnmurdoch
Extremely excited to be on a panel with the brilliant @_HannahRitchie @BristOliver @puzzlesthewill next Tuesday at @RoyalStatSoc, talking about how data & graphics have been used (and misused...) to communicate all things Covid.
Sign up here for free! rss.org.uk/training-events/events/events-2022/rss-events/… -
David Spiegelhalter
@d_spiegel
this is very interesting. I admit to having over-stated the deficit in young deaths in 2020, through insufficient attention to reporting delays
-
David Spiegelhalter
@d_spiegel
Can’t believe I’m still seeing this. @mlevitt_NP2013 made a silly error in March 2020, but for 2 years he has refused to admit it. Simple question: if the maximum IFR from Imperial is 9.3% (in over-80s), how could they possibly expect 116/700 = 17% deaths on Diamond Princess?
Michael Levitt
@MLevitt_NP2013
This is what I remember happening. Full of cited sources.
The Medium @d_spiegel medium.com/search used Report 9 to bel a year of extra risk of death
This disturbed & energized me as 10x too high. medium.com/@michael.levitt/the-excess-burden-of-death-fro…
My thanks to @HRenkel for keeping history straight!
Heather
@HRenkel
Report#9 from Imperial College of London & Neil Feguson fabricated Covid severity & the viability of lockdowns. Ventilator need created by a friend who stood to profit; a retracted paper, omitted data, exponential growth, invented a crisis & a solution heatherrenkel.github.io/blog/report_9.html
-
David Spiegelhalter
@d_spiegel
April 5th: You can also apply to give a 2-minute presentation on what has gone right, and wrong, in stats comms during the pandemic
rss.org.uk/news-publication/news-publications/2022/genera…
Royal Statistical Society
@RoyalStatSoc
Register for the first of our Covid Evidence sessions which will focus on stats communication during the pandemic
With @jburnmurdoch, @BristOliver, @_HannahRitchie and @puzzlesthewill
🗓️Tues 5 April, 3-5pm
👉rss.org.uk/training-events/events/events-2022/rss-events/… -
David Spiegelhalter
@d_spiegel
Big piece finally installed! Designed by my friend for Norfolk Broads house, reflecting her view of the reeds. Made in three panels, but with rather nice flow between them. Small picture shows the wonderful green/brown glass better. Other work here spiegelbull.com?page_id=11
-
David Spiegelhalter reposted
Royal Statistical Society
@RoyalStatSoc
📢You can now register to attend the first of our Covid-19 evidence sessions - to discuss the important statistical issues of the pandemic ahead of the official inquires
This session is themed around communication and is organised by @StatsJen, @d_spiegel & @TomChivers (1/3) -
David Spiegelhalter
@d_spiegel
Delighted to announce the winners of the inaugural #hardingprize for Useful and Trustworthy Communication, alongside @SMC_London and @senseaboutsci wintoncentre.maths.cam.ac.uk/news/winners-2021-harding-pr… two winners!
The ONS Covid infection survey and the Cochrane Review of Hydroxychloroquine for Covid-19
David Spiegelhalter
@d_spiegel
The prize of £3,141.59 will be shared (after all, it is World Pi day)
-
David Spiegelhalter reposted
Richard Evans
@RichardJaEvans
THREAD: On Dec 4 last year, Prof Norman Fenton said on LBC that "the [Covid] vaccines don’t seem to reduce all-cause mortality but rather produce a genuine spike in all-cause mortality shortly after vaccination". 1/10
-
David Spiegelhalter
@d_spiegel
if anyone happens to be in Valleta next Tuesday, I’m doing a gig at this wonderful venue www.kreattivita.org/en/event/cinexjenza-sex-murder-and-co…
-
David Spiegelhalter reposted
Office for Statistics Regulation
@StatsRegulation
We have written to @UKHSA and the @cabinetofficeuk following the use of unpublished data in a February COVID-19 press briefing.
As the data has since been published, we welcome transparency about cost estimates for the test and trace programme:
osr.statisticsauthority.gov.uk/correspondence/ed-humphers… -
David Spiegelhalter
@d_spiegel
This is brilliant. Both the code, and the final product. Can’t wait to use this when doing talks with @anthonybmasters !
-
David Spiegelhalter
@d_spiegel
this event is both online and in-person, and if you come along you can also register for a post-talk reception!
lse.eu.qualtrics.com/jfe/form/SV_4Vj12bYQ23w0mMe
LSE Events
@LSEpublicevents
7 March: David Spiegelhalter and Anthony Masters crunch the data on the COVID-19 pandemic, exposing the leading misconceptions about the virus and the vaccine, and answering our essential questions 🦠
🎟️ Register now!
@LSEStatistics #lsepostcovid
www.lse.ac.uk/Events/2022/03/202203071600/data -
David Spiegelhalter
@d_spiegel
a very useful primer from my colleagues at the Winton Centre
Ash Paul
@pash22
Communicating risks and benefits to cardiology patients heart.bmj.com/content/106/23/1862 via @mesotronium et al @dompagano @pomyers @VictorDayan1 @dr_benoy_n_shah @mmamas1973 @rahatheart1 @alf_collins @RichardLehman1
-
David Spiegelhalter reposted
LSE Statistics
@LSEStatistics
Don't miss out on our amazing PLP event on 7 March on COVID by Numbers with David Spiegelhalter & Anthony Masters!🦠📊
All those who register for & attend the event in person will be entered into a prize draw for a free copy of the book!📗Lucky things!
🔗www.lse.ac.uk/events/covid-by-numbers-making-sense-of-the…
LSE Events
@LSEpublicevents
🔜 Still time to register! #lsepostcovid
David Spiegelhalter and Anthony Masters crunch the data on the COVID-19 pandemic, exposing the leading misconceptions about the virus and the vaccine, and answering our essential questions 🦠
@LSEStatistics
www.lse.ac.uk/Events/2022/03/202203071600/data -
David Spiegelhalter reposted
Orry Robinson
@OrryRobinson
You can catch @d_spiegel and @anthonybmasters this Sunday 27th @niscifest as they crunch the data on the pandemic in Covid by Numbers #nisf22
BBC Radio Ulster
@bbcradioulster
🎧 Listen Back on @BBCSounds as @d_spiegel chats with @StevenRaineyNI about using data to make sense of #covid19
www.bbc.co.uk/sounds/play/p0bq6p1l -
David Spiegelhalter
@d_spiegel
This is a strong reprimand. But I guess the PM has other things on his mind today.
UK Statistics Authority
@UKStatsAuth
Sir David Norgrove has today written to the Prime Minister @10DowningStreet in regards to a statement on employment statistics. uksa.statisticsauthority.gov.uk/correspondence/sir-david-…
-
David Spiegelhalter
@d_spiegel
Extraordinary work
-
David Spiegelhalter reposted
-
David Spiegelhalter
@d_spiegel
But UKHSA continue to publish similar data. Dismal.
www.gov.uk/government/publications/covid-19-vaccine-weekl…
Paul Mainwood
@PaulMainwood
Don’t publish rates per head if you know full well your denominators are wrong. No, not even if it kind of looks OK and you’re sure it’ll all be fine.
Part 245 of a continuing series. -
David Spiegelhalter
@d_spiegel
I've seen this, and it looks excellent. And with Covid examples!
Sujit Sahu
@Sujit_K_Sahu
Glad to share a quick preview of the BMSTDR book - an introductory text on Bayesian modeling, computation, and spatial and spatio-temporal modeling with lots of examples and online R code! @RobCalver5 @CRC_MathStats @rstatstweet
-
David Spiegelhalter reposted
Hannah Fry
@FryRsquared
I think there are possible solutions on the horizon. Octopus, funded by UKRI and about to be launched, is one I’m particularly excited about: blog.science-octopus.org/update/2018/05/13/fixing-science…
-
David Spiegelhalter
@d_spiegel
excellent programme
www.bbc.co.uk/sounds/play/m0014ght
I think an image of Granny Mole would make a fine logo
BBC More or Less
@BBCMoreOrLess
Coming up at 9am on 📻@BBCRadio4 with @TimHarford:
- Children and Covid
- Covid modelling
- MPs' statistical savvy
- Nanomoles -
David Spiegelhalter reposted
John Roberts
@john_actuary
These two factors suggests that for every 100 deaths recorded under the 28 day measure, currently around 50 have COVID as the underlying cause, 25 to 30 have it as a contributory factor, and the balance are incidental deaths. But this distribution has only emerged recently.
4/ -
David Spiegelhalter reposted
Prof. Christina Pagel - @chrischirp.bsky.social
@chrischirp
THREAD: I keep being asked when we can go "back to normal" or "like it was before". My personal thoughts:
We've added a new disease to our population, more infectious and more severe than flu.
The world pre 2020 no longer exists - we may want it to, but it just doesn't. 1/13 -
David Spiegelhalter
@d_spiegel
Let’s look at this claim again. ONS now says 221 Covid deaths on 11th January, 77% ‘caused’ by Covid that week, so that’s ~ 170. Less than half 379. So this claim was untrue (which was clear at the time). But it got 3K retweets.
www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa… -
David Spiegelhalter
@d_spiegel
Summary: authors had already made up their minds, shoddy analysis, conclusions unjustified. eg Can’t add up re-analyses of same data as if independent new evidence! Marginal effects of mandatory measures may well be small. But this desperately poor ‘science’ doesn’t show it.
Health Nerd
@GidMK
This paper has been doing the rounds, claiming that lockdown was useless (the source of the 0.2% effect of lockdown claim). Dozens of people have asked my opinion of it, so here we go:
In my opinion, it is a very weak review that doesn't really show much, if anything 1/n -
David Spiegelhalter
@d_spiegel
I would like to strongly recommend this course from our Winton Centre - @mc_climent has done a wonderful job
Alexandra Freeman
@alex_freeman
A new course from the Winton Centre, on risk communication in public health!
Freely available at moodle.wintoncentre.uk
Should take less than 2hrs, and there's a certificate on completion.
We hope it's useful!
(Huge kudos to @mc_climent who put it together) -
David Spiegelhalter
@d_spiegel
changed now. And (time for my own correction) it was the Observer, not the Guardian.
David Spiegelhalter
@d_spiegel
appalling misleading headline from the Guardian: I said I didn’t take the pandemic ‘seriously enough’, but this is turned into ‘didn’t take Covid seriously’. Unacceptable.
-
David Spiegelhalter
@d_spiegel
appalling misleading headline from the Guardian: I said I didn’t take the pandemic ‘seriously enough’, but this is turned into ‘didn’t take Covid seriously’. Unacceptable.
-
David Spiegelhalter
@d_spiegel
Just to show I was not exaggerating on @DesIsDiscs, here is my Dad doing some escaping: note the 1950s Health and Safety comprising a man with a bucket of water
-
David Spiegelhalter
@d_spiegel
if anyone wanted to follow up on my Desert Island Discs tracks this morning, here is the family video of Numa Numa from 2007. Such a great song www.youtube.com/watch?v=WYg_gr52BA4
-
David Spiegelhalter
@d_spiegel
If you feel like a chortle over my (lack of) musical taste, listen to me talking to the wonderful @laurenlaverne on DID at 11am tomorrow, Radio 4. Rather rare to have a life’s ambition fulfilled.
-
David Spiegelhalter
@d_spiegel
To put recent Covid death figures in perspective, note that in Week 3 (ending Jan 21), fewer people were registered as dying due to all respiratory diseases (including Covid) than in the same week in 2015, 2017 and 2018. www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa…
-
David Spiegelhalter
@d_spiegel
Time to re-visit the claim that 379 UK deaths caused by Covid on Jan 11. I judged at the time that the true figure was less than half this. For Jan 11, dashboard coronavirus.data.gov.uk/details/deaths now says 264 ’28-day deaths’ and 217 registered occurrences (this will go up a bit) (1/2)
David Spiegelhalter
@d_spiegel
That week 77% of registered Covid-involved deaths had Covid as underlying cause www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa… So currently estimate 0.77 * 217 = 167 deaths caused by Covid on Jan 11. This will increase, but likely to be close to half of 379. (2/2)
-
David Spiegelhalter
@d_spiegel
Apologies. On @BBCMoreOrLess I just said that vaccine effectiveness estimates were based on Cox models: should have been clear that this is for effectiveness against *mortality*, but not symptomatic disease.
-
David Spiegelhalter
@d_spiegel
this is a great opportunity with a project that is aiming to completely change scientific publishing - the current model being broken
Octopus.ac
@octopus_ac
Dev opportunity to join the team and help build Octopus from the ground up 🏗️🐙 jobs.jisc.ac.uk/Jisc/JobDescription/C22Zb_Ni23k
-
David Spiegelhalter
@d_spiegel
So pleased to see the @royalsociety and @BritishAcademy_ giving such strong support to the excellent Core Maths programme www.thebritishacademy.ac.uk/news/core-maths-qualification…
-
David Spiegelhalter reposted
Office for Statistics Regulation
@StatsRegulation
We welcome @ONS’s clarification on data about those who have died from COVID-19. Read our statement 👉osr.statisticsauthority.gov.uk/news/osr-welcomes-ons-blog…
-
David Spiegelhalter reposted
ONS Focus
@ONSfocus
In our blog @ONSJames explains why it is both factually incorrect and highly misleading to say only 17,000 people have died of #covid19. This claim is based on the number of deaths where no other health conditions were noted on the death certificate. blog.ons.gov.uk/2022/01/26/to-say-only-17000-people-have-…
-
David Spiegelhalter
@d_spiegel
looking forward to appearing with this lot!
5x15
@5x15stories
Announcing @AdamRutherford @d_spiegel @JustinOnWeb and @EdwardShawcross at 5x15 in February ! More speakers TBA - mailchi.mp/5x15stories.com/feb5x15202222-1076898
-
David Spiegelhalter
@d_spiegel
really looking forward to doing this gig in Belfast with @anthonybmasters !
NI Science Festival
@niscifest
Twitter! It's Monday. We're just under 4 weeks away from the start of #nisf22!
Over 150 events across 11 days feat @theAliceRoberts, @anilkseth, @d_spiegel, @PastMortems, @ASIWYFA_BAND, @laoneill111, @itscolinblack, @maddiemoate & more!
Book now - nisciencefestival.com/events -
David Spiegelhalter reposted
stuart mcdonald
@ActuaryByDay
1. Saying there have been “only” 17,371 COVID deaths with no other underlying causes is misleading. This has been repeatedly addressed, including this thread from earlier today. Most older people have one or more pre-existing conditions. It is wrong to exclude their deaths.
Colin Angus
@VictimOfMaths
I've seen a few people recently shouting about how new ONS data shows that the "true" death toll from COVID in England & Wales is only 17,371, rather than the ONS figure of 157,816.
This is obviously nonsense, but evidently still needs debunking, so here goes... -
David Spiegelhalter
@d_spiegel
Thanks for pointing this out. I’ve contributed to this misunderstanding, and will now try and drop the ‘with’ term completely
-
David Spiegelhalter
@d_spiegel
Excellent animation showing effect of day of week on daily death reports. And all those headlines about deaths ‘soaring’ 🙁
-
David Spiegelhalter
@d_spiegel
On Dec 31st I guessed UK infections could be close to 500,000 a day. ONS survey now estimates 479,100 infections on 29th Dec www.ons.gov.uk/peoplepopulationandcommunity/healthandsoci… Not a bad call. Extraordinary numbers: around 1 in 140 people *per day*, but now dwarfed by France, Portugal, Denmark,..
-
David Spiegelhalter
@d_spiegel
I am so sad about this. David was a hero to statistics in general, and me in particular. He took me apart, in the nicest possible way, in my PhD Viva 43 years ago, probably after skimming it on the train. 1/
Stian Westlake
@stianwestlake
We're deeply saddened to learn of the death of Sir David Cox, a giant of statistics and a dear friend, supporter, and past president of the RSS.
David Spiegelhalter
@d_spiegel
He + his wartime generation set tone for UK stats; strongly opinioned, but without ego and bitterness. Daunting to have him in front row of audience: he would listen for a bit, see what was coming, then apparently fall asleep until the end, then ask most incisive questions 2/
David Spiegelhalter
@d_spiegel
I used to be mystified by his emphasis on statistics being the understanding of sources of variation – now that’s exactly how I see it.
I shall so miss his wisdom - he was not always gentle in his opinions, but he was always a gentleman. -
David Spiegelhalter reposted
Stian Westlake
@stianwestlake
We're deeply saddened to learn of the death of Sir David Cox, a giant of statistics and a dear friend, supporter, and past president of the RSS.
-
David Spiegelhalter
@d_spiegel
Tim is right - it’s very good, in spite of my interjections. Thanks to the great guests, and @ilangoodman for creating it.
Tim Harford
@TimHarford
This episode (riskytalk.libsyn.com/how-misinformation-spreads) of @RiskyTalkPod from @PuzzlesTheWill @ChloeColliver2 @LauraEdelson2 @Sander_vdLinden and @d_Spiegel discussing misinformation and what to do about it, was outstanding.
-
David Spiegelhalter
@d_spiegel
Brilliant. But since logistic regression is now considered 'machine learning', maybe Bayesian modelling will start being called ‘AI’? 😬
Paul Mainwood
@PaulMainwood
- "Is there any other point to which you would wish to draw my attention?"
- "To the curious role Artificial Intelligence played in the unprecedentedly data-rich environment of the COVID-19 epidemic; in predicting its course and shaping governmental response."
- "But Holmes ..." -
David Spiegelhalter
@d_spiegel
this was fun to do - tricky questions. But does contain the answer to that crucial question ‘How many beans make five?"
always possible
@always_possible
In this week’s 5 Big Questions interview we talk to one of the world’s pre-eminent statisticians, government advisers, and public science communicators, PROFESSOR SIR DAVID SPIEGELHALTER OBE FRS
www.alwayspossible.co.uk/podcast/david-spiegelhalter/ -
David Spiegelhalter
@d_spiegel
I prefer it the first way round!
-
David Spiegelhalter
@d_spiegel
I’ve been taken off the Sage list, as I only attended one meeting in 2020. So please could journalists no longer refer to me as a member of Sage? Thanks!
-
David Spiegelhalter
@d_spiegel
Remember Health Sec telling Parliament on Dec 13th that UKHSA estimated 200,000 Omicron infections a day in England? It now looks like this was around twice the true figure on that day, although roughly right a few days later inews.co.uk/news/health/covid-cases-daily-uk-omicron-vari…
David Spiegelhalter
@d_spiegel
ONS infection survey now estimates 180,000 infections on Dec 13th www.ons.gov.uk/peoplepopulationandcommunity/healthandsoci… UKHSA estimates about 60% Omicron (%SGTF on tests on 16th). So around 100,000 Omicron infections on 13th. But probably around 200,000 on Dec 16th, so he was only a few days out
-
David Spiegelhalter
@d_spiegel
5 things I can talk about for 30 minutes without preparation:
1. Covid stats
2 WW2
3 My children
4 Bayesianism
5 Walking
Hire me for 2.5 hours and get the lot 😀 (my enormous fee will go to charity)
Meaghan Kall has moved to Bluesky
@kallmemeg
5 things I can talk about for 30 minutes without preparation:
1. Current COVID-19 epidemiology
2. Survey design
3. The strengths and limitations of quality of life measures
4. RuPaul’s Drag Race
5. Adapting a residential property for a wheelchair user -
David Spiegelhalter reposted
Nick Triggle
@NickTriggle
Triggled? 😆 the figure is from FluMOMO used by PHE to work out flu deaths. More than 22k deaths that winter. Winters up to start of pandemic here assets.publishing.service.gov.uk/government/uploads/syste…
Dr. Deepti Gurdasani
@dgurdasani1
Um, literally just look up the ONS data... deaths from flu are ~1500 *per year*. How were 300 people dying from flu a day? We're being 'triggled'
Triggled: [verb] spun to misinform with propaganda to normalise and minimise death and suffering from SARS-CoV-2. (see also hopium) -
David Spiegelhalter reposted
NI Science Festival
@niscifest
Our full programme is revealed this Thursday w/ over 100 events celebrating the science of YOU, the Universe & everything in-between!
Events already on sale @theAliceRoberts, @d_spiegel, @maddiemoate, @gastronautTV, @DanSaladinoUK & @anilkseth!
Book now nisciencefestival.com/events -
David Spiegelhalter
@d_spiegel
Good news: admissions for flu not quite as tiny as last year, but close.
www.gov.uk/government/statistics/national-flu-and-covid-1… -
David Spiegelhalter
@d_spiegel
If now no need for confirmatory PCR after asymptomatic + LFD, why is PCR necessary after *symptomatic* + LFD? When chance of false positive is even lower? Genuinely unsure what the logic is.
-
David Spiegelhalter reposted
Barry Tennison
@Bari10
This is a VERY good podcast, with expert talk about social media, mis- and dis-information, what is known and what might be done (not very optimistic, though…)
Risky Talk
@RiskyTalkPod
Social media is awash with dodgy information. In the latest ep we find out:
🕵️ Who is pushing this content online and why?
💸 What are the different groups trying to achieve?
🏛️ What should we do?
👇 LISTEN HERE 👇
riskytalk.libsyn.com/how-misinformation-spreads -
David Spiegelhalter
@d_spiegel
After 50 short article in the Observer, @anthonybmasters and I try to sum up the experience of writing about Covid stats in 350 words. Our thanks to those who put the data out there, and who we learnt from! It’s been worth the occasional unpleasantness. www.theguardian.com/commentisfree/2022/jan/02/2021-year-w…
-
David Spiegelhalter
@d_spiegel
Not sure I deserve such kind comments from @AtheneDonald , but thanks all the same. And I realise that our abuse is trivial compared to what others suffer
Athene Donald
@AtheneDonald
In praise of scientists who are willing to speak up about #pandemic evidence and facts. Thanks to @d_spiegel @devisridhar @Sander_vdLinden and many others! occamstypewriter.org/athenedonald/2022/01/03/scientists-w…
-
David Spiegelhalter reposted
John Aston
@HardingProf
And I am really grateful for David Harding's continuing support of me as the Harding Professor of Statistics in Public Life
David Spiegelhalter
@d_spiegel
Delighted to see David Harding get a knighthood today. As part of his philanthropy, he supported me as Winton Prof (2007-2019) and the Winton Centre (2016-). He changed my life, and has done so much for the public communication of stats, risk and evidence. en.wikipedia.org/wiki/David_Harding_(financier)
-
David Spiegelhalter
@d_spiegel
Delighted to see David Harding get a knighthood today. As part of his philanthropy, he supported me as Winton Prof (2007-2019) and the Winton Centre (2016-). He changed my life, and has done so much for the public communication of stats, risk and evidence. en.wikipedia.org/wiki/David_Harding_(financier)
-
David Spiegelhalter
@d_spiegel
and so say all of us!
Phil Rodgers
@PhilRodgers
A particularly big thank you today to Covid data hero Pouria Hadjibagheri, a key member of the amazing team behind the UK Coronavirus Dashboard. Thank you @pouriaa and very best wishes for the future.
Pouria Hadjibagheri MBE
@Pouriaaa
Some personal news (a 🧵):
On 4 October 2021, I resigned from my role as the Deputy Head of Software Engineering for Web and Data at @UKHSA.
This means that 1 January 2022 will mark my final day in the agency — but *not* on the dashboard, so this is not goodbye.
...[1/10] -
David Spiegelhalter
@d_spiegel
really valuable explanation
-
David Spiegelhalter
@d_spiegel
Seems a bit harsh, just for being on the BBC (also I stopped being "Prof for Public Understanding of Risk" 2 years ago)
-
David Spiegelhalter reposted
Randall Munroe
@xkcd
Confounding Variables xkcd.com/2560
-
David Spiegelhalter
@d_spiegel
Just in time for Christmas, completed project for my daughter’s flat. designed by her friend - I couldn’t come up with anything so interesting. All clear glass, although some textured not very visible from this angle. Option to paint later. (Mug included for scale)
-
David Spiegelhalter
@d_spiegel
Great advice for dealing with vaccine-conspiratorial relatives at Christmas, by my colleague Sander van der Linden, apparently Cambridge’s (informal) "defence against the dark arts" teacher
www.theguardian.com/commentisfree/2021/dec/23/covid-myths… -
David Spiegelhalter
@d_spiegel
For decisions about money, health, or just what to believe, who has given you balanced, trustworthy info? Enter them, or yourself, for our £3,141.59 prize. Also a Weasel Words award for the best/worst example of persuasion dressed up as ‘just informing’ hardingprize.online
-
David Spiegelhalter
@d_spiegel
Sunrise in frosty Norfolk
-
David Spiegelhalter
@d_spiegel
In which @anthonybmasters and I feature number theatre, the rule of 72, natural logarithms, and the non-inexorable growth of Elvis Presley impersonators. And Omicron. www.theguardian.com/theobserver/commentisfree/2021/dec/19… (How long can we keep churning this stuff out, I hear you cry?)
-
David Spiegelhalter
@d_spiegel
Personal tweet: This is a very deserving cause, trying to mitigate the impact of the pandemic restrictions on marginalised women and girls (COI: I am very proud to be a patron of this great charity)
Karuna
@karuna_trust_uk
December is here and our Christmas campaign is underway! Can you support women and girls out of the shadows and into lives of freedom, autonomy & safety? Our goal is £75,000 for work that supports ending the #shadowpandemic. To donate and share: www.karuna.org/shadow-pandemic
Thank you! -
David Spiegelhalter
@d_spiegel
In which they have to explain why Kenneth and I start arguing right from the start!
Podcasts from The Economist
@EconomistPods
“You can measure some things but you have to infer what’s going on underneath. That is difficult and it leaves room for a huge variation in interpretation.” @d_spiegel of @Cambridge_Uni on making sense of data during the pandemic, on our “Babbage” podcast econ.trib.al/xIUIMfi
-
David Spiegelhalter reposted
Podcasts from The Economist
@EconomistPods
“You can measure some things but you have to infer what’s going on underneath. That is difficult and it leaves room for a huge variation in interpretation.” @d_spiegel of @Cambridge_Uni on making sense of data during the pandemic, on our “Babbage” podcast econ.trib.al/xIUIMfi
-
David Spiegelhalter
@d_spiegel
OSR to UKHSA re the 200,000 infections - "the delay between the use of the figure and the publication of the data was unsatisfactory”. But rightly congratulating for the daily Omicron reports
Office for Statistics Regulation
@StatsRegulation
We have written to @UKHSA about data to support Ministerial statements on the Omicron variant. osr.statisticsauthority.gov.uk/correspondence/ed-humphers…
-
David Spiegelhalter
@d_spiegel
Brilliant article by Michael Blastland on informing vs persuading “the temptation to apply a subtle thumb to the scale is strong”, and promoting our Harding Prize
www.timeshighereducation.com/blog/unvarnished-truths-brid… (need to register) hardingprize.online -
David Spiegelhalter
@d_spiegel
UKHSA have finally published their methodology behind the 200,000 infections quoted by Javid on Monday. But press got to see this on Tuesday. www.gov.uk/government/publications/covid-19-omicron-daily… Very poor practice
David Spiegelhalter
@d_spiegel
Office for Statistics Regulation: "Where models are referred to publicly, particularly to inform significant policy decisions, the model outputs, methodologies and key assumptions should be published at the same time" osr.statisticsauthority.gov.uk/why-trust-and-transparency…
-
David Spiegelhalter
@d_spiegel
Thanks to @ilangoodman!
Tim Harford
@TimHarford
Wow, the latest @RiskyTalkPod with @puzzlesthewill @ChloeColliver2 @LauraEdelson2 @Sander_vdLinden and @d_spiegel is absolutely outstanding.
riskytalk.libsyn.com/how-misinformation-spreads -
David Spiegelhalter
@d_spiegel
Has UKHSA publicly released a source for their 200,000 on Dec 13 figure? Clearly press have been briefed news.sky.com/story/covid-19-why-health-secretarys-200k-om…
-
David Spiegelhalter
@d_spiegel
Great PhD opportunity in "Public Expertise, Scientific Evidence, and Political Decision-making: Statistical and Public Policy Perspectives”. with a fine Cambridge team www.jobs.cam.ac.uk/job/32744/
-
David Spiegelhalter
@d_spiegel
This is a great episode with amazing guests. I learned so much.
Risky Talk
@RiskyTalkPod
Social media is awash with dodgy information. In the latest ep we find out:
🕵️ Who is pushing this content online and why?
💸 What are the different groups trying to achieve?
🏛️ What should we do?
👇 LISTEN HERE 👇
riskytalk.libsyn.com/how-misinformation-spreads -
David Spiegelhalter
@d_spiegel
Did this a while ago - but interesting to apply to Covid. There’s a lot of hedgehogs around …
-
David Spiegelhalter
@d_spiegel
and we also point out the one bit of good news for the NHS
-
David Spiegelhalter reposted
stuart mcdonald
@ActuaryByDay
The CMI report I referred to earlier has now been published.
Analysis of death rates shows that most of the excess reported by ONS this week is explained by population growth and ageing. The age-standardised excess was lower than the number of covid deaths.
COVID-19 Actuaries Response Group
@COVID19actuary
The Continuous Mortality Investigation (CMI) has published its latest Mortality Monitor to 26 November (week 47).
There were 2% more deaths this week than if death rates were the same as week 47 of 2019. That is 266 excess deaths in England and Wales this week. -
David Spiegelhalter
@d_spiegel
the interactive tool I just mentioned on Sky News www.bmj.com/content/375/bmj-2021-065312.full?ijkey=PuBVt2… Try exploring the effect of cleaning surfaces….
-
David Spiegelhalter
@d_spiegel
📢Nominations are open for the - 'Harding Prize for Trustworthy Communication’📢 For your own/ or others’ work. Prize? The glory! And £3,141.59 [we’re in a maths dept]. There is also a 'Weasel Words’ prize (£0) for persuasion dressed up as information hardingprize.online
-
David Spiegelhalter
@d_spiegel
Getting stick for saying on #r4today that govt policy ‘seems to have rather worked’. Not well-worded: I meant according to its own objectives of a delicate (now precarious?) balancing act, without the dire consequences some predicted. Other policies may have reduced Covid harm.
-
David Spiegelhalter
@d_spiegel
Just play with this! Check the estimated effects of visors, screens, hand-hygiene etc. Brilliant work by Will Stahl-Timmins @will_s_t and BMJ team. As I said on #r4today this morning, these estimates needed to be from expert judgements rather than statistical analysis.
Alexandra Freeman
@alex_freeman
How much difference does it make to open the windows?
Wear a mask?
What if we're eating?
Questions we've all been asking about covid - hoping to help answer them with this new tool:
www.bmj.com/content/375/bmj-2021-065312.full?ijkey=PuBVt2…
@d_spiegel @CathNoakes @harryrutter @will_s_t -
David Spiegelhalter
@d_spiegel
Personal message: This is an extremely deserving cause, given the particularly tragic effect of the pandemic on marginalised communities. Please donate. (COI: I am honoured to have joined Judi Dench as a patron of the Karuna Trust, and they get my speaking fees)
-
David Spiegelhalter
@d_spiegel
Strong letter from Chair of UK Statistics Authority to Dominic Raab on the leak of transgender prison stats to the Telegraph: “Such leaks risk both misinforming the public and undermining trust in official statistics.” 80 people had prior access!
uksa.statisticsauthority.gov.uk/correspondence/sir-david-… -
David Spiegelhalter
@d_spiegel
In which @anthonybmasters and I argue that meta-analysis is not some alchemy for turning leaden studies into gold. Wanted uncontroversial examples, so picked masks and ivermectin…. www.theguardian.com/theobserver/2021/nov/28/ivermectin-ma…
-
David Spiegelhalter
@d_spiegel
good to see this focus on the care needed to prevent stats being misused
Office for Statistics Regulation
@StatsRegulation
It is critical that producers of statistics do all they can to support users to interpret the data they publish. Today we have written to @ONS regarding its Deaths involving COVID-19 by vaccination status publication osr.statisticsauthority.gov.uk/correspondence/ed-humphers…
-
David Spiegelhalter
@d_spiegel
anyone fancy being Chair of the UK Statistics Authority? publicappointments.cabinetoffice.gov.uk/appointment/uk-st…
-
David Spiegelhalter
@d_spiegel
The booster data is very impressive, but could be better framed: 95% protection means 5% of the vulnerability of unvaccinated. Fortunately the UK has been bolstering away, compared with other countries www.theguardian.com/theobserver/commentisfree/2021/nov/21…
David Spiegelhalter
@d_spiegel
I meant ‘boostering’. But Twitter’s spell-correction to ‘bolstering’ seems quite reasonable
-
David Spiegelhalter
@d_spiegel
in which @anthonybmasters and I discuss how we know about the extraordinary effectiveness of boosters. And how negative framing might better communicate their importance www.theguardian.com/theobserver/commentisfree/2021/nov/21…
-
David Spiegelhalter
@d_spiegel
Our latest paper on how people react to quality-of-evidence scales. Crucially, they take notice of them.
Claudia R Schneider
@CRSchneider3
If you tell people the quality of the evidence behind a piece of public health advice, does it affect their response to it? Our work suggests –yes! + in ways important for ethical comms
All in paper out now, with @alex_freeman @d_spiegel @Sander_vdLinden
journals.plos.org/plosone/article?id=10.1371/journal.pone… -
David Spiegelhalter
@d_spiegel
‘Negative framing’ for vaccines? Protection against hospn after 2 Pfizer jabs waned from 99.7% to 92.7% after 20 weeks. May not sound much, but it means vulnerability relative to being unvaccinated went from 0.3% to 7.3%, more than 20-fold increase in risk www.bbc.co.uk/news/health-59260294
-
David Spiegelhalter
@d_spiegel
Encouraging new booster study. Regardless of initial vaccine, booster appears to give similar protection (93-94%) against symptomatic disease as seen soon after 2 Pfizer doses. Means those who initially had AZ will have more protection than they ever had www.gov.uk/guidance/monitoring-reports-of-the-effectivene…
-
David Spiegelhalter
@d_spiegel
In which we feature 4 recent statistical nasties, demonstrating the need for trustworthy communication
-
David Spiegelhalter
@d_spiegel
such a tricky issue - often asked about this
Gabriel Recchia
@mesotronium
We’ve got some new research out on how to incorporate uncertainty into icon arrays/cates plots/pictographs: www.sciencedirect.com/science/article/pii/S2772628221000030
These plots are great – but can you indicate uncertainty in them without confusing people? -
David Spiegelhalter reposted
Owen Arands
@OwenArandz
Just finished this brilliant book by @d_spiegel and @anthonybmasters, covers practically everything and should be mandatory reading for all journalists.
-
David Spiegelhalter
@d_spiegel
I often say that I make every mistake when dealing with the media. Just to prove it, just done a live interview on Zoom for Sky News in which I forgot to unmute myself…
-
David Spiegelhalter
@d_spiegel
This was such fun, being allowed to cover the meaning of probability, Bayesian modelling, communication risks, Covid, etc. Thanks Alexandre!
Alexandre Andorra
@alex_andorra
Special ep. of LBS: had the chance to talk w. @d_spiegel, the author of "#covid by Numbers" & "The Art of Statistics" 🤩
Talked about #chance, #risk, #uncertainty, #decisionmaking, the game of loop & special hats 🧢
See U in your favorite podcatcher or @ learnbayesstats.com/episode/50-talking-risks-embracing-un… -
David Spiegelhalter reposted
Full Fact
@FullFact
In October, Brazilian President Jair Bolsonaro told the world that people vaccinated against Covid were developing AIDS—and claimed English data proved it.
We reveal how misleading data from the @UKHSA has fuelled a global vaccine myth 👇
fullfact.org/health/phe-ukhsa-bolsonaro -
David Spiegelhalter reposted
Will Moy
@puzzlesthewill
A thread, which will inevitably not do justice. I think they have shown leadership on this. @RoyalStatSoc invited me to give the Beveridge Lecture on this in 2018 - video here rss.org.uk/training-events/events/key-events/beveridge-le… (thank you @d_spiegel) >
Colin Angus
@VictimOfMaths
OK everyone, can somebody *please* organise a proper, grown up discussion about the misrepresentation of statistics in public life, and what (if anything) we expect the producers of those statistics to do about it.
Feels like something @RoyalStatSoc or @UKStatsAuth should do... -
David Spiegelhalter
@d_spiegel
"Producers of data should do everything they can to minimise the potential for misinterpretation or misuse."
Office for Statistics Regulation
@StatsRegulation
If data are not presented and explained well, it can be harmful. In our latest blog, Deputy Director Mary Gregory discusses issues surrounding data published by @UKHSA osr.statisticsauthority.gov.uk/communicating-data-is-more…
-
David Spiegelhalter
@d_spiegel
if you’re in Cambridge early this evening, I’m doing a real true-life in-person talk!
Cambridge Philosophical Society
@CamPhilSoc
Should we trust statistics?
Professor Sir David Spiegelhalter with give Honorary Fellows Lecture next Wednesday 3rd November at 18.30
BOOK NOW: www.cambridgephilosophicalsociety.org/events/event/honora… -
David Spiegelhalter reposted
Royal Statistical Society
@RoyalStatSoc
We still have a few places left for our Q&A with
@d_spiegel and @anthonybmasters on their new book 'Covid by Numbers'
🗓️ Friday 5 November, 12-1pm (GMT)
Register here 👇
www.rss.org.uk/training-events/events/events-2021/rss-eve… -
David Spiegelhalter
@d_spiegel
This is about the apparent case-rates per 100,000 being higher in vaccinated than unvaccinated groups - the issue was pointed out weeks ago, including by us www.theguardian.com/theobserver/commentisfree/2021/sep/19…
Office for Statistics Regulation
@StatsRegulation
We have written to @UKHSA about its COVID-19 vaccine surveillance statistics.
osr.statisticsauthority.gov.uk/correspondence/ed-humphers… -
David Spiegelhalter
@d_spiegel
@shinecancersupport is running #shineconnect, a free, online conference for younger adults with cancer, 13 - 18 Nov. Register here: cvent.me/5xGQed (I spoke at this excellent meeting last year)
-
David Spiegelhalter
@d_spiegel
our piece of detective work….
-
David Spiegelhalter reposted
John Caudwell
@JohnDCaudwell
Morning! Tune in to @BBCRadio4 at 9.40am when I'll be on Broadcasting House with @bbcpaddy doing the Sunday papers review with fellow panelists Professor @d_spiegel and political journalist @nadinebh_.
Which stories have caught your eye this morning? #bbcbh #inthenews -
David Spiegelhalter reposted
CUNY GC Ed Psych
@CunyGCEdPsych
. @AmstatNews & @CunyGCEdPsych will host a symposium on teaching statistics Nov. 2 @ 2pm! Click below to register. Speakers: David Draper, @StatModeling Andrew Gelman, @nzgwynn, @d_spiegel & prof. David Rindskop!
@ucsantacruz, @Columbia, @Harvard, @Cambridge_Uni, @GC_CUNY -
David Spiegelhalter
@d_spiegel
Choosing a 2-day post-trip LFT provider? I thought Boots/Recova-19 sounded Ok, but should have checked the reviews uk.trustpilot.com/review/recova-19.com (Hint: when the website says ‘bar code on sample tube’, it should say ‘QR code on test cassette’)
-
David Spiegelhalter reposted
Dan Cookson
@danc00ks0n
A rare thing. A 3D chart that’s actually useful
-
David Spiegelhalter
@d_spiegel
tomorrow at 6 I will be at Waterstones in Cambridge talking to Talya Miron-Shatz about her excellent new book on how to improve decision-making about health care - both for patients and professionals
www.eventbrite.co.uk/e/your-life-depends-on-it-cambridge-… -
David Spiegelhalter
@d_spiegel
Important comments on transparency from OSR (COI, I’m a non-exec on the UKSA Board)
Office for Statistics Regulation
@StatsRegulation
Over the past 18 months we’ve talked a lot about transparency. In our latest blog we look at what we mean when we talk about transparency, and what we can all do to support it osr.statisticsauthority.gov.uk/transparency-open-communic…
-
David Spiegelhalter
@d_spiegel
Completely unacceptable that UKHSA put out absurd statistics showing case-rates higher in vaxxed than non-vaxxed (Fig 2), when just an artefact of using hopelessly biased NIMS popn estimates. Feeding conspiracy theorists worldwide.
www.gov.uk/government/publications/covid-19-vaccine-weekl… -
David Spiegelhalter reposted
Paul Pharoah (R soul) 🇪🇺🇵🇬🇿🇦🇬🇧
@paulpharoah
@Cambridge_Uni have published an account of the development of the PREDICT breast and prostate models. Thanks to many including @ProfWishart @d_spiegel @alex_freeman @mesotronium
www.cam.ac.uk/stories/predicting-better
Donations to keep the project alive would of course be welcome :-) -
David Spiegelhalter
@d_spiegel
This article is a bit more challenging than usual, but tries to get over the idea that a positive LFD is stronger evidence *for* the virus than a negative PCR is *against* the virus. So reasonable to be cautious after +ve LFD and -ve PCR
-
David Spiegelhalter
@d_spiegel
This looks a great job! Looking for both technical and communication skills
John Aston
@HardingProf
Im advertising a (hopefully!) really interesting job combining statistics, public policy and communication. Working with the Winton Centre @alex_freeman and @d_spiegel, and potentially many others across Cambridge including @BennettInst
Deadline 14 Nov.
www.jobs.cam.ac.uk/job/29725/ -
David Spiegelhalter
@d_spiegel
Great to be talking to Michael again!
Risky Talk
@RiskyTalkPod
Michael Blastland puts @d_spiegel and @anthonybmasters in the hot-seat: how should experts communicate about excess deaths, vaccine harms, lockdowns and more? riskytalk.libsyn.com/covid-by-numbers
-
David Spiegelhalter
@d_spiegel
$1,000 prize for the best ’small-world’ coincidence! Some great stories there already
www.yoursmallworldstory.com -
David Spiegelhalter
@d_spiegel
Amazon now charging the cover price for our book. So why not support your local bookshop instead?
-
David Spiegelhalter
@d_spiegel
our latest Risky Talk podcast! produced by the brilliant @ilangoodman
Alexandra Freeman
@alex_freeman
Out now - the latest RiskyTalk podcast from the Winton Centre, featuring @d_spiegel and @anthonybmasters talking about the lessons learned on communicating Covid statistics during the past 18 months.
Produced by @ilangoodman @RiskyTalkPod
riskytalk.libsyn.com -
David Spiegelhalter reposted
David Spiegelhalter
@d_spiegel
very excited about appearing in person at the Sidmouth Science Festival on the 15th sidmouthsciencefestival.org/events/communicating-statisti… (even though, for a North Devon boy, it’s a foreign country)
-
David Spiegelhalter
@d_spiegel
a unique chance to see both me and @anthonybmasters together! Monday at 10am in Cheltenham literature.cheltenhamfestivals.online/events/covid-by-num… (we never met when writing the book)
-
David Spiegelhalter
@d_spiegel
For anyone who has got our book, www.penguin.co.uk/books/446/446585/covid-by-numbers/97802…
you can now download a pdf with web-links for all the end-notes
github.com/covid-by-numbers/covid-by-numbers.github.io/bl… -
David Spiegelhalter
@d_spiegel
my take on OSR’s ‘Lessons Learned'
Office for Statistics Regulation
@StatsRegulation
“The pandemic has shown how much can be achieved with the right will and appropriate resources, and OSR’s ‘Lessons Learned’ point the way forward.”
@d_spiegel gives his reflections on lessons learnt for health and social care statistics.
osr.statisticsauthority.gov.uk/reflections-on-lessons-lea… -
David Spiegelhalter
@d_spiegel
extraordinary blog by @anthonybmasters on the effort in setting up reproducible code to extract graphs from ONS spreadsheets
anthonybmasters.medium.com/life-satisfaction-and-data-wra… (Of course this explains why he has done all the wonderful coding for our book!) -
David Spiegelhalter
@d_spiegel
very excited about appearing in person at the Sidmouth Science Festival on the 15th sidmouthsciencefestival.org/events/communicating-statisti… (even though, for a North Devon boy, it’s a foreign country)
-
David Spiegelhalter
@d_spiegel
Very pleased to be doing this. But, to avoid disappointment, I should be clear that I will certainly not be saying what will happen next!
Tortoise
@tortoise
.@d_spiegel has been described as a “superstar statistician” and as “Britain’s statistical national treasure”.
Join us for a ThinkIn in which he will debunk Covid misconceptions and explain, definitively, how we got here and what will happen next.
torto.se/3ozUqHC -
David Spiegelhalter
@d_spiegel
excellent lessons! I like the emphasis on independent leadership, transparency, collaboration, and data useful for evaluation. (COI: I am a non-exec on UKSA, who oversees the work of OSR).
Office for Statistics Regulation
@StatsRegulation
Today we published our report: Improving health and social care statistics: lessons learned from the COVID-19 pandemic.
osr.statisticsauthority.gov.uk/publication/improving-heal… -
David Spiegelhalter
@d_spiegel
excellent article by @VictimOfMaths (@anthonybmasters and I had been thinking of covering this in the Observer this week, but will need to think again..) theconversation.com/how-covid-deaths-compare-with-other-d…
-
David Spiegelhalter
@d_spiegel
Amazon does some very odd classification of books...
-
David Spiegelhalter
@d_spiegel
wow, great to be in the fine company! Why not get them all for the nerd in your life?
Tim Harford
@TimHarford
This is a VERY exciting week for nerdsome books. Four amazing books are being published.
I'm going to go from jovial to serious, but each one is well worth your attention... -
David Spiegelhalter
@d_spiegel
Some people in Norfolk seem to be preparing for rising sea levels 😉. But where are these wonderful conversions?
-
David Spiegelhalter
@d_spiegel
Quite a tricky puzzle to do face down - but wood grain and pencil marks helped. Shame about 2 missing pieces
-
David Spiegelhalter reposted
Peter Wells
@peterkwells
"MPs and civil servants called to account for misleading use of statistics" www.theguardian.com/politics/2021/oct/03/mps-and-civil-se…
OSR annual review: osr.statisticsauthority.gov.uk/publication/annual-review-… -
David Spiegelhalter
@d_spiegel
our first review! So pleased to see @BristOliver liked the book @anthonybmasters www.theguardian.com/books/2021/sep/30/covid-by-numbers-re…
-
David Spiegelhalter
@d_spiegel
the problem with counting some thing when it’s not agreed what it is
-
David Spiegelhalter reposted
Alexandra Freeman
@alex_freeman
Very glad to be able to announce that Predict:Breast Cancer (breast.predict.nhs.uk) is now available in both traditional and simplified Chinese. This hugely increases the number of patients who can now be helped with personalised treatment information
@paulpharoah @d_spiegel -
David Spiegelhalter
@d_spiegel
My colleagues have done this fine empirical work on risk matrices - I really like the idea of moving away from a symmetric grid to give a feel for non-linearity. About time that this familiar tool was investigated!
Gabriel Recchia
@mesotronium
Risk matrices are very familiar to those involved in health & safety and risk assessments – but do people actually understand them? We took a first look, published today: doi.org/10.1111/risa.13822
-
David Spiegelhalter
@d_spiegel
A chance to meet the stats watchdog! (COI; I'm a non-exec on UKSA, that oversees their work)
Office for Statistics Regulation
@StatsRegulation
What has it been like to be the statistics watchdog during a year when public interest in statistics has been so high?
Join us to hear about our work during the pandemic and how we regulate official statistics across the UK. 👇
www.eventbrite.co.uk/e/meet-the-uk-statistics-watchdog-ti… -
David Spiegelhalter
@d_spiegel
so pleased to be contributing to this
Prof Mark Lorch
@Mark_Lorch
A big #ff to the fab folk working to make #greatexplanations happen
@LordMartinRees @helenczerski @Kat_Arney @chronicleflask @scarlett_jazmin
@AdamHartScience @andymiah @FunSizeSuze
@garwboy @ChemistryKit @TheLabArtist
@d_spiegel @matthewcobb @unbounders @senseaboutsci 1/2 -
David Spiegelhalter
@d_spiegel
Yes we know More or Less covered this on Wednesday www.bbc.co.uk/sounds/play/m000zkzq But we wrote this on Tuesday! Interesting that both teams came up with an almost identical narrative structure.
-
David Spiegelhalter
@d_spiegel
Isn’t she brilliant?
-
David Spiegelhalter reposted
RSS History Section
@HistoryofStats
📈Keynote 1: RSS and the pandemic📉
🎙@Deborah_Ashby, Sheila Bird,@simonbriscoe & @anthonybmasters
🗓Monday, 6th September 2021
🕠17:30 - 18:45
📌Exchange Auditorium/ Livestream Plenary -
David Spiegelhalter
@d_spiegel
Always best to check the numbers actually represent!
-
David Spiegelhalter
@d_spiegel
there have been many comparisons between risks of vaccines for young people and risks if they get Covid. But surely the appropriate comparison is with not being vaccinated, which does not necessarily mean getting Covid www.bbc.co.uk/news/health-58423152
-
David Spiegelhalter
@d_spiegel
excellent clarification: analogies need to be used carefully, but this seems a very good one.
-
David Spiegelhalter
@d_spiegel
Apologies for errors in our latest Observer article. A demonstration of the care needed with stats…
-
David Spiegelhalter
@d_spiegel
this is an excellent and very valuable project by my colleagues
Alice Lawrence
@acethecurious
📢Research participants needed📢
Give us feedback🗣️on a website💻
which helps people make decisions💭❔
about #breastcancer treatments
& get a £10 M&S or Amazon voucher!
We especially welcome those under-represented in academic research!
Learn how to register below👇
Pls RT! -
David Spiegelhalter reposted
Adele Groyer
@AdeleGroyer
On behalf of @ActuaryByDay here is a summary of the ONS deaths data to week ending 13 August.
1,270 more deaths were recorded compared to the 2015-19 year average. That’s 14% higher.
Year-to-date there have been 364,655 deaths recorded; 8% higher than the 2015-19 average. -
David Spiegelhalter
@d_spiegel
Art of Statistics now out in US paperback!
www.basicbooks.com/titles/david-spiegelhalter/the-art-of-… -
David Spiegelhalter
@d_spiegel
*Personal Tweet* Extraordinary feat by part of my family climbing 3 Peaks (Snowdon, Scafell Pike, Ben Nevis) over 6 days last week in bad weather, with 3- and 6-year old grandchildren walking all the way. In aid of a fine charity working in Cameroon - uk.virginmoneygiving.com/fundraiser-display/showROFundrai…
-
David Spiegelhalter
@d_spiegel
I sound much better talking about excess deaths when it is not in English
Behrang Tajdin
@Behrang
بررسیهای بیبیسی فارسی نشان میدهد همهگیری #کرونا تا الان احتمالاً جان دویست هزار نفر را در ایران گرفته است؛ دو برابر تعداد رسمی قربانیان کووید-۱۹ در کشور
با سپاس از @d_spiegel و @MotallebiSaharThis media couldn't be displayed. See it on the original post. -
David Spiegelhalter
@d_spiegel
something on the need to consider a wide range of metrics when considering alternative policies
Royal Statistical Society
@RoyalStatSoc
RSS president @srichardson962, with fellow co-chair of our Covid-19 Task Force, @d_spiegel has written an article on how ideas from decision theory could lead to better, more rational decision-making when it comes to the pandemic
rss.org.uk/news-publication/news-publications/2021/genera… -
David Spiegelhalter
@d_spiegel
this is a fine development to help disrupt the anachronistic scientific publishing model - inspired funding by @ResEngland , and great leadership from my colleague @alex_freeman
Octopus.ac
@octopus_ac
Octopus is thrilled to have been awarded funding by @ResEngland to move to full launch!
-
David Spiegelhalter
@d_spiegel
good to see this. Masses of stats from Test and Trace, but they don't help us understand the effect of spending all those resources. This has been a problem from the very start.
Office for Statistics Regulation
@StatsRegulation
We are calling for government statistics to provide more insight into the Test and Trace programme. There are several important questions which cannot be answered by the statistics and we ask that @DHSCgovuk prioritises filling these gaps.
osr.statisticsauthority.gov.uk/correspondence/ed-humphers… -
David Spiegelhalter
@d_spiegel
non-White ethnicities have higher life expectancy than White - in stark contrast to their increased Covid risk
-
David Spiegelhalter
@d_spiegel
Here's a chance to say what you think about epidemic models! From a good team.
HPRU EZI
@HPRUezi
We want to know about your opinions on the use of transmission modelling in informing public health policy in the UK before and during the COVID-19 pandemic!
Please follow the link to complete our survey. forms.cloud.microsoft/Pages/ResponsePage.aspx?id=G96VzPWX…
@RunningMadProf @SallyBSheard @OutbreakJake -
David Spiegelhalter
@d_spiegel
Today only! The Art of Statistics only 99p on Kindle. Sounds a reasonable deal to me.
-
David Spiegelhalter
@d_spiegel
and we hope the bath analogy is useful!
-
David Spiegelhalter
@d_spiegel
Dominic Raab acknowledges JBC decision on France was mainly based on data from Reunion www.bbc.co.uk/sounds/play/m000y6rq (2:16) This should have been made clear at the time.
David Spiegelhalter
@d_spiegel
Good to see this letter from OSR to JBC. Unclear what data was used to support decision on France: data for mainland France (page 7) seem to show falling rates of beta variant www.santepubliquefrance.fr/dossiers/coronavirus-covid-19/…
Office for Statistics Regulation
@StatsRegulation
We have written to the Joint Biosecurity Centre about data to support decisions on the international travel traffic light system. It is our expectation that statistics and data which support decisions should be made available in an accessible form. osr.statisticsauthority.gov.uk/correspondence/mary-gregor…
-
David Spiegelhalter
@d_spiegel
Good to see this letter from OSR to JBC. Unclear what data was used to support decision on France: data for mainland France (page 7) seem to show falling rates of beta variant www.santepubliquefrance.fr/dossiers/coronavirus-covid-19/…
Office for Statistics Regulation
@StatsRegulation
We have written to the Joint Biosecurity Centre about data to support decisions on the international travel traffic light system. It is our expectation that statistics and data which support decisions should be made available in an accessible form. osr.statisticsauthority.gov.uk/correspondence/mary-gregor…
-
David Spiegelhalter
@d_spiegel
Shouldn’t this be: “scientists should rise above polarised policy debates”? Odd to just blame media.
theconversation.com/covid-media-must-rise-above-pitting-s… -
David Spiegelhalter
@d_spiegel
sorry, even after retirement you’ll be wondering if you will be found out
-
David Spiegelhalter reposted
Royal Statistical Society
@RoyalStatSoc
Yesterday may have been “freedom day” but now is not the time to play fast and loose with statistics. As the public are being asked to rely more on their own judgement it is vital that the government’s use of statistics helps inform that. We have some recommendations (a thread)
-
David Spiegelhalter
@d_spiegel
did anyone explain to him that a disease like Covid, that multiplies everyone’s risk of death by a similar amount, will leave average age of death unchanged but still be very nasty?www.bbc.co.uk/news/uk-politics-57854811
-
David Spiegelhalter reposted
Tom Chivers
@TomChivers
In which I argue that we should we should use our AstraZeneca stockpile to let older children get vaccinated, and people to mix-and-match vaccines at shorter intervals, if they (and the children’s parents) decide the benefits outweigh the risks unherd.com/2021/07/give-astrazeneca-to-the-kids/
-
David Spiegelhalter reposted
Justin Sandefur
@JustinSandefur
How many people have died in India during the pandemic?
There's no truly satisfactory answer. @abhishekecon @arvindsubraman and I have a new paper presenting 3 alternative estimates from 3 different sources, and discussing flaws of each.
cgdev.org/sites/default/files/three-new-estimates-indias-…
1/ -
David Spiegelhalter
@d_spiegel
Usual excellent visualisation by @binocularity , including some comparisons between projections from May and what actually happened
Nick Holliman
@binocularity
Quick visual summary of just some of @anthonybmasters & @d_spiegel 's points in here, sensitivity to variations in assumptions of vaccine efficacy is also a key issue, as are broader uncertainties.
app.powerbi.com/view?r=eyJrIjoiMGE5NzU0N2MtN2JiZi00MGMzLT… -
David Spiegelhalter
@d_spiegel
And we also look at the importance of ‘framing’: 94% vs 97% vaccine effectiveness against hospitalisation does not sound much of a difference - but it means twice as many going to hospital.
-
David Spiegelhalter
@d_spiegel
And a fine collaborator!
-
David Spiegelhalter
@d_spiegel
Isn’t there an essential mismatch in policy now? Opening up by relaxing measures, and yet (predictably) causing massive disruption to economy and families by telling millions each week to self-isolate. www.theguardian.com/world/2021/jul/16/summer-chaos-predic… And to what benefit?
David Spiegelhalter
@d_spiegel
T+T has masses of stats on its activity, but we still can’t assess what it achieves. How many contacts are infected? (last year, 6% of contacts became cases www.nature.com/articles/s41586-021-03606-z) What is cost-effectiveness for cases prevented? Overlap between people called and pinged?
David Spiegelhalter
@d_spiegel
Crucially, is 10 -days self-isolation necessary? LSHTM models
www.lshtm.ac.uk/newsevents/news/2021/modelling-study-esti…
say repeat testing almost as effective as quarantine. Trials comparing alternative strategies could have been done and published months ago. Summary: looks a bit of a mess. -
David Spiegelhalter
@d_spiegel
watching RSS Excellence in Stats awards. www.youtube.com/watch?v=Zpsyy5X8_UA So much great work done over the pandemic. Excellent winner of Campion award for Official Statistics: ONS ethnicity analysis was complex and important, and well-communicated.
-
David Spiegelhalter
@d_spiegel
This mention is richly deserved!
Pouria Hadjibagheri MBE
@Pouriaaa
The state of the UK’s statistical system 2020/21 by @StatsRegulation
Thank you! 🎊😀🎉
See the report:
osr.statisticsauthority.gov.uk/publication/the-state-of-t… -
David Spiegelhalter
@d_spiegel
If this event actually takes place in person, I suppose I shall have to change out of shorts mjauk.org/2021/07/14/mja-honours-davids-stats-tastic-work/
-
David Spiegelhalter
@d_spiegel
never done one in my life (although invented some! www.jstor.org/stable/2335498 )
JJ
@iiiinahoj
True story
-
David Spiegelhalter
@d_spiegel
a welcome statement on transparency and trustworthy use of statistics: "Where statistics and data are used publicly to inform Parliament, the media and the public, they are published in an accessible form with appropriate explanations of context and sources."
Office for Statistics Regulation
@StatsRegulation
Today we have published a statement on data transparency and the role of Heads of Profession for Statistics osr.statisticsauthority.gov.uk/news/osr-statement-on-data…
-
David Spiegelhalter reposted
Alexandra Freeman
@alex_freeman
And launching Predict:Breast Cancer in Portuguese! Brazil is one of the biggest users of the site and it's great to be able to provide it now in a form that can be shared directly with patients.
Huge thanks to @paulpharoah and the translators -
David Spiegelhalter
@d_spiegel
this is excellent (and explains why I follow @JamesWard73 so carefully)
James Ward
@JamesWard73
in case you're wondering how the hospital admissions in my model compare to those from Warwick, Imperial and LSHTM released today, here's a chart with all of them. (it shows both my central case, and the downside scenario with a larger behaviour change at "Step 4") 1/2
-
David Spiegelhalter
@d_spiegel
valuable opinion from one of leaders of the team evaluating and advising on the Covid App
www.turing.ac.uk/blog/demonstrating-impact-nhs-covid-19-app
Christophe Fraser Group
@ChristoPhraser
The vaccine is making a huge difference. Personal opinion, I think there is a good case for many people who are contact traced or pinged, incl. double vaxxed, to be offered daily testing with lateral flow instead of isolation. But let's see the data on the ongoing trials. 9/
-
David Spiegelhalter reposted
Nick Holliman
@binocularity
Broader context than usual, but related to "Living with the Virus", @anthonybmasters and @d_spiegel latest weekly stats article.
What are the wider costs, and what is tolerable?
@MSPowerBI here: app.powerbi.com/view?r=eyJrIjoiZGNhZGYzZGItYWZmZi00NGNmLW… -
David Spiegelhalter reposted
James Ward
@JamesWard73
So it’s time for me to come off the fence re 19th July. I’m in favour of:
a) Opening up on 19/7, but with stronger mitigating baseline measures than the govt is proposing
b) Opening up vaccinations to 12-17s as supply permits, starting with the vulnerable.
Long🧵follows... -
David Spiegelhalter
@d_spiegel
Oh dear, Sky headline gets the denominator wrong: it’s 0.005% of infected children, not of Covid deaths. Is this the most common statistical error?!
-
David Spiegelhalter
@d_spiegel
Latest PHE data: www.gov.uk/government/publications/investigation-of-novel… Out of 257 Delta deaths, 118 (46%) fully vaxxed, 92 (36%) unvaxxed. What we would expect with high coverage by very effective – but not perfect – vaccine.
David Spiegelhalter
@d_spiegel
Many more fully-vaxxed older people dying (118) than unvaxxed younger people (21) due to very strong risk-gradient with age: unvaxxed 30 year-old has lower risk than vaxxed 60-year-old. 71 deaths in unvaxxed over-50s - communities with lower uptake will be hit hard in new wave
-
David Spiegelhalter
@d_spiegel
Our App allows you to interrogate E+W ONS death registration data, by period, area, place-of-death, Covid/Total/Excess deaths, registrations/occurrences.
wintoncentre.maths.cam.ac.uk/coronavirus/covid-excess/ Warning: may be slow downloading. Thanks for @Henry_Jon for coding
David Spiegelhalter
@d_spiegel
Place-of-death reveals continued shift to deaths at home. Tables of all counts provided, and can look at individual areas. NB caveat on calculation of excess occurrences by area and place.
-
David Spiegelhalter
@d_spiegel
excellent opportunity to see the staggeringly linear relationship between log-death rates and age! Meaning risks increase exponentially between 5 and 95, doubling for each 5-6 years older
Pouria Hadjibagheri MBE
@Pouriaaa
We have just released a number of new features on the #ukcovdi19 Dashboard (🧵)
1. We now have an option to view graphs in logarithmic scale - I know some of you really wanted this! Feedback are welcome.
Many thanks to @statsgeekclare and @DaveJephson for the QA.
1/🧵 -
David Spiegelhalter reposted
Consilium Scientific
@consiliumsci
Book your tickets for an in - person talk by Professor David Spiegelhalter on July 27 in London! www.eventbrite.co.uk/e/naughty-numbers-in-the-news-ticket… @d_spiegel @david_colquhoun @HealthWatchUK @ImproveHealthR @LSEHealthPolicy
-
David Spiegelhalter reposted
International Science Council
@ISC
(🌟 +New speakers!)
Join @ISC-@UNDRR official side-event at the High-Level Political Forum #hlpf on 7 July to discuss:
✔ Lessons learned from the pandemic
✔ How scientists can better reach policy-makers
✔ How to make better evidence-based decisions
council.science/events/policymaking-emergency-covid-hlpf/ -
David Spiegelhalter
@d_spiegel
delighted to be part of this!
King's Maths School
@kingsmathschool
🥁 Pls retweet! More outreach
For teachers, students, maths educators + anyone interested in incisive ideas.💡
Talks: @BiggestTriangle @honeypisquared @d_spiegel @JohnMOxford @SophieBays @DrBarons @TFrancome + Dave Hewitt at @lborouniversity
Info ➡️ www.kingsmathsschool.com/mathemagora-2021 -
David Spiegelhalter
@d_spiegel
Yes, perhaps a bit harsh. But I’ve had worse referee’s reports.
-
David Spiegelhalter
@d_spiegel
this seems good advice
Sense about Science
@senseaboutsci
Talking about Covid conspiracy: this is a short guide made with help from people who have been inclined to them or debated them. Created from the research of Peter Knight @OfficialUoM @UKRI_News and team, with us. Download it here and share: senseaboutscience.org/activities/talking-about-conspiracies/
-
David Spiegelhalter
@d_spiegel
Now we are seeing a return to non-Covid health stories, it’s time for press officers, journalists and anyone else to use RealRisk!
realrisk.wintoncentre.uk (with links to our videos) -
David Spiegelhalter reposted
TMiP
@TMiPUK
Tickets are still available for the Talking Maths in Public conference in August - £25 for four half-days of online sessions from great speakers and a chance to meet the maths-comm community: talkingmathsinpublic.uk/programme
-
David Spiegelhalter
@d_spiegel
Really excellent summary of our article - great to see the Tolerability of Risk visualisation
Nick Holliman
@binocularity
A quick visual summary of this article by @anthonybmasters & @d_spiegel
What is tolerable (by need or choice) may not also be acceptable to everyone. Always a balance of risk.
@MSPowerBI version here:
app.powerbi.com/view?r=eyJrIjoiYTY5Njk1ODktNDRlNi00MjI3LW… -
David Spiegelhalter
@d_spiegel
Would getting Covid deaths down to the level of flu be ‘acceptable’?
-
David Spiegelhalter reposted
Prof. Christina Pagel - @chrischirp.bsky.social
@chrischirp
THREAD on England short term Covid outlook:
Thread on some of the implications of latest SAGE SPI-M models about English roadmap, step 4 & Delta.
TLDR: medium term v uncertain but in short term expect hospital admissions to keep rising. -
David Spiegelhalter
@d_spiegel
just listened to this. I must say it’s rather good, with excellent guests who really know their stuff. Of course I would say that, wouldn’t I - but it’s true.
Risky Talk
@RiskyTalkPod
How should we communicate evidence at the highest levels of government?
📈 No. 10’s Chief Data Analyst @statsArePower
🧪former chief scientific adviser to Home Office @HardingProf
💷 Enlightened Economist @DianeCoyle1859
join @d_spiegel to explore
riskytalk.libsyn.com/communicating-evidence-to-policymakers -
David Spiegelhalter
@d_spiegel
“The pandemic and its responses were an overall life-saver for young people."
-
David Spiegelhalter
@d_spiegel
not sure about the last line, but definitely a great team to work with!
Alexandra Freeman
@alex_freeman
Calling web developers who want to work on important/interesting stuff (and don't mind not being paid a fortune)!
Come and spend some time with us to make important sites to help patients make difficult healthcare choices.
Working with @d_spiegel a bonus!
www.jobs.cam.ac.uk/job/28635/ -
David Spiegelhalter
@d_spiegel
looking forward to appearing with a real artist
New Scientist
@newscientist
Don't Miss: Data art with David Spiegelhalter and Stefanie Posavec www.newscientist.com/article/2280243-dont-miss-data-art-w…
-
David Spiegelhalter reposted
Nick Holliman
@binocularity
One day quotable, the next less so...
In light of which a modified version of this visual, the linked version will update automatically after a little time.
app.powerbi.com/view?r=eyJrIjoiYTdkZjE4ZjYtZTMxNC00NTBjLT… -
David Spiegelhalter
@d_spiegel
yesterday’s 350 word statistical primer!
-
David Spiegelhalter
@d_spiegel
usual excellent visualisation from @binocularity !
Nick Holliman
@binocularity
A quick visual summary of data on this week's article by @anthonybmasters & @d_spiegel
The outlook is uncertain, although we have survived a variant once already (B.1.1.7). The data on the effect of variants is analysed as fast as it (reliably) arrives.
app.powerbi.com/view?r=eyJrIjoiMDc5YWYxMDQtZTk3Ny00OTkyLW… -
David Spiegelhalter
@d_spiegel
.. and that the great majority of 7 million contacts told to isolate were not infected
-
David Spiegelhalter reposted
Office for Statistics Regulation
@StatsRegulation
Today we have written to producers of health-related statistics in England to acknowledge their hard work in producing detailed statistics about COVID-19 vaccinations and outline further improvements we think are important osr.statisticsauthority.gov.uk/correspondence/ed-humphers…
-
David Spiegelhalter
@d_spiegel
Very honoured to be speaking about the stats of Covid in India (and Nepal) tomorrow at 5pm, as a new Patron of the Karuna Trust karuna.org/india-in-crisis Limited places available
-
David Spiegelhalter reposted
John Kerr
@SciComGuy
Navigating the 'minefield' of vaccine communication. Great to see @alex_freeman and our Winton Centre work noted in this @Nature piece.
nature
@Nature
The available COVID vaccines are safe and effective, but they come with risks that researchers need to communicate transparently. go.nature.com/3fdjwal
-
David Spiegelhalter reposted
Stian Westlake
@stianwestlake
Given the emerging evidence of the harm that Covid-19 can cause in pregnancy, it's time to prioritise pregnant women for vaccinations.
Great piece by @RoyalStatSoc President @srichardson962 www.thetimes.co.uk/article/pregnant-women-need-to-be-at-f… -
David Spiegelhalter
@d_spiegel
This looks excellent. And if you are looking for a book that embodies these principles, I happen to have one already prepared… www.penguin.co.uk/books/294/294857/learning-from-data/978…
Sander Greenland
@Lester_Domes
Monday 24th May at 4 pm UK time:
"Advancing statistics reform: Counting and causation before probability and inference".
Zoom link is below the abstract here:
www.bristol.ac.uk/integrative-epidemiology/seminars/2021/…
It is supposed to be recorded and posted here for convenient viewing:
www.bristol.ac.uk/integrative-epidemiology/outputs/talkvi… -
David Spiegelhalter reposted
Alexandra Freeman
@alex_freeman
Calling programmers who want to work on something interesting and life-changing for patients!
We're making our risk comms tools like breast.predict.nhs.uk (used 30,000 times per month) into an Open Source generic tool.
Come and help us!
Pls RT
www.jobs.cam.ac.uk/job/28635/ -
David Spiegelhalter reposted
Royal Statistical Society
@RoyalStatSoc
One of the four tests that the government set for proceeding with the roadmap out of lockdown is that “our assessment of the risks is not fundamentally changed by new Variants of Concern” (A thread)
-
David Spiegelhalter
@d_spiegel
great Covid statistical musical video from Italy! www.youtube.com/watch?v=glueeURebDU (A limited genre, to be true, but surely one that can only grow)
-
David Spiegelhalter reposted
Nick Holliman
@binocularity
No graphs.... doesn't the Guardian/Observer have visualization staff....😉
This week's quick visual summary of @anthonybmasters and @d_spiegel insightful writing.
app.powerbi.com/view?r=eyJrIjoiOTdlOTliZWEtZTNmNy00MWMzLT… -
David Spiegelhalter reposted
Gabriel Recchia
@mesotronium
Thanks to @ConversationUK for the opportunity to talk about this research in greater detail: theconversation.com/who-was-better-at-predicting-the-cour… @alex_freeman @d_spiegel
-
David Spiegelhalter
@d_spiegel
Delighted to be doing this with a @anthonybmasters Just the teensiest issue of finishing writing it….
The Bookseller
@thebookseller
.@penguinpress has signed Covid by Numbers, a "vital" data-driven examination of the coronavirus pandemic by leading statisticians @d_spiegel and @anthonybmasters. Discover more here: www.thebookseller.com/news/penguin-press-signs-vital-book…
-
David Spiegelhalter reposted
Simon Singh
@SLSingh
This year we ran weekly online Masterclasses with 500 students, to increase the number & diversity of great mathematicians. TEACHERS can now nominate students for 2021/22. Free places for disadvantaged students, big subsidies for the rest. Nominate here talent-ed.uk/mmt-teachers
-
David Spiegelhalter
@d_spiegel
Very low number of deaths reported by ONS today. My guess: they are the lowest age-adjusted rates *ever seen* in the UK for the end of April. But I shall wait to see what the excellent actuaries say!
www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa… -
David Spiegelhalter
@d_spiegel
Basically, Covid takes whatever risk you’ve got, and multiplies it - but a common relative risk has a much bigger impact on those who are already vulnerable. This virus is a bully.
Krishnan Bhaskaran
@krishnan_lshtm
New paper out in @LancetRH_Europe, asking: to what extent is a person's risk of dying from COVID-19 governed by their broader risk of death? @OpenSafely @ehr_lshtm @EBMDataLab @TPP_SystmOne @d_spiegel 👇
-
David Spiegelhalter
@d_spiegel
usual excellent visualisation of our article!
www.theguardian.com/theobserver/commentisfree/2021/may/09…
Nick Holliman
@binocularity
This topic needs more (good) publicity: evidenced prediction is delivering useful understanding in many areas of our lives.
Here is a visual description of @anthonybmasters @d_spiegel article today. Plus @JuniperConsort1 results on the summer ahead.
app.powerbi.com/view?r=eyJrIjoiZWI3MWY1OGEtZWY1NS00OTNhLT… -
David Spiegelhalter
@d_spiegel
This has just started - "You got one! Now give one!” seems a good idea. You can give with Gift Aid get1give1worldwide.org
-
David Spiegelhalter
@d_spiegel
Graphics to provide background to today’s announcement about AZ vaccine
Alexandra Freeman
@alex_freeman
Here are newly updated graphics from us for the latest data from the UK's @MHRAgovuk on blood clots and the AZ vaccine.
Full info is at wintoncentre.maths.cam.ac.uk/news/latest-data-mhra-blood-…
There are important notes on the site, which I'll also thread here. -
David Spiegelhalter
@d_spiegel
a great article on misinformation by Tim Harford www.ft.com/content/5cea69f0-7d44-424e-a121-78a21564ca35 My take: as well as focussing on whether to trust others, maybe we should also question whether to trust ourselves
-
David Spiegelhalter
@d_spiegel
a fine read
Tim Harford
@TimHarford
Incidentally, if you like the combination of curious stories and solid advice for making sense of the headlines, you'll love "How To Make The World Add Up" - published in paperback TODAY. Pick up a copy on your way to the polling station; it might help: timharford.com/books/worldaddup/
-
David Spiegelhalter reposted
Gabriel Recchia
@mesotronium
Our paper on how well experts and laypeople forecast the size of the COVID-19 pandemic in the UK (back in April) is out: journals.plos.org/plosone/article?id=10.1371/journal.pone… @d_spiegel @alex_freeman Short answer: Not very, though experts were much better than nonexperts. Both groups underestimated it 1/8
-
David Spiegelhalter
@d_spiegel
brilliant article by Tom Chivers unherd.com/2021/05/have-we-seen-the-end-of-covid/?tl_inbo… here’s the relevant Bill of Mortality (including 10 deaths from the King’s Evil)
-
David Spiegelhalter
@d_spiegel
I am honoured to be chairing this event this afternoon, celebrating the life and work of the great Harvey Goldstein
Royal Statistical Society
@RoyalStatSoc
We're holding an online event in May (jointly with @BritishAcademy_ ) celebrating the life and work of much-missed #rssfellow Harvey Goldstein. Speakers include @d_spiegel @Klharron Tim Cole, William Browne and George Leckie. Register at
rss.org.uk/training-events/events/events-2021/a-celebrati… -
David Spiegelhalter
@d_spiegel
excellent summary. Note the apparent protective effect of vaccination starting 21 days *before* the jab! Amazing! But, of course, you can’t get jabbed if you have had, or suspect you have had, a recent infection. Such is reverse causation.
Nick Holliman
@binocularity
A quick visual summary of this @anthonybmasters & @d_spiegel article: www.theguardian.com/theobserver/commentisfree/2021/may/02…
Will COVID-19 vaccines reduce virus transmission?
(Spoiler: Yes!)
app.powerbi.com/view?r=eyJrIjoiYzVmYzE5ZmItOWM1Yy00M2ExLW…
Use large screen (or zoom in) for details: -
David Spiegelhalter reposted
Tim Harford
@TimHarford
Excellent conversation with two of my favourite thinkers @d_spiegel @DianeCoyle1859 and other smart folk - about evidence communication in government.
I recommend the "Risky Talk" podcast.
Risky Talk
@RiskyTalkPod
How should we communicate evidence at the highest levels of government?
📈 No. 10’s Chief Data Analyst @statsArePower
🧪former chief scientific adviser to Home Office @HardingProf
💷 Enlightened Economist @DianeCoyle1859
join @d_spiegel to explore
riskytalk.libsyn.com/communicating-evidence-to-policymakers -
David Spiegelhalter
@d_spiegel
These excellent guests have got trough-loads of experience and wisdom - strongly recommend a listen
Risky Talk
@RiskyTalkPod
How should we communicate evidence at the highest levels of government?
📈 No. 10’s Chief Data Analyst @statsArePower
🧪former chief scientific adviser to Home Office @HardingProf
💷 Enlightened Economist @DianeCoyle1859
join @d_spiegel to explore
riskytalk.libsyn.com/communicating-evidence-to-policymakers -
David Spiegelhalter
@d_spiegel
Great news, especially as the Rev Thomas is buried just down the road www.ft.com/content/0dd561d4-f841-41b8-ac63-529d48ddd1d2 (and maybe his theorem could become a bit less ‘obscure’ :-) )
-
David Spiegelhalter reposted
Alexandra Freeman
@alex_freeman
Our work from last year on how to communicate personalised risks from COVID-19 is out:
royalsocietypublishing.org/doi/10.1098/rsos.201721
In it we describe a series of studies by the Winton team
@d_spiegel @SciComGuy @mesotronium @CRSchneider3 @sarahdryhurst @acethecurious @lfinikarides @LuoniGiulia -
David Spiegelhalter
@d_spiegel
‘Obscure’?!! Some of us have spent much of our working lives exploring Bayes theorem. (Great article - I suspect this is the sub-editor’s rather than Tom’s label)
Sean Mulvany
@seanmulvany
Longform from @TomChivers in @guardianscience today. How to make sense of the accuracy and reliability of a diagnostic or a diagnosis, from #covid to cancer. Answer: with Bayesian thinking. @d_spiegel www.theguardian.com/world/2021/apr/18/obscure-maths-bayes…
-
David Spiegelhalter reposted
John Kerr
@SciComGuy
Another study from our new pre-print: Do reminders to continue protective behaviours (masks 😷, distancing 🧍↔️🧍 etc.) after vaccination increase intention to do so? & might they have perverse effects on vaccination intentions?
Neither it seems!
www.medrxiv.org/content/10.1101/2021.03.19.21253963v1.ful… 1/n -
David Spiegelhalter reposted
Alexandra Freeman
@alex_freeman
What happens when you tell people the quality of the evidence underlying a health claim?
In short: it matters!
In fact, unless stated otherwise people seem to behave as if quality were high – which raises qs about not stating when quality is low.
connect.medrxiv.org/qr/2021.04.07.21255010. -
David Spiegelhalter reposted
Alexandra Freeman
@alex_freeman
It's the last few days to apply to come and work with us! If you've enjoyed our evidence communication around COVID and vaccines, come and join our team. We aim for diversity of opinions, skills and backgrounds so if you share our aims, apply!
www.jobs.cam.ac.uk/job/28736/ -
David Spiegelhalter
@d_spiegel
Sky News report that "one third of local authority areas reported a higher than average infection rate”. Wow. news.sky.com/story/covid-19-the-areas-where-infections-ar… [also there is more recent data available digital.nhs.uk/dashboards/progression-full-width]
-
David Spiegelhalter
@d_spiegel
And of course we did *not* approve the current sub-head
David Spiegelhalter
@d_spiegel
Our 350-word attempt to look at the AZ clot risks from multiple perspectives: different comparators, personal vs societal. Crucial last point from ‘tolerability of risk’ framework: nothing is ‘safe’; some risks are ‘acceptable’, some ‘unacceptable’, in-between area is negotiable
-
David Spiegelhalter
@d_spiegel
Our 350-word attempt to look at the AZ clot risks from multiple perspectives: different comparators, personal vs societal. Crucial last point from ‘tolerability of risk’ framework: nothing is ‘safe’; some risks are ‘acceptable’, some ‘unacceptable’, in-between area is negotiable
-
David Spiegelhalter reposted
Graham Easton
@Drgrahameaston
Sir David Spiegelhalter offers tips on communicating about risks and statistics during Covid . Analogies, relative risks, blood clots and sensitivity!
Open access link: media.qmplus.qmul.ac.uk/media/The+ClinComm+Podcast+Episod… -
David Spiegelhalter reposted
Simon Calder
@SimonCalder
For one night only, the analytical supergroup that is @TimHarford and @d_spiegel took to the virtual stage to talk about why numbers deserve to be treated with respect and precision; this one also has pictures, but not too much is lost if you simply listen
www.youtube.com/watch?v=8yTeHCeXVkA&t=17s -
David Spiegelhalter
@d_spiegel
I have good memories of an enjoyable and lively conversation from rather a long time ago.
Not sure of a suitable caption... -
David Spiegelhalter
@d_spiegel
Did an enjoyable interview for Premio Cosmos 2021 - included an explanation of the AZ vaccine clot slides used at the recent briefing www.youtube.com/watch?v=ow5FECigH00&t=119s
-
David Spiegelhalter
@d_spiegel
Some context for “1 in 100,000” - the risk of a blood-clot per *week* on contraceptive pill. (5 per 10,000 per year www.thelancet.com/journals/lancet/article/PIIS0140-6736(2…
) But suggested vaccine side-effect looks more serious -
David Spiegelhalter
@d_spiegel
background for the slides used in the briefing yesterday. Brilliant work by @alex_freeman and @HardingProf, and great presentation by JVT wintoncentre.maths.cam.ac.uk/news/communicating-potential…
-
David Spiegelhalter
@d_spiegel
nice work from @alex_freeman and @HardingProf
Alexandra Freeman
@alex_freeman
Our graphics to illustrate the potential benefits and potential harms of the AstraZeneca vaccine as announced by the MHRA this afternoon.
Benefits accrued over 16 weeks, at three different levels of exposure to the virus. -
David Spiegelhalter
@d_spiegel
usual brilliant interpretation by @binocularity!
Nick Holliman
@binocularity
Complex topic this week by @anthonybmasters & @d_spiegel beautifully summarised:
"Of the deaths that we are not seeing, many are the shadows of those who were taken early"
My own father was killed by #covid in London in April 2020
Interactive @MSPowerBI
app.powerbi.com/view?r=eyJrIjoiMzJlNjUwM2ItZDY4ZS00YTcyLT… -
David Spiegelhalter
@d_spiegel
350 words by @anthonybmasters and me on the noteworthy current deficit in deaths. A good thing, but why?www.theguardian.com/theobserver/commentisfree/2021/apr/04…
-
David Spiegelhalter
@d_spiegel
No medical intervention is 'safe', and balance of benefit to risk is crucial. MRHA has identified 30 serious thrombotic events with low platelets out of 18 million AZ vaccinations so, if causal, average risk of less than 1 in 500,000 is low.
David Spiegelhalter
@d_spiegel
But important to know whether this risk is substantially higher in younger age groups, as has been found elsewhere. Vaccinating younger people is as much for wider community as for their direct benefit, and so reasonable to try to avoid causing harm, even if extremely rarely
David Spiegelhalter
@d_spiegel
Vital that the vaccination roll-out is not delayed. As Sir Patrick Vallance showed last week, even at current low levels of the virus, a month's delay in vaccinating 500K people between 44-54 expected to lead to around 85 cases requiring hospitalisation, of which perhaps 5 die
-
David Spiegelhalter
@d_spiegel
can I echo that appalling smugness at being nominated for "Dr Katharine Giles Award for best popular article written by a scientist or engineer" from ABSW! Awed to be among these people
Tom Chivers
@TomChivers
Smug as all heck to be shortlisted for science journalist of the year by the Association of British Science Writers! www.absw.org.uk/pages/115-absw-awards-2021-shortlists congratulations Some great candidates in there
-
David Spiegelhalter reposted
Ramai Santhirapala
@ramai23
Read @alex_freeman @Tammy_Hoffmann #chrisdelmer and my evaluation of the course in @bmj_company #ebm:
ebm.bmj.com/content/early/2020/09/21/bmjebm-2020-111521 -
David Spiegelhalter reposted
Robert Cuffe
@robertcuffe
Get your positive rapid tests double-checked
-
David Spiegelhalter
@d_spiegel
Guidance for any statisticians willing to engage with the media on COVID (and anything else), by @kjm2 and me. rss.onlinelibrary.wiley.com/doi/10.1111/1740-9713.01508 Basically, try and sound human, and be prepared for it all to go wrong
-
David Spiegelhalter
@d_spiegel
Great guests discussing then value of stories based on data, and how to make them accessible to audiences
Risky Talk
@RiskyTalkPod
Journalism is about compelling stories - but statistics require nuance and caveat. How do you do both?
@caelainnbarr, @TomChivers and Paul Milner join @d_spiegel to talk statistical style, vivid data vis and good storytelling:
riskytalk.libsyn.com/doing-journalism-with-data -
David Spiegelhalter
@d_spiegel
usual brilliant portrayal of our brief article - with some great quotes and links
Nick Holliman
@binocularity
The more science @anthonybmasters and @d_spiegel summarize the more they illustrate the huge efforts to find robust scientific ways forward in the pandemic.
Interactive #powerbi link
app.powerbi.com/view?r=eyJrIjoiNDIxOGZiNDktOTVlYS00NGJiLW… -
David Spiegelhalter
@d_spiegel
our usual caveat: we are only given 350 words!
-
David Spiegelhalter reposted
The BMJ
@bmj_latest
We are able to estimate risk of death from covid-19 based on identified risk factors. But what do these numbers mean for an individual patient, and how should they modify their activities and interactions in light of them? www.bmj.com/content/372/bmj.n637
@sfinnikin @d_spiegel -
David Spiegelhalter
@d_spiegel
Personal tweet. This is the latest campaign from Karuna, the excellent charity to whom I donate my speaking fees. karuna.org/breaking-free video too www.youtube.com/watch?v=vFWVsrURjw8 I strongly recommend them.
-
David Spiegelhalter reposted
Tim Harford
@TimHarford
On 27 March I'm talking to @d_spiegel on the topic of "What Do All The Numbers Mean"? Hosted by @Cambridge_Fest
It's free. Sign up and drop in!
www.festival.cam.ac.uk/events/how-make-world-add -
David Spiegelhalter
@d_spiegel
this is excellent. And Tim is very generous in his acknowledgements! timharford.com/2021/03/cautionary-tales-catching-a-killer…
-
David Spiegelhalter reposted
Tim Harford
@TimHarford
Tune in to "How To Vaccinate The World" to hear Susan Ellenberg, @sheencr and @d_spiegel discuss what we've learned so far from the data about vaccines.
11.30am GMT, @BBCRadio4 and then on BBC Sounds: www.bbc.co.uk/programmes/m000py6s -
David Spiegelhalter
@d_spiegel
With the usual caveat that we only have 350 words for a complex topic!
-
David Spiegelhalter
@d_spiegel
Excellent guide for communicating stats about the pandemic
gss.civilservice.gov.uk/policy-store/tips-for-communicati…
@stattymichelle @GSSgoodpractice @AnalysisFuncti1 -
David Spiegelhalter
@d_spiegel
Back-of-envelope. If 5 events *were* caused by AZ vaccine, expect 1 event in 2 mil jabs. If 2 mil people in 50s put off their jab by 1 week, expect ~2,000 to get infected, 20-30 seriously ill, 5 to die. So benefit/risk ratio high even for *1* week. But risk goes on in unjabbed!
-
David Spiegelhalter reposted
Kai Kupferschmidt
@kakape
Why did European countries pause immunizations with AstraZeneca vaccine? What have doctors actually seen?
@GretchenVogel1 and I try to sift through the evidence so far.
Story here, thread to come: www.sciencemag.org/news/2021/03/it-s-very-special-picture… -
David Spiegelhalter reposted
Office for Statistics Regulation
@StatsRegulation
Thinking of changing career? We are looking people who can apply strong numerical and analytical skills to support our regulatory work.
💰 £36,027
Statistics Regulator: www.civilservicejobs.service.gov.uk/csr/index.cgi?SID=c2V…
Statistics Regulator (Economy) www.civilservicejobs.service.gov.uk/csr/index.cgi?SID=am9… -
David Spiegelhalter
@d_spiegel
reports of 6 cases of cerebral venous sinus thromboses (CVST) in younger women Germany following 1.6 million AZ vaccines www.pei.de/SharedDocs/Downloads/EN/newsroom-en/hp-news/fa… UK reports 3 cases after 10 mil AZs, 1 after 10 mil Pfizer www.gov.uk/government/publications/coronavirus-covid-19-v… analysis-print EMA will look at age profile
-
David Spiegelhalter
@d_spiegel
with cases rising, pausing a vaccine rollout may not be the appropriate 'cautious’ choice
Anthony Cox
@Cox_A_R
I've written a short piece about the precautionary principle and the AstraZeneca Vaccine issues theconversation.com/blood-clot-fears-how-misapplication-o…
-
David Spiegelhalter
@d_spiegel
very good FAQ sheet from the German Paul-Ehrlich Institute about the suspected adverse events www.pei.de/SharedDocs/Downloads/EN/newsroom-en/hp-news/fa…
-
David Spiegelhalter
@d_spiegel
Important that EMA look at possible link of *all* Covid vaccines with low platelets, not just AZ. pharmaphorum.com/news/eu-experts-call-for-allergic-reacti…
www.health.com/condition/infectious-diseases/coronavirus/… -
David Spiegelhalter
@d_spiegel
cool animation from ONS showing, for people of each year of age, how the proportion with antibodies has changed over time
www.ons.gov.uk/peoplepopulationandcommunity/healthandsoci… -
David Spiegelhalter
@d_spiegel
my take on why we need to be very careful when interpreting stories about the AZ vaccine and blood clots www.theguardian.com/commentisfree/2021/mar/15/evidence-ox…
-
David Spiegelhalter
@d_spiegel
Our attempt at summarising Bayes theorem in words: "If you pick what looks like a needle from a haystack, then your confidence that it really is a needle depends on how many are in the haystack in the first place."
-
David Spiegelhalter
@d_spiegel
This week’s effort by me and @anthonybmasters to explain something statistical in 350 words: why we need to be cautious about lateral flow tests in schools www.theguardian.com/theobserver/commentisfree/2021/mar/14…
-
David Spiegelhalter
@d_spiegel
I had forgotten about this article about Fukushima from 10 years ago. It could have had a more coherent story, but did have some good details and I hope has stood the test of time fairly well
-
David Spiegelhalter
@d_spiegel
Excellent memories of John Haigh from Rob. John was always amazingly helpful when I wanted to check a probability claim about games or gambling - a kind and generous colleague.
Rob Eastaway
@robeastaway
I'm sad to report that John Haigh, the statistician and probability guru, died on Tuesday. I've written some memories of him here:
robeastaway.com/blog/john-haigh
(cc @royalstatsoc @d_spiegel) -
David Spiegelhalter
@d_spiegel
this is good stuff
Royal Statistical Society
@RoyalStatSoc
📢Today we launch our memo to government on the ten statistical lessons they must learn from Covid-19
rss.org.uk/news-publication/news-publications/2021/genera… (a thread) -
David Spiegelhalter
@d_spiegel
really pleased to be part of this and celebrate Harvey’s work. And also that he was such a fine man
Royal Statistical Society
@RoyalStatSoc
We're holding an online event in May (jointly with @BritishAcademy_ ) celebrating the life and work of much-missed #rssfellow Harvey Goldstein. Speakers include @d_spiegel @Klharron Tim Cole, William Browne and George Leckie. Register at
rss.org.uk/training-events/events/events-2021/a-celebrati… -
David Spiegelhalter
@d_spiegel
The usual masterful display ! Amazed you can do these so fast - it takes us ages to do the 350 words!
Nick Holliman
@binocularity
Another great summary for Sunday morning reading from @anthonybmasters + @d_spiegel
If you like to click links follow this for a #powerbi version:
app.powerbi.com/view?r=eyJrIjoiOGE5NTc4ZGYtZDg2OS00NDM1LT… -
David Spiegelhalter
@d_spiegel
Our latest piece, on variants. No apologies for celebrating the UK’s sequencing effort. Usual caveat: we only have 350 words, and that’s challenging
-
David Spiegelhalter
@d_spiegel
Feeling a bit smug having the coolest mask in town #census2021
-
David Spiegelhalter reposted
Nick Triggle
@NickTriggle
The great school reopening. How #covid testing could do more harm than good, why we maybe missing a trick with masks and why it may not matter in infection rates rise www.bbc.co.uk/news/health-56285179
-
David Spiegelhalter reposted
Andrew Croxford
@andrew_croxford
Wasn’t happy with the negative press this excellent vaccine received at the start of the year. It wasn’t warranted, obviously overblown and now seems to be neck and neck with the Pfizer/BioNTech vaccine in terms of efficacy.
www.economist.com/graphic-detail/2021/03/06/new-data-show…
Andrew Croxford
@andrew_croxford
I implore everyone, including my mum, who have had the AstraZeneca vaccine to ignore the headlines and be confident that the vaccine you have received is HIGHLY LIKELY to keep you out of hospital.
There is no 100% in biology.
If you want 100%, talk to a chemist. -
David Spiegelhalter reposted
Jon Deeks FMedSci
@deeksj
@RoyalStatSoc has put out a statement on School Testing stressing
the need for TRANSPARENCY
the risks of INACCURATE results
the need for GOOD SCIENCE
sharing EXISTING DATA
rss.org.uk/RSS/media/File-library/News/2021/RSS-statement… -
David Spiegelhalter
@d_spiegel
latest data from PHE suggest we may soon be showing a deficit (negative excess) of daily all-cause deaths www.gov.uk/government/statistics/national-flu-and-covid-1… Good news, but needs cautious interpretation in the face of lots of risk and uncertainty
-
David Spiegelhalter reposted
Cliff Pickover
@pickover
"Stained Glass Brain"
By Prof. David Spiegelhalter. Source: www.psychol.cam.ac.uk/BCNI/publicinfo/stainedglass/view -
David Spiegelhalter
@d_spiegel
excellent musical explanation of R0 and herd immunity! (In Italian) www.youtube.com/watch?v=JvNsllGq9MM
-
David Spiegelhalter
@d_spiegel
This report identifies issues that go well beyond the exam saga. Strongly recommended! (COI: I was on the Expert Oversight Group)
Office for Statistics Regulation
@StatsRegulation
Our UK-wide review of the approach for awarding exam results last summer identifies the wider lessons for government and other public bodies. We are concerned that public confidence in the use of statistical models has been damaged. osr.statisticsauthority.gov.uk/news/osr-publish-review-of…
-
David Spiegelhalter
@d_spiegel
I’ve usually added 5-10% to the dashboard ‘deaths within 28 days of +ve test’. But since late Jan, dashboard undercount has increased. ??? A greater proportion being ICU patients who survived more than 28 days? If so can expect divergence to increase www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa… Tab 9
-
David Spiegelhalter
@d_spiegel
latest data from ONS. Big deficit in deaths not involving Covid. Will we get to 'no excess deaths' by Easter? Looks likely www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa…
-
David Spiegelhalter
@d_spiegel
excellent coincidence
Richard Harwood KC
@RichardHarwood2
Feathers flying in Manchester Crown Court today
-
David Spiegelhalter
@d_spiegel
For deaths and admissions, big gap opening up now between over and under 65s. For over 75s, deaths each day have been essentially halving every *week*. Extraordinary.
-
David Spiegelhalter reposted
Jim McManus OCDS,FFPH Views my own
@jimmcmanusph
The Biologist , the @RoyalSocBio magazine, has a great feature on this year’s best science podcasts. For anyone interested in communicating risk David Spiegelhalter’s series is a must riskytalk.libsyn.com #publichealth #covid19
-
David Spiegelhalter
@d_spiegel
so please don’t tell us off for insufficient detail :-)
-
David Spiegelhalter
@d_spiegel
good to see that weekly bulletins from NHS England now have lots more detail on vaccination rollout. And even some denominators www.england.nhs.uk/statistics/statistical-work-areas/covi…
-
David Spiegelhalter
@d_spiegel
latest data from dashboard: new cases seem to be falling faster in older age-groups, but bumpiness leads to wide intervals
-
David Spiegelhalter
@d_spiegel
Declines in admissions and deaths are steeper in older than younger groups. 40% weekly reductions in deaths is very encouraging, to put it mildly
-
David Spiegelhalter
@d_spiegel
deaths in over 85s falling faster than 65-74s, as might be expected from vaccine. But why are deaths in under 65s apparently falling so fast? An artefact of long-stay cases dying after 28 days and so not being counted using this measure? Need to wait for registrations.
-
David Spiegelhalter
@d_spiegel
usual amazingly fast production of summary from @binocularity !!
Nick Holliman
@binocularity
-
David Spiegelhalter
@d_spiegel
Our latest offering - also see latest evidence that, after the
second dose pf AZ, "efficacy was higher in those with a longer prime-boost interval"
www.thelancet.com/journals/lancet/article/PIIS0140-6736(2… -
David Spiegelhalter
@d_spiegel
good analysis refuting misinformation about deaths in 2020 anthonybmasters.medium.com/there-were-over-600-000-death-…
-
David Spiegelhalter
@d_spiegel
I strongly recommend this programme on having the humility to admit what we don't know. @sarahdryhurst does a fine job in explaining our research that suggests this can be done without losing trust
Dr. Sarah Dryhurst
@sarahdryhurst
Can we show more humility in acknowledging what we don't know about #covid19, #gdp and the #economy? Can we be more muscular in communicating #uncertainty? Fantastic narrative by Michael Blastland and @CarolineBayley that I was honoured to be a part of! www.bbc.co.uk/programmes/m000s81m
-
David Spiegelhalter
@d_spiegel
this is a great team to work with. Well I would say that, wouldn’t I? But I mean it
Alexandra Freeman
@alex_freeman
We're looking for a software developer to join the Winton Centre team, helping develop our lovely risk communication tools like breast.predict.nhs.uk. Clojure/clojurescript skills required, ideally. Closing date 31st March.
www.jobs.cam.ac.uk/job/28635/ -
David Spiegelhalter
@d_spiegel
Good to see formal criteria for new shielded group: absolute risk of catching+death in 1st wave > 0.5% OR relative risk > 10 digital.nhs.uk/coronavirus/risk-assessment/population But what about people on *current* shielded list who do *not* meet these criteria?
-
David Spiegelhalter
@d_spiegel
Official government report says “The impact of contact tracing is relatively small” (p22) www.gov.uk/government/publications/the-r-m-model-technica… - surprised this hasn't had more attention
David Spiegelhalter
@d_spiegel
They estimate R reduction of 18-33% *compared to a scenario with only social distancing restrictions and no self-isolation* - but nearly all due to isolation: contact tracing alone reduced the R number by only 2-5%. So could funds for centralised contact-tracing be better used?
-
David Spiegelhalter
@d_spiegel
good article, which repeatedly highlights the lack of publicly available data on all these vital issues www.bbc.co.uk/news/56031531
-
David Spiegelhalter
@d_spiegel
Deaths in over-75s have been falling even faster than deaths in under-75s - roughly halving every 15 days.
-
David Spiegelhalter
@d_spiegel
JBC’s modelling of TTI estimates R reduction of 18-33% *compared to a scenario with only social distancing restrictions and no self-isolation*. Contact tracing alone reduced R by 2-5% www.gov.uk/government/publications/the-r-m-model-technica… ie contact tracing added very little over people with symptoms isolating
-
David Spiegelhalter
@d_spiegel
It’s a bit terse, but a lot to get into our 350-word ration!
-
David Spiegelhalter
@d_spiegel
this is so much better than clapping
Dr Dom Pimenta MD
@DrDomPimenta
It’s Valentine’s Day today.
Many NHS staff will be working this weekend & so many are beyond exhausted as it is.
So why not #sharethelove & send a little the NHS way by donating to @theHWF? Funding counselling, wellbeing & rest rooms for staff.
Donate | www.gofundme.com/f/nhs-valentines-day -
David Spiegelhalter
@d_spiegel
In the 10 days up to feb 6th, deaths were falling at twice the rate in over-85s compared to under-65s. This is good news. [sorry, just put up yesterday’s graph!]
-
David Spiegelhalter
@d_spiegel
thread on the stats of side effects of breast cancer therapies - very important but challenging area
Alexandra Freeman
@alex_freeman
Patients said they wanted more info about the possible side effects of their breast cancer therapies, but there wasn’t good evidence so we started doing meta-analyses. The first is now out, on bisphosphonates. We found 24 possible side effects...
doi.org/10.1371/journal.pone.0246441
🧵 -
David Spiegelhalter
@d_spiegel
Well let’s see how this goes down in Germany, France, Austria, Poland, Sweden, Denmark, the Netherlands and Spain, who are all currently denying their older citizens this protection
Chris Smyth
@Smyth_Chris
Significant move from @WHO - they are following Britain in recommending the Oxford AZ vaccine be given 8-12 weeks apart, saying a longer gap "associated with greater vaccine efficacy" and will reach more people.
And recommend for over 65s too
www.who.int/publications/i/item/WHO-2019-nCoV-vaccines-SA… -
David Spiegelhalter
@d_spiegel
Stats regulator calls for better public data on percentage vaccine take up by age band, ethnicity and priority group. And to stop publicly quoting data that are not in the public domain. But acknowledges intensive work taking place to provide more comprehensive vaccinations data
Office for Statistics Regulation
@StatsRegulation
During the pandemic, Government statisticians and analysts across the UK have worked intensively to provide a comprehensive picture of the roll-out of vaccinations. We are now calling for further progress.
osr.statisticsauthority.gov.uk/news/statement-osr-calls-f… -
David Spiegelhalter
@d_spiegel
Deaths of over-85s falling faster than for under-65s (P=0.05, 2-sided). Seems a bit soon to be due to vaccines alone, but have not done modelling. log-plot as visually easier to compare straight lines @BristOliver (Hope I’ve got the code right this time)
-
David Spiegelhalter
@d_spiegel
ben, great news!
Ben Goldacre
@bengoldacre
I’ve been asked by Secretary of State for Health and Social Care to lead a new review into how we can get better, broader, safer use of health data in the UK. We will be hearing about barriers and great practice from across the ecosystem. More tomorrow!
www.gov.uk/government/news/new-review-into-use-of-health-… -
David Spiegelhalter
@d_spiegel
Call for proposals for a randomised trial of COVID-19 vaccine dosing intervals www.nihr.ac.uk/funding/randomised-trial-to-evaluate-vacci… - this will help the world
-
David Spiegelhalter
@d_spiegel
Our review of adverse effects of bisphosphonates in
breast cancer journals.plos.org/plosone/article?id=10.1371/journal.pone… extraordinary analysis from Chris Jackson (@cjackstats) revealing new side-effects and not detecting others that had been reported -
David Spiegelhalter reposted
Adam Finn
@adamhfinn
Please take a look at this piece if you are interested in how vaccines really work and how we measure those effects in populations and decide how to deploy them. www.theguardian.com/commentisfree/2021/feb/08/compare-vac…
-
David Spiegelhalter
@d_spiegel
“embarrassing tale” about procurement of failed test www.dailymail.co.uk/news/article-9229507/amp/Taxpayers-fo… - all predicted by @deeksj back in August
www.theguardian.com/world/2020/aug/27/data-secrecy-covid-… -
David Spiegelhalter
@d_spiegel
Yesterday’s effort by me and @anthonybmasters to explain the over-65s and Oxford/AZ vaccine issue in 350 words. www.theguardian.com/theobserver/commentisfree/2021/feb/06… What should we try next week?
-
David Spiegelhalter
@d_spiegel
many apologies for wasting people’s time on previous tweet this morning - over 85s are *not* yet showing faster decrease in admissions. Thanks to @alexselby1770 for pointing out my error. Moral, don’t rush coding on Sunday mornings
-
David Spiegelhalter
@d_spiegel
I think I made an error in my last tweet about vaccines - deleted and checking analysis!
-
David Spiegelhalter reposted
Sense about Science
@senseaboutsci
Test and Trace model shows 90% of its claimed effectiveness is down to people isolating themselves in response to symptoms, not the intervention of Test and Trace. www.huffingtonpost.co.uk/entry/test-and-trace-dodgy-stati…
-
David Spiegelhalter
@d_spiegel
How effective is the Oxford coronavirus vaccine for the over-65s? This week’s attempt by @anthonybmasters and me to explain statistical things in 350 words www.theguardian.com/theobserver/commentisfree/2021/feb/06…
-
David Spiegelhalter
@d_spiegel
President of the Royal Statistical Society argues for more embedded experiments to properly evaluate different strategies: proper trials have been essential for treatments and vaccines theconversation.com/vaccine-rollouts-school-testing-and-c…
-
David Spiegelhalter
@d_spiegel
Great interview with Fiona Fox from the wonderful @SMC_London thebiologist.rsb.org.uk/biologist-covid-19/189-biologist/… (if you didn’t hear it, Nick Robinson ended up asking my opinion on Dominic Cummings, leading to some laughter)
-
David Spiegelhalter
@d_spiegel
This is a very useful explanation of all those different ways of counting Covid deaths, including full R code for graphics!
-
David Spiegelhalter reposted
Tim Johns
@timoncheese
If you find the news a bit glum lately do this dance when BBC News comes on to inspire joy
-
David Spiegelhalter
@d_spiegel
this is useful - suggests that you should only listen to the daily reported deaths on Thursdays, Fridays and Saturdays
Iain Dillingham
@iaindillingham
You can see the cyclical pattern more clearly if we separate the data by day of the week. In other words, Sunday's chart shows only the data for Sundays; Monday's chart shows only the data for Mondays; etc.
-
David Spiegelhalter
@d_spiegel
I know it's what everyone says on Twitter, but I was really true-life delighted to be asked to be part of this course: thanks to James and his fine team for putting it together
Office for Statistics Regulation
@StatsRegulation
There was a time when it would be inconceivable to imagine Government leaders addressing the nation using charts and statistics, which has now become normal. Our guest blog from @KuhtJames looks at the importance of government leaders being data literate osr.statisticsauthority.gov.uk/blog/upskilling-government…
-
David Spiegelhalter
@d_spiegel
sorry, linked to wrong paper! papers.ssrn.com/sol3/papers.cfm?abstract_id=3777268
David Spiegelhalter
@d_spiegel
Vital data from Oxford www.medrxiv.org/content/10.1101/2020.12.27.20248896v2.ful… After standard SD/SD doses, 74 (vaccine) vs 197 (control) symptomatic cases (63% efficacy); and 8 vs 45 (82% efficacy) after 12 wk gap 😀😀 But no apparent effect on asymptomatic infections (41 vs 42 in UK study) 😟
-
David Spiegelhalter
@d_spiegel
Nice of LSE to feature @kjm2 and my blog on communicating through the media in the time of Covid blogs.lse.ac.uk/impactofsocialsciences/2021/02/03/communi…
-
David Spiegelhalter
@d_spiegel
Vital data from Oxford www.medrxiv.org/content/10.1101/2020.12.27.20248896v2.ful… After standard SD/SD doses, 74 (vaccine) vs 197 (control) symptomatic cases (63% efficacy); and 8 vs 45 (82% efficacy) after 12 wk gap 😀😀 But no apparent effect on asymptomatic infections (41 vs 42 in UK study) 😟
David Spiegelhalter
@d_spiegel
50% reduction in any PCR+ after 2 doses. But far fewer of these infections will be symptomatic, so may transmission reduction be considerably larger than 50%? [Warning, only a statistician]
-
David Spiegelhalter
@d_spiegel
robust blog on need for better data from NHS and PHE on vaccine rollout
Simon Briscoe
@simonbriscoe
We rightly celebrate the vaccine roll out but why soundbites not data on vaccination rates in care homes and by other demographics from @NHSEngland & @PHE_uk?
Refusing to publish data or incapable of publishing data? simonbriscoeblog.wordpress.com/2021/02/01/invisible-covid… Scotland is OK - see
@P_H_S_Official -
David Spiegelhalter reposted
Kevin McConway
@kjm2
People have been kindly saying that our "Top tips to statisticians", d-spiegel.medium.com/top-tips-to-statisticians-communicat…, makes points that are useful to anyone communicating evidence. But @d_spiegel and I couldn't possibly comment because of our point 2, "Don’t be lured out of your comfort zone"...
-
David Spiegelhalter reposted
Royal Statistical Society
@RoyalStatSoc
Co-chair of our Covid-19 Task Force @d_spiegel and RSS stats ambassador @anthonybmasters in the @ObserverUK this weekend on how the Covid vaccine is prioritised
www.theguardian.com/theobserver/commentisfree/2021/jan/31… -
David Spiegelhalter
@d_spiegel
glad you think this is useful - it was fun putting it together
-
David Spiegelhalter
@d_spiegel
Nick, thanks so much for this! Amazing to put this together so quickly - great communication of main points, and uncertainties @anthonybmasters
Nick Holliman
@binocularity
@sarahdryhurst wondering if we would make this (even) clearer, using our work from KTN workshop?
I am not sure how easy it is to comprehend text only including numbers. I seem to remember some eye tracking work on numbers in sentences slowed reading? -
David Spiegelhalter reposted
Sonia Sodha
@soniasodha
Cynical interventions in the vaccine debate risk undermining public confidence. What Labour, Macron and the BMA got wrong last week.
www.theguardian.com/commentisfree/2021/jan/31/politiician… -
David Spiegelhalter
@d_spiegel
Substantial excess deaths this winter. But fewer than might expect from Covid deaths, as non-Covid-caused deaths are still at low levels
-
David Spiegelhalter
@d_spiegel
Behind the numbers: how the recipients of the Covid vaccine are prioritised in UK: with @anthonybmasters www.theguardian.com/theobserver/commentisfree/2021/jan/31… [only allowed 350 words, so rather terse]
-
David Spiegelhalter
@d_spiegel
Useful guide: COVID-19 Vaccination Statistics by Anthony B. Masters link.medium.com/uWGO00dktdb
-
David Spiegelhalter
@d_spiegel
NIHR case study on our work with @chrischirp on public engagement with website to explain complex and sensitive children’s heart surgery data. Proud of this. www.nihr.ac.uk/case-studies/putting-public-engagement-at-…
-
David Spiegelhalter
@d_spiegel
COVID-19 (largely) multiplies existing frailties. Delighted to have helped a little with this fine work.
Krishnan Bhaskaran
@krishnan_lshtm
NEW @OpenSAFELY analysis (PREPRINT - NOT PEER REVIEWED)
We carried out parallel analyses of COVID-19 deaths and non-COVID deaths, to answer the question: to what extent is a person's risk of dying from COVID-19 governed by their broader risk of death?
www.medrxiv.org/content/10.1101/2021.01.15.21249756v2 -
David Spiegelhalter reposted
EXPeditions
@joinExpeditions
“We are always making choices about what data we collect, what we do to it, and how we present those conclusions”. Share @D_Spiegel’s Passion for Statistics on EXP-Editions. #dataprivacyday #datacollectionday @Cambridge_Uni
share.exp-editions.com/en_US/exp/5fb290a83f678 -
David Spiegelhalter reposted
POLITICOEurope
@POLITICOEurope
OPINION: 100,000 coronavirus deaths is an appalling number. Yet it is still a substantial underestimate of the real number of coronavirus fatalities.
Statistician David Spiegelhalter explains what data we should follow for a more accurate picture.
www.politico.eu/article/coronavirus-deaths-statistics-dat… -
David Spiegelhalter reposted
Royal Statistical Society
@RoyalStatSoc
What is R? What can we tell from international comparisons? What's a 90% effective vaccine? Our Covid-19 Task Force has answered some of the most common questions around the stats. Take a look on our new Covid-19 FAQ page 👉rss.org.uk/policy-campaigns/policy/covid-19-task-force/co…
-
David Spiegelhalter
@d_spiegel
brilliant work
Ben Goldacre
@bengoldacre
The NHS in England has rapidly delivered mass vaccination. Here we share an OpenSAFELY analysis that gives insights on exactly which patients received the COVID-19 vaccine in the earliest parts of the vaccination campaign. www.medrxiv.org/content/10.1101/2021.01.25.21250356v1.ful…
-
David Spiegelhalter reposted
John Burn-Murdoch
@jburnmurdoch
NEW: white people in England aged 80+ are being vaccinated at twice the rate of black people, and rates in deprived neighbourhoods are lagging behind less deprived areas www.ft.com/content/a831fcdd-5793-3374-b281-80c4afe06640?a
The gaps are widening, suggesting fundamental challenges in ensuring equal protection -
David Spiegelhalter
@d_spiegel
yes, a wonderful counterfactual worthy of Robert Harris (if he hadn’t already done Bletchley)
Duncan Bell
@DrDuncanBell
In this episode, David Spiegelhalter observes that if F. P. Ramsey had lived only a year longer - he died at 26 - he would have been the supervisor/Director of Studies of Alan Turning, an incoming undergraduate at King's. Worthy of a novel.
-
David Spiegelhalter
@d_spiegel
this is very good. I am so impressed by the way the JCVI has made bold, evidence-based decisions on prioritisation and vaccine strategy, based on a strong ethical framework
Adam Finn
@adamhfinn
JCVI normally does its work in relative obscurity. Currently it has been thrust into the public eye amidst public interest in COVID19 vaccines & related policy. This article provides a clear summary of the way it works & the principles driving its advice www.sciencedirect.com/science/article/pii/S26667762203002…
-
David Spiegelhalter
@d_spiegel
COVID deaths within 28 days of +ve test may reach 100,000 today. But ONS data www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa… show that over 100,000 people in the UK had died with COVID on death certificate by *Jan 7th*, nearly 3 weeks ago. 108,000 by Jan 15th, and total now will be nearly 120,000
David Spiegelhalter
@d_spiegel
Around 90% of these had COVID as immediate cause of death, and so perhaps we can say that around 100,000 people in the UK have now died *because of* COVID. An awful total
-
David Spiegelhalter reposted
Risky Talk
@RiskyTalkPod
And here’s the video from @d_spiegel himself (the original bacon-sandwich-muncher-number-cruncher)
Risky Talk
@RiskyTalkPod
🗣️ Relative risks often drive misleading headlines about health. Here's how to recognise them, and how to use absolute risks for clearer communication from @d_spiegel
youtu.be/ELNt0XZYsAQ -
David Spiegelhalter
@d_spiegel
JCVI’s argument for vaccine delay strategy, with PHE’s statistical analysis
www.gov.uk/government/publications/prioritising-the-first… Also monitoring strategy here assets.publishing.service.gov.uk/government/uploads/syste… But really need randomised trial -
David Spiegelhalter
@d_spiegel
pleased to hear @adamhfinn on @r4Today essentially telling the BMA where to stick their letter criticising vaccine strategy
-
David Spiegelhalter
@d_spiegel
Very sad about this news: I learned much of my maths in that pub www.bbc.co.uk/news/uk-england-oxfordshire-55763746
-
David Spiegelhalter
@d_spiegel
thanks for the video plug!
Dorothy Bishop (inactive account)
@deevybee
Well, I got sufficiently irritated by @c4news reporting yesterday to write a short blogpost on why reporters should stop converting absolute risks to relative risks. @Anja_Popp @krishgm @d_spiegel @alex_freeman
deevybee.blogspot.com/2021/01/time-to-ditch-relative-risk… -
David Spiegelhalter
@d_spiegel
a striking visualisation of daily changes in positive cases by area priceless-swirles-45b099.netlify.app/index.html - (would make an intriguing tapestry project)
-
David Spiegelhalter reposted
Risky Talk
@RiskyTalkPod
🎲 Odds Ratios are odd, but they crop up often in medical & social research thanks to some convenient statistical properties. @d_spiegel takes you through what they mean and how translate them into more accessible terms when communicating. 👍 youtu.be/ixKhS0Silb4
-
David Spiegelhalter
@d_spiegel
genius
BBC More or Less
@BBCMoreOrLess
By popular demand, here's our sea shanty on some fishy figures...
-
David Spiegelhalter reposted
Office for Statistics Regulation
@StatsRegulation
Today we have written to producers of COVID-19 vaccine statistics across the UK. The publication of vaccine statistics is a welcome step forward in supporting transparency and we look forward to seeing the data develop in order to keep the public informed. osr.statisticsauthority.gov.uk/correspondence/ed-humphers…
-
David Spiegelhalter
@d_spiegel
Awful misleading BBC News headline y'day “highest number of deaths recorded in a single day since the pandemic began” www.bbc.co.uk/iplayer/episode/m000rhjf/bbc-news-at-six-19… implying 1,610 died in last 24 hrs. But Tuesday’s *reported* deaths always high: truth probably nearer to 1,000 (which is high enough) 1/2
David Spiegelhalter
@d_spiegel
1,073 died on 8th April, using criterion of 28 days after +ve test coronavirus.data.gov.uk/details/deaths But few tests in April – in fact *1,456* died on Apr 8th and registered with Covid www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa… So note to BBC: current Covid death rate much less than 1st wave peak 2/2
-
David Spiegelhalter
@d_spiegel
glad you like it!
Ed Sykes
@edmsykes
I'm late to this but here's an absolutely fantastic tool for anyone wanting to communicate risk from their research. Hats off to everyone at @d_spiegel @ilangoodman @mc_climent and all involved.
-
David Spiegelhalter
@d_spiegel
Bank holidays complicate today’s ONS E+W data www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa… Since Sept, > 26,000 excess deaths over 5-year avg, but 31,000 have Covid as primary cause. So non-Covid deaths are low: distancing -> minimal flu, and many among the 59,000 Mar-Jun excess. Graph by @Henry_Jon
-
David Spiegelhalter
@d_spiegel
my (inevitably doomed) attempt to explain everything about excess deaths in 330 words. www.theguardian.com/commentisfree/2021/jan/17/statistics-… First of a series
-
David Spiegelhalter reposted
Global Forum for Media Development #MediaDev
@MediaGFMD
📆 Event: Join @EJNetwork on January 27 for a panel discussion with @jamesrbuk, @anjahuja, @wiczipedia, @kateomega, and @d_spiegel on the role journalists can play in tackling COVID-19 #misinformation.
Details here: ethicaljournalismnetwork.org/giving-it-our-best-shot-what…
#mediadev -
David Spiegelhalter reposted
John Burn-Murdoch
@jburnmurdoch
NEW video from me & @tomhannen addressing more myths that have been used to play down Covid this winter.
Featuring:
• Why delays in death registration (& bad charts) wrongly led people to think there was minimal excess mortality
• "Excess winter deaths" are not "excess deaths"
Financial Times
@FinancialTimes
The FT's @jburnmurdoch explains why the UK's Covid-19 death rate is likely to rise further. Watch the full video: www.ft.com/video/0cd6f9f9-664e-40f9-bad4-dde59d7c746c?pla…
This media couldn't be displayed. See it on the original post. -
David Spiegelhalter reposted
Hannah Ritchie
@_HannahRitchie
Important slide of COVID demographics in England.
What impact does vaccinating over-75s or over-65s have?
→ vaccinate small % of population
→ target most deaths
→ large amount of hosp. admissions but less ICU
→ small % of cases
From @chrischirp at @IndependentSage briefing -
David Spiegelhalter reposted
Anna Powell-Smith
@darkgreener
The UK is currently publishing less data on its otherwise impressive vaccine rollout than other countries.
I've done some international comparative analysis for @CFPData.
Thread
(1/9) -
David Spiegelhalter
@d_spiegel
Extraordinary data from Scotland on excess deaths by cause and location in 2020 www.nrscotland.gov.uk/files/statistics/covid19/covid-deat… 6,686 deaths involving COVID-19 closely match 6,704 excess over 5-year average. But 570 *fewer* deaths than normal in hospitals, even with over 3,000 Covid deaths! ?????
David Spiegelhalter
@d_spiegel
Deaths at home normally around 12,000, but *17,000* in 2020 (>40% extra). Extra 1,800 cancer deaths at home, 1,000 from heart disease and strokes. So massive change in place of death in Scotland – very similar to England and Wales. But what has the end-of-life care been like?
-
David Spiegelhalter
@d_spiegel
excellent article on ‘scotch-egging’ capx.co/stop-the-scotch-egging-and-focus-on-the-big-picture/
-
David Spiegelhalter
@d_spiegel
Ivan Wise’s wonderful Better Known betterknown.co.uk podcast celebrating its 150th episode with Jim Al-Khalili’s 6 things that should be better known. My list featured whelks
betterknown.co.uk/2020/06/07/david-spiegelhalter/ -
David Spiegelhalter
@d_spiegel
Top Tip #43. Why I like Zoom interviews. Can position text just below laptop camera, and so only have to lower eyes a little to check notes.
Channel 4 News
@Channel4News
Statistician Prof David Spiegelhalter, from Cambridge University, tells @jonsnowC4 the vaccine rollout will soon have a “dramatic effect” on the Covid death toll - but only after a “massively tough time for the NHS” over the next few weeks.
www.channel4.com/news/vaccine-rollout-will-have-dramatic-… -
David Spiegelhalter reposted
COVID-19 Actuaries Response Group
@COVID19actuary
In our latest bulletin @john_actuary extends our earlier work on the impact of the vaccine on deaths, to additionally consider when we’ll see an impact on reported case numbers, hospitalisations and ICU admissions.
www.covid-arg.com/post/how-soon-to-see-the-vaccine-s-bene… -
David Spiegelhalter
@d_spiegel
So we need good public data on actual vaccinations - dashboard currently only gives a weekly update coronavirus.data.gov.uk/details/healthcare
-
David Spiegelhalter reposted
Christopher Snowdon
@cjsnowdon
Every time I tweet these stats I get the same old questions and objections in the replies, despite having blocked the worst smileys. So I’ve written out the answers to save time.
-
David Spiegelhalter reposted
RSS History Section
@HistoryofStats
As the fabulous @florencemuseum closes indefinitely after the pandemic effects, find out how you can support its reopening in the near future. 👇
We would really love the memory of our first @RoyalStatSoc female fellow to live for ever, wouldn't you?
Florence Nightingale Museum 💙
@florencemuseum
Help define & support our post-pandemic future! Trustees of the Florence Nightingale Museum role with FLORENCE NIGHTINGALE MUSEUM London (Central), London (Greater) jobs.theguardian.com/job/7150297/trustees-of-the-florence… #jobs
-
David Spiegelhalter
@d_spiegel
Just finished this Christmas beast of a jigsaw. Grey, gloomy and very difficult. But got to the end eventually. There’s a contemporary lesson there somewhere...
-
David Spiegelhalter
@d_spiegel
Personal: I strongly recommend this fine, heart-warming book (ages 2-10) www.amazon.co.uk/Nain-Rocks-Afro-Lorreta-Chindo-Besong/dp… Small girl wants long blonde hair: visits Africa, sees culture, has hair styled. Then loves being “little brown girl with an afro” COI: author is my (unofficially adopted) daughter!
-
David Spiegelhalter reposted
Theo Sanderson
@theosanderson
1/ The MHRA has approved a longer gap between doses for both the AstraZeneca vaccine and the Pfizer vaccine. The latter has concerned some people. Specifically many are citing a figure of 52% for protection after the first dose.
Here is why this 52% figure isn't useful [1/n] -
David Spiegelhalter
@d_spiegel
ONS shows total deaths 13% above 5-year average. www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa… Non-Covid-caused deaths at low level. Deficit of alzheimers and dementia - many brought forward to spring. Also, predictably, almost no flu deaths. Graph by @Henry_Jon
David Spiegelhalter
@d_spiegel
PHE's Sentinel Trusts show Covid admissions running at 400+ the rate of flu admissions www.gov.uk/government/statistics/national-flu-and-covid-1… Flu admission rates at extremely low level
-
David Spiegelhalter reposted
Dr Dom Pimenta MD
@DrDomPimenta
We have lost over 640 healthcare workers to COVID so far. They leave behind grieving partners and children.
Today @theHWF launches a Family Fund to take care of the families of those who took care of ours.
Donate | www.gofundme.com/f/help-them-help-us-christmas-appeal
Read more about the project below 👇🏽 -
David Spiegelhalter
@d_spiegel
For time of year, total and excess deaths are without recent precedent. But deaths without Covid as primary cause are at lowest level for 10 years. Minimal flu, and some brought forward by first-wave?
-
David Spiegelhalter
@d_spiegel
Also reported by Paul Gallagher inews.co.uk/news/health/uk-coronavirus-death-toll-latest-… That's people dying from Covid as primary cause
BBC Newsnight
@BBCNewsnight
“More people have been killed by Covid in the last nine months in the UK than civilians were killed in the UK during WWII.”
Statistician Prof @d_spiegel says he hopes "the spirit of Beveridge" will follow the pandemic and levelling up will be "more than a slogan"
#newsnight -
David Spiegelhalter reposted
Royal Statistical Society
@RoyalStatSoc
The Government has said that a dashboard with vaccine stats will be available next week. This is good news – but the devil will be in the detail. A thread from our Covid-19 Task Force. (1) rss.org.uk/policy-campaigns/policy/covid-19-task-force/
-
David Spiegelhalter reposted
Percival Merganser
@lennylaw
Sir David Spiegelhalter joins the queue of people telling Peter Hitchens that he’s a twat and explaining exactly why.
-
David Spiegelhalter
@d_spiegel
another example of where good collection and communication of data seems to be a low priority
Rowland Manthorpe
@rowlsmanthorpe
What's happening with the data about the vaccine? Well, let's put it this way: there's a lot to sort out
A THREAD on my reporting today -
David Spiegelhalter
@d_spiegel
wow, Our World in Data now showing both Sweden and Germany having a higher daily Covid death rate than the UK ourworldindata.org/coronavirus-data-explorer?zoomToSelect…
-
David Spiegelhalter
@d_spiegel
13,500 excess deaths over 5-year avg in E+W Sept 5 - Dec 4, unlike anything recent. But non-Covid deaths at level of ‘good year’ + minimal flu so far. Will deaths in January exceed a ‘bad year’? Graph by @Henry_Jon
www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa… -
David Spiegelhalter
@d_spiegel
This should be interesting (and with luck will include an illuminating exchange with Nick Robinson….)
Jim Al-Khalili
@jimalkhalili
If you're interested in how scientists have been coping with communicating their Covid-related research over the past year, I'm presenting a doc on @BBCRadio4 tomorrow evening at 8pm called Scientists in the Spotlight
www.bbc.co.uk/programmes/m000q9gj -
David Spiegelhalter
@d_spiegel
Talking with the excellent @MisakCheryl about my hero Frank Ramsey - this afternoon at 4.30 on Radio 4
Oxford Philosophy
@OUPPhilosophy
Join @MisakCheryl tomorrow at 16:30 GMT as she and David Spiegelhalter (@d_spiegel) discuss the astonishing life of philosopher and all-round polymath #frankramsey on @BBCRadio4's 'Great Lives' - we'll be tuning in! www.bbc.co.uk/programmes/m000q8pq
-
David Spiegelhalter
@d_spiegel
Our topical festive baubles ....
-
David Spiegelhalter
@d_spiegel
Great work from my colleagues on vaccine hesitancy around the world, and relation to trust in science and anxiety about Covid
John Kerr
@SciComGuy
Interesting results when we ask people if they would accept a (hypothetical) COVID-19 vaccine...
All across the world, men, people who trust science & people who worry about COVID more likely to say yes.
But the proportion seems to be declining...
www.medrxiv.org/content/10.1101/2020.12.09.20246439v1
1/X -
David Spiegelhalter
@d_spiegel
this is worth bookmarking
ONS Focus
@ONSfocus
Today we've launched the Coronavirus Latest Insights interactive tool which brings together the latest data on #covid19 from the ONS to make it easier to follow the progression of the pandemic. www.ons.gov.uk/peoplepopulationandcommunity/healthandsoci…
-
David Spiegelhalter
@d_spiegel
I do recommend this podcast - guests to be treasured! And rather topical
Risky Talk
@RiskyTalkPod
💉 Communicating about vaccines has never been higher stakes. How do we do it well? Joining @d_spiegel to explore the value of scientific rigour and building relationships are @ProfHeidiLarson and @ajpollard1 riskytalk.libsyn.com/communicating-about-vaccines
-
David Spiegelhalter
@d_spiegel
Looks like I’m about 13 millionth in the queue for a vaccine www.telegraph.co.uk/news/2020/12/09/priority-list-pfizer-… But why am I still in front of a younger person with Type 2 diabetes? medium.com/wintoncentre/why-i-am-in-a-higher-priority-gro…
-
David Spiegelhalter
@d_spiegel
Some videos that may be useful for communicating stats and risk concepts. Don’t ask how many takes were necessary...
Risky Talk
@RiskyTalkPod
Announcing our Winton Centre Youtube channel! 🍿
🗣️The boss @d_spiegel on how to communicate research stats to a wide audience. ORs, HRs, Relative vs Absolute risks and more.
❗️Intro to our new tool RealRisk
🎙️ Episodes of Risky Talk.
Enjoy!
www.youtube.com/channel/UCrxUG_3Cdxmn3J4QX_w-HKw?cbrd=1&u… -
David Spiegelhalter
@d_spiegel
I should emphasise this follows from their excellent 2-stage testing strategy - pooled testing did see 10 false positives (0.5%), and then corrected by individual tests www.cam.ac.uk/sites/www.cam.ac.uk/files/documents/pooled_…
David Spiegelhalter
@d_spiegel
0 positive out of 10,121 tests on asymptomatic people! Cambridge Lighthouse Lab does not seem to suffer much from false positives
Phil Rodgers
@PhilRodgers
Just nine cases in week 9 of @Cambridge_Uni's testing programme, all from tests on people with symptoms. The University also tested 10,121 people without symptoms, and found zero cases. Details at www.cam.ac.uk/coronavirus/stay-safe-cambridge-uni/data-fr…
-
David Spiegelhalter
@d_spiegel
A week today - Great Lives on the great Frank Ramsey www.bbc.co.uk/programmes/m000q8pq
-
David Spiegelhalter
@d_spiegel
0 positive out of 10,121 tests on asymptomatic people! Cambridge Lighthouse Lab does not seem to suffer much from false positives
Phil Rodgers
@PhilRodgers
Just nine cases in week 9 of @Cambridge_Uni's testing programme, all from tests on people with symptoms. The University also tested 10,121 people without symptoms, and found zero cases. Details at www.cam.ac.uk/coronavirus/stay-safe-cambridge-uni/data-fr…
-
David Spiegelhalter reposted
Alexandra Freeman
@alex_freeman
How certain do people think a negative (or positive) COVID test result is? Is that affected by the info used to explain test results?
We tested info from NHS, CDC, and New Zealand’s page about test uncertainty against people who got no information…
www.medrxiv.org/content/10.1101/2020.12.04.20243840v1 -
David Spiegelhalter
@d_spiegel
sorry, deleted and retweeted this as made mistake with number of deaths
David Spiegelhalter
@d_spiegel
Vaccine priorities unaffected if look at years of life medium.com/wintoncentre/why-i-am-in-a-higher-priority-gro… Back-of-envelope analysis, for 1,000 infections, expect 11 deaths each losing 9 years of life: around 100 years lost in total. Current 30,000 daily infections lead to around 330 deaths, 3000 years lost
-
David Spiegelhalter
@d_spiegel
Vaccine priorities unaffected if look at years of life medium.com/wintoncentre/why-i-am-in-a-higher-priority-gro… Back-of-envelope analysis, for 1,000 infections, expect 11 deaths each losing 9 years of life: around 100 years lost in total. Current 30,000 daily infections lead to around 330 deaths, 3000 years lost
-
David Spiegelhalter
@d_spiegel
Blog medium.com/wintoncentre/why-i-am-in-a-higher-priority-gro… on why I have been put in a higher vaccine priority group than younger people with health conditions. Ranked broadly by risk of dying if infected: ~1% for me. But risk dominated by age, eg a 40-year-old white man with transplant has lower risk
David Spiegelhalter
@d_spiegel
I am not saying this ranking is 'correct': just providing an explanation. I hope groups 5 and 6 all get vaccine at the same time
-
David Spiegelhalter
@d_spiegel
This is the charity that my speaking-fees go to. They are doing a great job - it’s such important work. Maybe have a look?
Karuna
@karuna_trust_uk
Education is life-saving for girls at risk of child marriage, trafficking & domestic bondage. Help us be ready for girls when schools reopen in #india & #nepal, so that no girl is left behind. Your donation this Christmas will be doubled! www.karuna.org/her-education
#givingtuesday -
David Spiegelhalter
@d_spiegel
Celia Hoyles’ useful blog on maths anxiety blogs.ucl.ac.uk/ioe/2020/12/01/maths-anxiety-how-can-we-o… - - I certainly feel physically ill when confronted with some government ‘maths’ ...
-
David Spiegelhalter
@d_spiegel
Latest @ONS data www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa… 12,535 deaths registered in E+W in week ending Nov 30th. 5-year average is 10,380, max over last 10 years was 10,882 in 2019. Unlike any recent year. Graph by @Henry_Jon
Post not available here, view on X
David Spiegelhalter
@d_spiegel
Between Sept 5th and Nov 20th, 12,907 deaths involving Covid registered in UK, roughly 3,000 since then, making 16,000 so far in the second wave. Sadly, prediction that second wave means tens of 000s of Covid deaths looks like it will be fulfilled, likely >20K by Christmas
David Spiegelhalter
@d_spiegel
Once again, there were over 1,000 additional deaths in private homes compared with normal, a 40% increase. This seems to be a long-term change in the way people are dying in this country, and deserves close attention
-
David Spiegelhalter
@d_spiegel
Wonderful!
@tombutton.bsky.social
@MathsTechnology
But then I'd have never written this - it's the one thing that I've created that I'm most proud of 😀
-
David Spiegelhalter
@d_spiegel
Data used for tiering decisions available from this page (check link at bottom) www.gov.uk/government/publications/coronavirus-cases-by-l… Excellent to see this transparency, but couldn’t some of Test and Trace’s billions be spent on a decent website/dashboard, given all our sacrifices?
-
David Spiegelhalter
@d_spiegel
This horrible graph still there www.gov.uk/government/publications/covid-19-winter-plan/c… Kevin McConway @kjm2 x.com/kjm2/status/1331197478750785548 has shown it was knocked up in Excel, and on a linear scale the UK looks much lower relative to others. Spelling errors would be corrected, but what about this misleading mess?
Kevin McConway
@kjm2
Definitely log. Codged up in Excel, with log scale:
-
David Spiegelhalter
@d_spiegel
ah here it is!
www.gov.uk/government/publications/coronavirus-cases-in-e… (Thanks to @ArtySmokesPS) I should have tried a bit harder in my searching...
David Spiegelhalter
@d_spiegel
Very detailed slide pack from PHE and NHS Test and Trace with reasons for every Tier decision
assets.publishing.service.gov.uk/government/uploads/syste… But it doesn’t seem to turn up on a Google search! Need a good website with all this info there -
David Spiegelhalter
@d_spiegel
Very detailed slide pack from PHE and NHS Test and Trace with reasons for every Tier decision
assets.publishing.service.gov.uk/government/uploads/syste… But it doesn’t seem to turn up on a Google search! Need a good website with all this info there -
David Spiegelhalter
@d_spiegel
excellent explanation!
-
David Spiegelhalter
@d_spiegel
Always a pleasure to contribute to the excellent online Plus Magazine for people interested in what maths can do plus.maths.org/content/david-spiegelhalter-wins-faraday-p…
-
David Spiegelhalter
@d_spiegel
this joint RSS - Full-Fact course on fact-checking looks excellent www.rss.org.uk/training-events/training/public-courses/pr…
-
David Spiegelhalter
@d_spiegel
our suggestions for checking for trustworthy communication www.theguardian.com/commentisfree/2020/nov/26/pandemic-ea… from @alex_freeman , Michael Blastland, @Sander_vdLinden , @MarteauTM and me
-
David Spiegelhalter
@d_spiegel
useful blog on the vaccine studies - I had no idea the criteria for a ‘case’ were so different
Prof. Jennifer Visser-Rogers
@StatsJen
Those AstraZeneca and Oxford University vaccine results were pretty puzzling and complex. I’ve gone through the protocol to try to make sense of it all and ask the questions we still don’t know answers too! phastar.com/blog/254-astrazeneca-covid19-vaccine-trial-re…
-
David Spiegelhalter
@d_spiegel
looks like a great offer!
Dr Dom Pimenta MD
@DrDomPimenta
So my book DUTY OF CARE is NOW just 99p on Kindle for a LIMITED TIME (30/11)
All royalties go to the Healthcare Workers’ Foundation, (@theHWF) helping NHS staff mentally & physically at this tough time.
So buy at least FIVE copies NOW: www.amazon.co.uk/Duty-Care-everyone-should-COVID-19-ebook…
RT & share! 🎄 -
David Spiegelhalter
@d_spiegel
ghastly graph from government’s Winter Plan
www.gov.uk/government/publications/covid-19-winter-plan/c… - is this a logarithmic or linear scale? -
David Spiegelhalter
@d_spiegel
Being questioned by PACAC tomorrow at 9am parliamentlive.tv/Event/Index/2d536fc1-a070-42d3-8c1d-d8a… Hope to shout our 5 rules for evidence communication: 1) Inform, not persuade 2) Balance but not false balance 3) Disclose uncertainties 4) State evidence quality 5) Pre-bunk misinformation www.nature.com/articles/d41586-020-03189-1
-
David Spiegelhalter
@d_spiegel
very timely, our latest Risky Talk podcast on vaccine communication - with mega-stars Andrew Pollard and @ProfHeidiLarson
riskytalk.libsyn.com/communicating-about-vaccines -
David Spiegelhalter
@d_spiegel
featuring one of my graphs!
Jemima Kelly
@jemimajoanna
A follow-up to my Notebook on our tolerance of flu deaths, looking at our perception of Covid risk for @FTAlphaville www.ft.com/content/879f2a2b-e366-47ac-b67a-8d1326d40b5e via @financialtimes
-
David Spiegelhalter
@d_spiegel
Singing and COVID-19:
This is how we’ve been doing it
www.youtube.com/watch?v=5LgkfK3scPw - also new guidance for doing it in real time! www.gov.uk/government/publications/covid-19-suggested-pri… -
David Spiegelhalter
@d_spiegel
I think this is a brilliant article by my colleague Alex Freeman - such a good analogy
Alexandra Freeman
@alex_freeman
Some musings from me on risk perception/communication on BMJ blog today.
We throw numbers around when talking about health risks - they’re a way of precisely defining a concept. But it’s like communicating colour through hexademical codes or wavelengths…
blogs.bmj.com/bmj/2020/11/18/alexandra-freeman-the-strang… -
David Spiegelhalter
@d_spiegel
just to be clear, this is referring to cases in hospital
David Spiegelhalter
@d_spiegel
Apology: I made a mistake today on Today and BBC News - compared current Covid cases in UK (16K) with April in *England* (17K), Should have said 19.5K for UK in April. But story unchanged - this is very likely to be exceeded.
-
David Spiegelhalter
@d_spiegel
Apology: I made a mistake today on Today and BBC News - compared current Covid cases in UK (16K) with April in *England* (17K), Should have said 19.5K for UK in April. But story unchanged - this is very likely to be exceeded.
-
David Spiegelhalter reposted
Alexandra Freeman
@alex_freeman
All the work by @_gfarmer and team on developing this hugely widely-used tool.
More improvements coming (such as translation into French, which we hope to release next week)
Paul Pharoah (R soul) 🇪🇺🇵🇬🇿🇦🇬🇧
@paulpharoah
The important process that underpinned the development of the web interface for the #predict models
breast.predict.nhs.uk
prostate.predict.nhs.uk
by the Winton Centre team - @alex_freeman @d_spiegel and others is out.
www.medrxiv.org/content/10.1101/2020.11.16.20232348v1 -
David Spiegelhalter
@d_spiegel
Our ‘manifesto’ for trustworthy evidence communication! See the Supplement for more details
Alexandra Freeman
@alex_freeman
We’ve got a comment in Nature today on why science needs good ‘evidence communication’ and not the typical rules of rhetoric, designed to persuade rather than inform: www.nature.com/articles/d41586-020-03189-1
@d_spiegel @Sander_vdLinden @MarteauTM -
David Spiegelhalter
@d_spiegel
For some reason this tweet got deleted, and so here it is again for those who missed this wonderful example for teaching percentages
-
David Spiegelhalter
@d_spiegel
what a great demonstration of the need for @Nat_Numeracy !! I hope every maths class is using this in their lessons today. Remember, this man has 144K Twitter followers...
-
David Spiegelhalter
@d_spiegel
excellent article by @TomChivers on incentives in pharma unherd.com/2020/11/should-big-pharma-profit-from-covid/?t…
-
David Spiegelhalter
@d_spiegel
Here’s a short video from EXPeditions featuring my dogma on trust and data, in bite-size chunks (and higher-than-usual production values) share.exp-editions.com/en_US/exp/5faedcb32ce8c
-
David Spiegelhalter reposted
The BMJ
@bmj_latest
Spending the equivalent of 77% of the NHS annual revenue budget on an unevaluated, underdesigned national programme leading to an insufficiently supported intervention—in many cases for the wrong people—cannot be defended, say @DrMikeWGill and Muir Gray www.bmj.com/content/371/bmj.m4436
-
David Spiegelhalter
@d_spiegel
Back in March I pointed out that Imperial infection fatality rate estimates closely matched average annual mortality risk. Based on 200,000+ deaths, their latest estimates show an even closer match! www.imperial.ac.uk/mrc-global-infectious-disease-analysis…
David Spiegelhalter
@d_spiegel
This is an average, but as first approximation might say that catching Covid roughly doubled an individual’s risk of dying in 2020. But IFR in second wave should be substantially lower.
-
David Spiegelhalter
@d_spiegel
(3rd time lucky!) Free registration for great day discussing uncertainty about Covid Fri 20th www.bristol.ac.uk/integrative-epidemiology/seminars/2020/… see also humble BMJ editorial www.bmj.com/content/bmj/371/bmj.m3979.full.pdf
-
David Spiegelhalter
@d_spiegel
me explaining Bayesian ideas - www.youtube.com/watch?v=06kTysxv5uw (Note to self. Sit on hands rather than wave them at the camera).
-
David Spiegelhalter
@d_spiegel
So delighted that John Aston www.statslab.cam.ac.uk/~jada2/ will become the new Harding Professor of Statistics in Public Life after he steps down as CSA at Home Office. Lots of material to work with...
-
David Spiegelhalter
@d_spiegel
The Art of Statistics is the 99p Kindle Deal of the Day, today! www.amazon.co.uk/Art-Statistics-Learning-Pelican-Books-eb… Why not get two?
-
David Spiegelhalter
@d_spiegel
these crochet representations of different R numbers are wonderful - by a professor of biostatistics in Norway www.youtube.com/watch?v=9lxKUulMxH8
-
David Spiegelhalter
@d_spiegel
I’m talking to the wonderful @TimHarford tomorrow at 6pm - being streamed live on the Royal Society Youtube channel www.youtube.com/watch?v=JW9plVfanjo
The Royal Society
@royalsociety
Join us online this Thursday for the 2020 Michael Faraday Prize Lecture, where @d_spiegel will discuss communicating statistics in the time of #covid and how to understand risk with @TimHarford. Find out more: royalsociety.org/science-events-and-lectures/2020/11/fara…
-
David Spiegelhalter reposted
Bennett Institute for Public Policy
@BennettInst
TONIGHT - Hear @AnnaVignoles @MarteauTM @d_spiegel & @lawrennd talk to @EhsanMasood about their experiences & reflections of advising government during #covid_19 & how well Westminster has used the key scientific evidence to inform its decisions?
www.eventbrite.com/e/science-evidence-and-government-refl… -
David Spiegelhalter
@d_spiegel
a clinical trial statistician’s look at the vaccine results, having to work out what the data show as it was not given in the press release!
Prof. Jennifer Visser-Rogers
@StatsJen
Days like today make me so proud to be a clinical trial statistician! Here’s a little blog piece I’ve written looking at the numbers behind COVID-19 vaccine announcement today phastar.com/blog/250-statisticians-view-on-pfizer-covid19…
-
David Spiegelhalter
@d_spiegel
wonderful example of Simpson’s Paradox! Well-spotted Kevin!
Kevin McConway
@kjm2
Media: worst ever week for Test & Trace; they only reached 59.9% of identified contacts. But the % reached went up this week for contacts managed by local health protection teams AND for contacts not managed by them (What used to be called 'complex' and 'non-complex' cases.) How?
-
David Spiegelhalter
@d_spiegel
More from OSR on trust and transparency: "Timely and transparent publication of information negates the need for information leaks – which are the antitheses of the expectations set out in the Code of Practice for Statistics"
Office for Statistics Regulation
@StatsRegulation
Following on from our statement earlier today regarding transparency of data related to COVID-19, Mary Gregory looks at the importance of transparency for public trust in government and the decisions it makes. osr.statisticsauthority.gov.uk/why-trust-and-transparency…
-
David Spiegelhalter
@d_spiegel
Vaulable statement from OSR "where models are referred to publicly, particularly to inform significant policy decisions, the model outputs, methodologies and key assumptions should be published at the same time" (COI: I am a non-exec on the UKSA board)
Office for Statistics Regulation
@StatsRegulation
Today we have released a statement on use of data by governments across the UK as part of their communications on the Covid-19 pandemic. osr.statisticsauthority.gov.uk/news/osr-statement-regardi…
-
David Spiegelhalter
@d_spiegel
Yesterday PM said things were worse than Reasonable Worst Case Scenario, and yet this RWCS has not been published, just leaked to some media, eg Spectator
www.spectator.co.uk/article/classified-covid-in-winter-20…
So much for transparency. Need fixed timetable for releasing Sage docs, not just at whim of No 10
David Spiegelhalter
@d_spiegel
Perhaps they were embarrassed at RWCS showing odd plateau through Oct/Nov - didn't consider risks from schools and universities opening? Unless unpredictable catastrophe, if actual is worse than RWCS then by definition it was not a RWCS - in fact a UWCS (Unreasonable …)
-
David Spiegelhalter
@d_spiegel
In the room yesterday at No 10 was Johnson, Whitty, Valance, and an enormous unmentioned elephant with ‘Centralised Test and Trace’ written on its belly. 6 weeks ago Sage said it was probably having ‘marginal impact’ www.gov.uk/government/publications/summary-of-the-effecti… and it’s been downhill since then
-
David Spiegelhalter
@d_spiegel
Media asking about why varying estimates of R. 8 groups all estimated UK’s median R on October 8th, from similar data sources, and got non-overlapping intervals. Suggests intervals are too tight - so best to take an individual estimate with caution www.gov.uk/government/publications/spi-m-o-consensus-stat…
-
David Spiegelhalter
@d_spiegel
Excellent new Covid RED dashboard from UCL covid.i-sense.org.uk Would be good to also know (a) how many contacts isolate (b) how many would not have done so anyway (c) how many actually infected. All necessary for T+T to be effective - need this info!
-
David Spiegelhalter
@d_spiegel
good to see this plug for Core Maths
Glaister
@p_glaister
MUCH statistical/thinking/understanding/literacy in #coremaths - designed to support & integrate this into those subjects & others too! @TimHarford @robeastaway @Bobby_Seagull @FryRsquared @d_spiegel
2 letters in The Times draw attention to this
see www.stem.org.uk/news-and-views/opinions/core-maths-most-s… too -
David Spiegelhalter
@d_spiegel
this is an excellent collaboration! (COI, I am joint chair of the RSS COVID-19 Task Force)
Royal Statistical Society
@RoyalStatSoc
We are pleased to announce today our partnership with the @turinginst to boost the government’s data analysis capabilities and support its response to Covid-19
rss.org.uk/news-publication/news-publications/2020/genera… -
David Spiegelhalter
@d_spiegel
another fine article on the importance of acknowledging uncertainty in communications ebm.bmj.com/content/early/2020/10/18/bmjebm-2020-111541.f… by @mgtmccartney, Frank Sullivan and @carlheneghan advise that "admissions of uncertainty in such communications is regarded as a strength"
-
David Spiegelhalter
@d_spiegel
excellent book!
Tim Harford
@TimHarford
HAPPY WORLD STATISTICS DAY EVERYONE!
World Statistics Day only comes every five years - like the Olympics - so it's time to express a little mindful gratitude for all the statisticians and other wonderful nerds out there helping us to understand the world. -
David Spiegelhalter
@d_spiegel
youtu.be/XmiEzi54lBI this link may be more reliable!
David Spiegelhalter
@d_spiegel
I did an ultra low-tech video tutorial on Covid false-positive rates with @jamesgrime for Mathsworld - still relevant! www.youtube.com/watch?v=XmiEzi54lBI
-
David Spiegelhalter
@d_spiegel
Is this originally Puritan rather than Quaker - Cromwell’s Law x.com/DrJohnCosgrove/status/1290565460354899968 Or did he take it from the Quakers?
Nieve Thomas
@nievethomas
There is an apposite piece of advice from the Quakers that comes to mind: “Think it possible that you may be mistaken”
-
David Spiegelhalter
@d_spiegel
I did an ultra low-tech video tutorial on Covid false-positive rates with @jamesgrime for Mathsworld - still relevant! www.youtube.com/watch?v=XmiEzi54lBI
-
David Spiegelhalter
@d_spiegel
excellent article on acknowledging uncertainty abut Covid from @mendel_random, @MarcusMunafo and Michael Blastland. www.bmj.com/content/371/bmj.m3979 The Footnote nicely demonstrates their own epistemic humility
-
David Spiegelhalter reposted
Royal Statistical Society
@RoyalStatSoc
Our Covid-19 Task Force comments on the importance of scientific scrutiny of Covid-19 tests
-
David Spiegelhalter
@d_spiegel
Getting some good feedback on #realrisk - all thanks to @ilangoodman @mc_climent Mike Pearson, Jin Park and @alex_freeman
Alexandra Freeman
@alex_freeman
Calling all science/health journalists, press officers and those trying to communicate health research:
Delighted to announce the launch of RealRisk, @d_spiegel and the Winton Centre's tool to help you get your numbers across clearly and accurately.
realrisk.wintoncentre.uk -
David Spiegelhalter
@d_spiegel
If you are interested in the future of education for mathematical sciences, please respond to this call!
The Royal Society
@royalsociety
We have released a new Call for Views to inform our understanding of the changing nature of mathematics and implications for the future of education, society and work. Read more and submit a response: #mathsfutures #mathsed #futureofed royalsociety.org/topics-policy/projects/mathematical-futu…
-
David Spiegelhalter
@d_spiegel
If you have anything to do with communicating health research, have a look at the Winton Centre’s RealRisk!
Alexandra Freeman
@alex_freeman
Calling all science/health journalists, press officers and those trying to communicate health research:
Delighted to announce the launch of RealRisk, @d_spiegel and the Winton Centre's tool to help you get your numbers across clearly and accurately.
realrisk.wintoncentre.uk -
David Spiegelhalter
@d_spiegel
“the most consistent predictor of decreased susceptibility to misinformation about Covid-19 was numerical literacy” - great study (COI, by my colleagues) royalsocietypublishing.org/doi/10.1098/rsos.201199 Data literacy is vital
www.theguardian.com/world/2020/oct/14/poor-numerical-lite… -
David Spiegelhalter
@d_spiegel
Good to have continuing excess home deaths scrutinised - see also Ed’s article news.sky.com/story/coronavirus-excess-fatalities-how-pand…
Ed Conway
@EdConwaySky
Thread: Throughout the pandemic there's been plenty of attention focused on #covid19 deaths and on what’s happened in hospitals and in care homes - for obvious and understandable reasons. But for months there’s been another phenomenon which I find deeply unsettling.
-
David Spiegelhalter
@d_spiegel
excellent article
Stephen John Senn
@stephensenn
“Forget the clash of grand ideas of Sweden versus New Zealand. Just stop bungling the basics.” Excellent piece by @TimHarford in today’s @FT weekend magazine.
-
David Spiegelhalter reposted
Isaac Newton Institute
@NewtonInstitute
In the latest #inipodcast we're joined by the esteemed @d_spiegel to discuss "number theatre" 🎭, ranting 😠, blame 👊, speculation 🤷♂️, "The Art of Statistics" ✅, "Following the Science" ❌ and more. Listen here 👉 www.newton.ac.uk/news/podcast or via your podcast app of choice 🎧
-
David Spiegelhalter
@d_spiegel
our team’s research and advice on communicating Covid risk
Alexandra Freeman
@alex_freeman
Preprint out – our work on communicating personal risks from COVID-19:
doi.org/10.1101/2020.10.05.20206961
In summary, for those developing personal risk calculators:
1) Are you trying to change behaviour or simply inform people? It changes your communication approach. UK public opinion: -
David Spiegelhalter
@d_spiegel
shouldn’t @BBCNews report this, instead of grossly misleading statements like "the UK has hit 10,000 new cases for the first time since mass testing began” www.bbc.co.uk/news/uk-politics-54407656
UK COVID-19
@UKCovid19Stats
Nearly 13K COVID-19 positives reported for the UK yesterday.
Technical issues delaying the reporting of data (now resolved) means announcements for the next few days will be inflated to clear the backlog.
This chart shows most of yesterdays reported positives by specimen date. -
David Spiegelhalter
@d_spiegel
BMI 30-35 hardly changes risk, BMI 35-40 raises to 4-5%
David Spiegelhalter
@d_spiegel
mortality rate for a 74-year-old white man, otherwise healthy, with Covid-19? from this
alama.org.uk/covid-19-medical-risk-assessment/
3-4% during the peak of pandemic in the UK, would presumably be somewhat less now. -
David Spiegelhalter
@d_spiegel
mortality rate for a 74-year-old white man, otherwise healthy, with Covid-19? from this
alama.org.uk/covid-19-medical-risk-assessment/
3-4% during the peak of pandemic in the UK, would presumably be somewhat less now. -
David Spiegelhalter
@d_spiegel
For those following The Shipman Files www.bbc.co.uk/iplayer/episodes/m000n1h4/the-shipman-files… details of all deaths are here webarchive.nationalarchives.gov.uk/20090808221518/http://… - we have extracted data on 215 victims dspiegel29.github.io/ArtofStatistics/00-1-age-and-year-of…
which are plotted using R code dspiegel29.github.io/ArtofStatistics/00-1-age-and-year-of… Too awful
David Spiegelhalter
@d_spiegel
The most chilling graph I know is that of the times of day his victims died, compared with those of patients of other GPs
dspiegel29.github.io/ArtofStatistics/00-2-shipman-times/0…
The sharp spike around 2.30 is when he did his home visits.
David Spiegelhalter
@d_spiegel
Among others, I gave evidence at the Public Inquiry. He could have been caught earlier, if someone had been collecting and looking at the data. A sequential probability ratio test would have confidently identified a problem after only around 40 victims dspiegel29.github.io/ArtofStatistics/10-3-4-Shipman-cumul…
David Spiegelhalter
@d_spiegel
All this is covered in The Art of Statistics.
-
David Spiegelhalter reposted
Millfield School
@MillfieldSenior
🏆Our amazing mathematician Yuka spoke to @bbcsomerset @Claire7Carter this morning after she was named as the top female mathematician in the world following her performance in the 2020 International Mathematical Olympiad!
Listen back here from 2:47:00 - www.bbc.co.uk/sounds/play/p08rjgrh -
David Spiegelhalter
@d_spiegel
Great to hear that Yuka Machino (17) was the top female at the (online) International Maths Olympiad - her Gold medal helped UK into a fine 9th place out of 105. Now try the sample problem… bmos.ukmt.org.uk/home/imo-2020-pr.pdf
-
David Spiegelhalter
@d_spiegel
the 30% extra deaths at home just goes on and on - is this a benefit or a harm?
www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa… -
David Spiegelhalter reposted
Data & Statistical Science for Health, LSHTM
@LSHTM_datastats
We are excited to announce that Sir David Spiegelhalter @d_spiegel will give a webinar "Communicating Statistics in the Time of Covid" (abstract below) on 18/11/2020 at 17:30- 19:30 hrs UK time.
Details on how to join will appear on our event page soon. -
David Spiegelhalter
@d_spiegel
exactly what we should be debating. Surprised that govt announcements don’t mention age - the overwhelmingly important risk factor. Intergenerational contacts seem the crucial area
Adam Kucharski
@adamjkucharski
A key aspect I feel is missing from this piece is social contact structure, i.e. how to protect older groups without either a) shutting them off completely from society, or b) controlling infection among their contacts (which will affect younger groups):
This post is unavailable. See it on X. -
David Spiegelhalter
@d_spiegel
I presume he meant a 'very high false *negative* rate'. And ‘7%’ study shows 86% pickup with day 6 test www.gov.uk/government/publications/phe-investigation-into…
Also see
www.medrxiv.org/content/10.1101/2020.07.24.20161281v2 for proper analysis
Sky News
@SkyNews
Foreign Secretary Dominic Raab says the "challenge" with testing for #covid19 in airports is 'the very high false positive rate' and adds 'only 7% of tests will be successful in identifying those who have the virus'.
#kayburley
news.sky.com/story/coronavirus-live-latest-uk-news-update… -
David Spiegelhalter
@d_spiegel
very thoughtful piece on risk-based shielding from Raj Bhopal www.bmj.com/content/370/bmj.m3702/rr-2 "No one is advocating “letting the infection rip”, a phrase only heard from the lips of people who do not want this discussed"
-
David Spiegelhalter
@d_spiegel
and this one www.phc.ox.ac.uk/research/covid-19/projects/covid-19-risk…
george davey smith
@mendel_random
Several tools under development www.som.org.uk/occupational-health-experts-explain-how-th… www.bma.org.uk/media/2768/bma-covid-19-risk-assessment-to… etc. Communication approaches need developing; as this has been an epidemic of miscommunication needs to be got right!
-
David Spiegelhalter
@d_spiegel
very nice analysis by @TomChivers (with assistance from @kjm2)
unherd.com/2020/09/do-we-need-a-second-lockdown/?tl_inbou… (and apologies if I inadvertently assisted this incorrect claim that most positive tests are false) -
David Spiegelhalter
@d_spiegel
very refreshing to see this perspective expressed without needing abuse of either individuals or the science www.bbc.co.uk/news/health-54228649
-
David Spiegelhalter
@d_spiegel
another fine addition to this podcast series - thanks to @ilangoodman for producing these!
Risky Talk
@RiskyTalkPod
Chemical spills, food scares, pandemics: how can you manage a crisis with good communication?
Govt. advisor & security expert @DrBrookeRogers joins @foodgov veteran Lord John Krebs to share lessons from the frontlines with @d_spiegel in the latest ep!
riskytalk.libsyn.com/communicating-in-a-crisis -
David Spiegelhalter
@d_spiegel
Tweet below is for mass testing. Very different if test 1000 people *with symptoms*. then from PHE data expect 40-50 covid, detect say 30-40. So most positive tests are correct. The current rise is real.
David Spiegelhalter
@d_spiegel
If you test 1000 people at random, latest ONS figures estimate 1 will have the virus, and let’s assume you find them. But with an FPR of 0.8%, that’s 8/1000, and so you expect to find 8 false positives. That’s 9 positive tests, only one of which has the virus . Hope this is ok
-
Post not available here, view on X
David Spiegelhalter
@d_spiegel
If you test 1000 people at random, latest ONS figures estimate 1 will have the virus, and let’s assume you find them. But with an FPR of 0.8%, that’s 8/1000, and so you expect to find 8 false positives. That’s 9 positive tests, only one of which has the virus . Hope this is ok
-
David Spiegelhalter
@d_spiegel
Brooke Rogers and John Krebs are excellent here, talking about communicating about risk in a crisis. Somewhat relevant
Risky Talk
@RiskyTalkPod
Chemical spills, food scares, pandemics: how can you manage a crisis with good communication?
Govt. advisor & security expert @DrBrookeRogers joins @foodgov veteran Lord John Krebs to share lessons from the frontlines with @d_spiegel in the latest ep!
riskytalk.libsyn.com/communicating-in-a-crisis -
David Spiegelhalter
@d_spiegel
interesting job….
Office for Statistics Regulation
@StatsRegulation
As part of our work to review the approach taken to develop models to award exam results this year, we have set up an Expert Oversight Group to provide technical oversight to the OSR project team osr.statisticsauthority.gov.uk/our-regulatory-work/osr-re…
-
David Spiegelhalter
@d_spiegel
very nice, of course shape depends so much on testing regime. let’s see what happens over the next few weeks, and whether the christmas tree starts bulging
Keith Edkins 🔶 🇺🇦🇬🇱🇨🇦🇵🇦
@keithedkins
Here is a version with the gender split
-
David Spiegelhalter
@d_spiegel
excellent, would love to see this with UK data!
Stats and R
@statsandr
It is likely that someone created a graph like this, but I am unaware of it. Anyway, here is the age and sex specific incidence per 100 000, which shows the cases in relation to the population
-
David Spiegelhalter
@d_spiegel
Very interesting - shows the crucial need to report ages of cases and not just total number
Krista Fischer
@kristafischer16
Tried to reproduce your graph with Estonian data - curiously similar (despite of the >100x smaller scale on the X-axis)
-
David Spiegelhalter
@d_spiegel
My BMJ paper on using the idea of ‘normal’/background/actuarial risk to communicate the huge range of risks from COVID-19 experienced by people of different ages www.bmj.com/content/370/bmj.m3259.full?ijkey=f1FMGWUz2L39…
Post not available here, view on X
David Spiegelhalter
@d_spiegel
The risk of death from COVID-19 during the pandemic, if not currently infected, was equivalent to experiencing around 5 weeks extra ‘normal’ risk for those over 55, and steadily less for younger age-groups, reducing to just 2 extra days for school-children
David Spiegelhalter
@d_spiegel
The risk of death from COVID-19 *if infected*, roughly similar to the risk of dying from all other causes over the next year, therefore essentially doubling the risk. Bit more than a year's worth for over 55s, less for under 55s
David Spiegelhalter
@d_spiegel
Conclusion: dominating effect of age highlights the need for caution at intergenerational meetings. As a simple guide, younger people should be particularly cautious around those over 55. (NB Only look at death: do not yet have good data on long-term morbidity)
-
David Spiegelhalter reposted
Eric Topol
@EricTopol
Understanding, contextualizing the risk of death from #covid19
www.bmj.com/content/370/bmj.m3259
@bmj_latest A provocative new analysis by @d_spiegel showing how there is a close match to "normal" age-related risk of all-cause death -
David Spiegelhalter
@d_spiegel
Yes, important to point out that young people always had twice the infections of older (although not diagnosed). But current age distribution is still, I think, extraordinary
Helen Ward
@profhelenward
This is the link to the preprint (not peer reviewed) and antibody levels by age (right hand column). Most of these infections occurred in march&April www.medrxiv.org/content/10.1101/2020.08.12.20173690v2
-
David Spiegelhalter
@d_spiegel
what must the authors of this analysis be thinking when they hear the ‘only 7%’ claim from the PM?
-
David Spiegelhalter
@d_spiegel
I suspect this is the source of the 7%. It says “The outputs as presented in the following work is highly dependent on a range of assumptions"
Simon Calder
@SimonCalder
I've asked the DfT for the source
Can't be this SAGE paper from June, because:
1 It's purely theoretical; since then much real-world research has appeared, not least from Jersey
2 The paper makes it clear the 7% prediction is for asymptomatic carriers only
assets.publishing.service.gov.uk/government/uploads/syste… -
David Spiegelhalter
@d_spiegel
and in any case a single test seems a straw man - need proper cost-effectiveness analysis of reasonable repeat strategies. I assume someone is doing this!
David Spiegelhalter
@d_spiegel
PM says airport test would only pick up 7% of cases, and this means "93% of the time you could have a real false sense of security”. This is hopelessly wrong. Even if 7% were true (evidence?) the vast majority of people will have *correct* negative tests www.bbc.co.uk/news/uk-54045057
-
David Spiegelhalter
@d_spiegel
PM says airport test would only pick up 7% of cases, and this means "93% of the time you could have a real false sense of security”. This is hopelessly wrong. Even if 7% were true (evidence?) the vast majority of people will have *correct* negative tests www.bbc.co.uk/news/uk-54045057
-
David Spiegelhalter
@d_spiegel
thoughtful SAGE statement on mass testing as balance to recent hype: neatly covers today’s story of importance of thresholds for positivity www.gov.uk/government/publications/tfms-consensus-stateme…
-
David Spiegelhalter reposted
Sam Joiner
@samjoiner
For today's Times @Smyth_Chris and I analysed the latest Public Health England figures on coronavirus cases by age group. It revealed that most confirmed cases are now in younger people in an “extraordinary” shift. 1/4
-
David Spiegelhalter reposted
Tim Harford
@TimHarford
ht @d_spiegel
This looks to be a very useful resource for calculating your "Covid age" - adjusting for various other risk factors, and therefore figuring out whether you are at high risk if infected: alama.org.uk/covid-19-medical-risk-assessment/ -
David Spiegelhalter
@d_spiegel
oh dear, doing Any Questions tomorrow. www.bbc.co.uk/programmes/m000m606 that's 22 hours to become wise
-
David Spiegelhalter
@d_spiegel
my latest rant in support of stats. www.theguardian.com/commentisfree/2020/aug/31/politicians… NB as usual, I don’t write the headline (also, spot the Blackadder rip-off)
-
David Spiegelhalter reposted
Ben Ansell
@benwansell
This is great government data availability work. And for those of us who use R (or JS) they are developing packages for these languages as well. There's already helpful R code up at the link coronavirus.data.gov.uk/developers-guide
-
David Spiegelhalter
@d_spiegel
Strong response from RSS to Ofqual’s claims
Stian Westlake
@stianwestlake
We’ve issued a response to Ofqual’s baseless criticism of the @RoyalStatSoc’s letter to @StatsRegulation, in which we called for a review of this year’s A-level and GCSE adjustment models.
For anyone interested, here’s a THREAD. -
David Spiegelhalter
@d_spiegel
BBC reruns of Total Wipeout take me back to my outing to Buenos Aires in 2011. I did my stats and trained for 3 mins exertion to get through the qualifier. youtu.be/906b2wrz49k The highlight of my career.
David Spiegelhalter
@d_spiegel
I got a Times article out of it which tells the whole sorry story. The full episode is available here www.youtube.com/watch?v=VjoHC__Sifk
-
David Spiegelhalter
@d_spiegel
New podcast with great guests !
Risky Talk
@RiskyTalkPod
"This phrase 'following the science' makes me want to scream and shout.." 😱 @d_spiegel
How can govt, scientists and the media communicate evidence effectively in a pandemic? With superstars Tracey Brown @senseaboutsci @vsmacdonald & @chrischirp riskytalk.libsyn.com/communicating-evidence-in-a-pandemic -
David Spiegelhalter
@d_spiegel
so pleased to get my cool @RoyalStatSoc and @statsyss Nightingale Bicentenary badge!
-
David Spiegelhalter reposted
Royal Statistical Society
@RoyalStatSoc
Events this year have shown just how important statistics are to helping us make sense of what is going on around us. Our #statsoftheyear initiative highlights what numbers can tell us - and we want to hear what stats you think sum up 2020! Nominate here: rss.org.uk/statsoftheyear/
-
David Spiegelhalter reposted
Stian Westlake
@stianwestlake
We at @RoyalStatSoc have written to the @StatsRegulation asking them to review the 2020 exam adjustment model and the process used to design it. A review will help remedy the harms caused, and ensure this doesn't happen again.
rss.org.uk/RSS/media/News-and-publications/News/2020/14-0… -
David Spiegelhalter reposted
Tom King
@Acertaintom
The piece by @d_spiegel in HDSR still seems important:
hdsr.mitpress.mit.edu/pub/56lnenzj
One component of a functioning evaluation system. But listing that component would help us to set out what else is still needed for effective governance -
David Spiegelhalter
@d_spiegel
excellent reporting by @robertcuffe www.bbc.co.uk/news/health-53772459 Shows crucial issue is *ages* of cases, not just numbers: new data from PHE and JBC below shows few >65 in Blackburn, so few hospitalisations
assets.publishing.service.gov.uk/government/uploads/syste… -
David Spiegelhalter
@d_spiegel
PHE's new definition of Covid deaths leads to massive reduction, but BBC 10 pm News still reports that 18 people ‘died from Covid in the last 24 hour period’. No they didn’t - these are reported deaths. Why can’t the BBC just get this right?
-
David Spiegelhalter
@d_spiegel
the Washington Post’s view on the change in the way COVID deaths are counted (with some quotes from me) www.washingtonpost.com/world/britain-says-it-overcounted-…
-
David Spiegelhalter
@d_spiegel
Good new info from PHE, JBC and NHS TT on LA Watchlist areas, www.gov.uk/government/publications/national-covid-19-surv… with graphs of cases, including % positive, and detailed maps.
-
David Spiegelhalter reposted
Office for Statistics Regulation
@StatsRegulation
During the coronavirus pandemic a key question has been: How many people have died because of COVID-19?
In her new blog, our Deputy Director for Regulation, Mary Gregory, explains why this seemingly straightforward question is difficult to answer.
osr.statisticsauthority.gov.uk/the-challenges-of-counting… -
David Spiegelhalter reposted
Samira Asma
@DrSamira_Asma
“...keep an open mind”
“keep a bit back — with a little bit of probability, a little tiny bit — for the fact that you may be wrong,”
~David Spiegelhalter on Cromwell’s law.
understandinguncertainty.org/node/97 -
David Spiegelhalter
@d_spiegel
Bayes hits the New York Times! www.nytimes.com/2020/08/04/science/coronavirus-bayes-stat… - with me on Cromwell’s Law and modelling with humility
-
David Spiegelhalter
@d_spiegel
This headline (not my choice) is not very appropriate - article is more about counting than giving reasons, and I could have listed many possible factors. But at least I gave around 1000 people the pleasure of pointing out how ignorant I was 😀www.theguardian.com/commentisfree/2020/aug/02/uk-covid-19…
-
David Spiegelhalter
@d_spiegel
excellent communication - whole numbers, what low and high mean as benchmarks, current numbers being compared. Waving hands around to show magnitude is vivid (but perhaps not so effective if on radio…)
Robert Cuffe
@robertcuffe
Thanks - rting so my mum can see the clip (if she ever joins Twitter)
-
David Spiegelhalter
@d_spiegel
great to see this wonderful museum re-opening!
Florence Nightingale Museum 💙
@florencemuseum
Hand sanitisers, timed tickets, thorough cleaning…. We think Florence Nightingale, the pioneer of sanitary conditions and improved public health, would approve of the measures we’re putting in place for our re-opening on August 1
-
David Spiegelhalter
@d_spiegel
Overall ONS registered deaths www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa… have been below 5-year average for 5 weeks now. But still >100 a day happening at home rather than hospital: still not back to normal wintoncentre.maths.cam.ac.uk/coronavirus/covid-excess/
-
David Spiegelhalter
@d_spiegel
Good Guardian reporting www.theguardian.com/world/live/2020/jul/28/uk-coronavirus…
on increasingly bizarre PHE daily death count, which includes any of the 250,000+ people who have tested positive and then died of any cause. DHSC has suspended it. Makes international comparisons even more misleading than usual
David Spiegelhalter
@d_spiegel
eg ONS data for week ending 17th July www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa…: 284 Covid deaths registered in England: 184 actually occurred. But PHE dashboard count increased by a remarkable *574*, > 2 x registrations, and > 3 x occurrences. Why are they still publishing this?
-
David Spiegelhalter
@d_spiegel
Very odd. Main DHSC Covid site says "We’re pausing the publication of the daily figure” pending review www.gov.uk/guidance/coronavirus-covid-19-information-for-… . But linked dashboard reports "Daily number of COVID-19 associated UK deaths: 110” coronavirus.data.gov.uk So paused or not?
-
David Spiegelhalter
@d_spiegel
Chu Lancet paper www.thelancet.com/journals/lancet/article/PIIS0140-6736(2… often quoted re 1 vs 2 m distancing advice. But arbitrary straight-line assumption in Fig 3A *forces* relative effect of 1 vs 2m to be the same as 0 vs 1m, regardless of what data show. Hope nobody is taking results in Fig 3B seriously
-
David Spiegelhalter
@d_spiegel
Our new interactive site displaying wonderful ONS data on all-cause, COVID-19 and excess deaths in England and Wales. Shows that ‘excess’ can be very sensitive to choices wintoncentre.maths.cam.ac.uk/coronavirus/covid-excess/
-
David Spiegelhalter
@d_spiegel
Blog by Sebastian Walsh on asking the right questions blogs.bmj.com/bmj/2020/06/02/sebastian-walsh-we-are-askin… "Furthermore, it has never been “safe” to attend school. In the 2017-18 flu season, in which (unlike now) children were super-spreaders, 16 school children died. Were schools safe then? "
-
David Spiegelhalter
@d_spiegel
Another strong rebuke from @UKStatsAuth to Matt Hancock for the way data on tests is reported: 'The aim seems to be to show the largest possible number of tests, even at the expense of understanding.' Data should help the public understand both the epidemic and testing programme
UK Statistics Authority
@UKStatsAuth
Sir David Norgrove has today replied to @MattHancock, Secretary of State for Health and Social Care regarding the Government’s COVID-19 testing data uksa.statisticsauthority.gov.uk
-
David Spiegelhalter
@d_spiegel
And predictions for April 5th might have been slightly more impressive had they not been made on .. April 4th
www.theguardian.com/world/2020/may/31/covid-19-expert-kar…
Max Roser
@MaxCRoser
have a look what that research actually says wellcomeopenresearch.org/articles/5-89
In the interview the claim is 'our predictions have been accurate to within a day or two, so there is a predictive validity to our models that the conventional ones lack.'
The predictions that the authors make: -
Kevin McConway
@kjm2
Claims that children with COVID are as infectious as adults are often based on zoonosen.charite.de/fileadmin/user_upload/microsites/m_cc… . But their data suggest there *is* a difference in viral load of children and adults medium.com/@d_spiegel/is-sars-cov-2-viral-load-lower-in-y…. @kjm2 and @D_Spiegel recommend that the pre-print is withdrawn.
David Spiegelhalter
@d_spiegel
Thanks, I do know what a preprint server is. And this is not on one. So it needs to be withdrawn from the website, and replaced by a new paper - apparently this is going to happen. But still should be on a proper preprint server.
-
David Spiegelhalter
@d_spiegel
Is SARS-CoV-2 viral load lower in young children than adults? Jones et al virologie-ccm.charite.de/fileadmin/user_upload/microsites… claim no evidence for this, but their statistical analysis is inappropriate. @kjm2 and @d_spiegel recommend the pre-print is withdrawn medium.com/@d_spiegel/is-sars-cov-2-viral-load-lower-in-y…
David Spiegelhalter
@d_spiegel
Their data show 27% (95% interval 8% to 91%) (geometric) mean viral load in 1-10s compared to adults. Finding robust to alternative definitions of ‘young’.
David Spiegelhalter
@d_spiegel
Their analysis inappropriately * was not based on a pre-specified question * treated age as categorical * found significant difference among age-groups, but then lost it again by carrying out 45 pointless additional comparisons. Minor point: could have used parametric analysis.
-
David Spiegelhalter
@d_spiegel
‘The pointy end of probability’. My new podcast with wonderful Tamsin Edwards @flimsin is very special: uncertainty, climate science, communication and personal stories of dealing with cancer. Thanks to @ilangoodman for producing
riskytalk.libsyn.com/confronting-uncertainty-personally-a… -
David Spiegelhalter
@d_spiegel
My blog on the COVID risks faced by *population* during the peak of the epidemic (great ONS E+W data)
medium.com/@d_spiegel/what-are-the-risks-of-covid-and-wha… 1 in 1800 died over 5 weeks. Risks doubled if a man, and for each 6–7 years extra age: compared to 20-year-old, an 80-year-old had ~ 500 times risk -
David Spiegelhalter
@d_spiegel
excellent blog by the chief stats regulator: “Being trustworthy depends not on conveying an aura of infallibility, but on honesty and transparency.”
Office for Statistics Regulation
@StatsRegulation
Have we all become armchair epidemiologists? Read the latest blog from Ed Humpherson, DG Regulations www.statisticsauthority.gov.uk/the-armchair-epidemiologists/
-
David Spiegelhalter
@d_spiegel
Polite request to PM and others: please stop using my Guardian article to claim we cannot make any international comparisons yet. I refer only to detailed league tables-of course we should now use other countries to try and learn why our numbers are high www.theguardian.com/commentisfree/2020/apr/30/coronavirus…
-
David Spiegelhalter
@d_spiegel
Latest excellent ONS data www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa…: 22,351 deaths registered in week ending April 17th, a huge excess of 11,854. 74% of extra labelled as COVID, leaving around 3096 (26%) extra ‘non-COVID’ deaths. Proportion is larger than last week
Post not available here, view on X
David Spiegelhalter
@d_spiegel
COVID deaths in hospitals now around 400-450 a day, new CQC data www.ons.gov.uk/peoplepopulationandcommunity/healthandsoci… show around 350-400 a day in English care-homes. Add deaths at home, this suggests there are now about as many COVID deaths out of hospital as in hospital.
-
David Spiegelhalter
@d_spiegel
Whilst fully acknowledging limitations of daily reported deaths, our new charts show the power of a simple procedure to smooth out erratic reporting artefacts
wintoncentre.maths.cam.ac.uk/coronavirus/chart/ - shape of Italy’s reduction is notable -
David Spiegelhalter
@d_spiegel
Slightly delayed blog on Tuesday’s ONS data, showing 8000 excess deaths in E+W in the week ending April 10th medium.com/@d_spiegel/covid-and-excess-deaths-in-the-week… But what has caused around 1800 extra non-COVID deaths?
-
David Spiegelhalter
@d_spiegel
I wish the daily press conferences featured this graph from NHS England - clear peak 2 weeks ago on April 8, and slow steady improvement www.england.nhs.uk/statistics/statistical-work-areas/covi…
-
David Spiegelhalter
@d_spiegel
New @ONS data: 18,516 deaths registered in the week ending April 10th, 7,996 more than usual. 6,213 COVID, so 22% of this excess did not have COVID-19 mentioned (was 41% last week). More readiness to put COVID on certificate?
www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa… -
David Spiegelhalter
@d_spiegel
Good article cautioning about predictive models, eg IHME jamanetwork.com/journals/jama/fullarticle/2764824
-
David Spiegelhalter
@d_spiegel
Latest data from ONS 30 mins ago www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsa… Just drawn this extraordinary, and shocking, graph. Why the massive rise in non-covid deaths? How much is judgement of registering doctor? How much is collateral damage of lockdown???
-
David Spiegelhalter
@d_spiegel
My new blog on whether Covid raises everyone’s relative risk of dying by a similar amount. medium.com/wintoncentre/does-covid-raise-everyones-relati… Latest ONS data shows that, of recent death registrations, the proportion linked to Covid does not depend on age.
David Spiegelhalter
@d_spiegel
This means that population death rates for Covid increase exponentially with age, as a fixed proportion of overall age-specific risk. Supports idea that, for general public, Covid hugely multiplies an individual’s background risk of dying over a few weeks.
-
David Spiegelhalter
@d_spiegel
excellent discussion of death reporting from CEBM Oxford www.cebm.net/covid-19/reconciling-covid-19-death-data-in-…
-
David Spiegelhalter
@d_spiegel
excellent stats coverage from @robertcuffe and @chrisjeavans at BBC www.bbc.co.uk/news/health-52133054
-
David Spiegelhalter
@d_spiegel
Congratulations to @UKStatsAuth, @StatsRegulation, @ONS, @NHSuk, @PHE_uk, @DHSCgovuk for improvements in Covid mortality stats. UKSA and OSR have been pushing hard www.statisticsauthority.gov.uk/news/statement-from-the-of… DHSC now clear that daily counts are in-hospital only and delayed www.gov.uk/guidance/coronavirus-covid-19-information-for-…
-
David Spiegelhalter
@d_spiegel
My blog suggesting that getting COVID-19 is, rather remarkably, like packing a year’s worth of risk into a week or two
medium.com/wintoncentre/how-much-normal-risk-does-covid-r… -
David Spiegelhalter
@d_spiegel
Thoughtful blog from Michael Blastland on expert disagreement and tolerance of uncertainty medium.com/wintoncentre/quick-thoughts-on-coronavirus-and…
-
David Spiegelhalter
@d_spiegel
My take on communicating about coronavirus: medium.com/wintoncentre/communication-about-coronavirus-c…
-
David Spiegelhalter
@d_spiegel
@kjm2 and I on misuse of live facial recognition stats: "When looking for a needle in a haystack, even for the sharp-eyed, there’s a lot of bits of straw that look like needles” medium.com/wintoncentre/live-facial-recognition-how-good-…
-
David Spiegelhalter
@d_spiegel
Should live facial recognition be described as ’80% accurate’? @kjm2 and I take apart the stats
medium.com/wintoncentre/live-facial-recognition-how-good-… -
David Spiegelhalter
@d_spiegel
'Although numbers are often treated as cold, hard facts, we should be willing to acknowledge how uncertain they can be’. In my blog for Scientific American
blogs.scientificamerican.com/observations/the-problem-wit… -
David Spiegelhalter
@d_spiegel
fine (free) article 'How to communicate evidence to patients’ dtb.bmj.com/content/57/8/119
COI: by my excellent colleague, Alex Freeman -
David Spiegelhalter
@d_spiegel
How to communicate uncertainty about facts, numbers, science? We have a new paper out rsos.royalsocietypublishing.org/lookup/doi/10.1098/rsos.1…
@RSocPublishing @alex_freeman @Sander_vdLinden @nuffieldfound -
David Spiegelhalter
@d_spiegel
This paper motivates the call for the end of significance. A 25% mortality reduction, but because P=0.06 (two-sided), they declare it ‘did not reduce’ mortality. Appalling. jamanetwork.com/journals/jama/article-abstract/2724361
-
David Spiegelhalter
@d_spiegel
this is an excellent commentary on stats in research medium.com/@darren_dahly/statistical-reform-be2d4f742a4 - but would be good to know how many in UK stats depts do form serious applied collaborations. Is it really 'the exception’?
-
David Spiegelhalter
@d_spiegel
Great news that Brad Efron has won the 2019 International Statistics Prize for the 1977 bootstrap idea - so simple and yet so powerful www.statslife.org.uk/news/3982-2018-international-prize-i…
-
David Spiegelhalter
@d_spiegel
It's the press release. The paper uses the term ‘statistical’ 10 times, ‘machine learning’ 5 times, and never mentions AI.
Stephen John Senn
@stephensenn
From, hypothesis, to hope, to hype. How journalism works. Warp factor 8 applied to “we could potentially use this AI tool to intervene at an earlier stage” @theipaper
-
David Spiegelhalter
@d_spiegel
shocking that alcohol paper www.thelancet.com/journals/lancet/article/PIIS0140-6736(1… broke Lancet guidelines in only reporting relative risks: needed fine press office to extract absolute risks from them
medium.com/wintoncentre/the-risks-of-alcohol-again-2ae8cb… -
David Spiegelhalter
@d_spiegel
why you should put on someone else’s glasses and squint at the graph when judging employment trends medium.com/wintoncentre/has-uk-employment-gone-down-maybe…
-
David Spiegelhalter
@d_spiegel
Just about all I can say about risk and uncertainty communication www.annualreviews.org/eprint/SYTjSuHHYvXUxEZYydYi/full/10…
-
David Spiegelhalter
@d_spiegel
BBC Trust report on statistics says avoid false balance and challenge faulty stats on the spot: @BBCr4today failed with Lawson
-
David Spiegelhalter
@d_spiegel
Great RSS meeting on stats and media in Leeds next Wednesday - register here www.eventbrite.co.uk/e/afternoon-meeting-statistics-the-m…
-
David Spiegelhalter
@d_spiegel
Daily Mail’s own special way of showing a change from 0.6% to 0.7% www.dailymail.co.uk/news/article-4248690/Economy-grew-0-7…
-
David Spiegelhalter
@d_spiegel
Alcohol guideline risks: units go medicine, poison, poison, … understandinguncertainty.org/medicine-poison-poison-poiso…
-
David Spiegelhalter
@d_spiegel
Jeremy Hunt, the Guardian, and the importance of getting the stats right understandinguncertainty.org/jeremy-hunt-guardian-and-imp…
-
David Spiegelhalter
@d_spiegel
Useful fact sheets on risk communication from Patient Information Forum www.pifonline.org.uk/communicating-risk-in-health-informa…
-
David Spiegelhalter
@d_spiegel
Data: can we cope? Blog by Kevin McConway on the articles in Science on data deluge, but which ignore uncertainty understandinguncertainty.org/node/1259
-
David Spiegelhalter
@d_spiegel
great infographic in Guardian www.theguardian.com/news/datablog/2011/jan/14/mortality-s… - see also www.nhs.uk/Tools/Pages/NHSAtlasofrisk.aspx for an animated version of causes of death
































































































































































































































































































































































