John Burn-MurdochCompte certifié

@jburnmurdoch

Stories, stats & scatterplots for | Daily updates of the coronavirus trajectory tracker | john.burn-murdoch@ft.com |

Doncaster   London
Inscrit en juin 2009

Tweets

Vous avez bloqué @jburnmurdoch

Êtes-vous sûr de vouloir voir ces Tweets ? Les voir ne débloquera pas @jburnmurdoch

  1. Tweet épinglé
    12 mai

    NEW: we’ve updated our excess mortality tracker, the gold-standard measure for Covid deaths, allowing like-for-like comparisons btwn countries UK had 50,000 more deaths than usual in March & April vs 27,000 reported Covid deaths at the time Free to read:

    Afficher cette discussion
    Supprimer
  2. il y a 3 heures

    • Antibody testing suggests ~15% of NYers (~20% in NYC) have had Covid vs • "I did some back of the envelope extrapolations and found that 83% of NYers have had Covid. Here are my charts from Google Sheets." I know which one I’m going with...

    Afficher cette discussion
    Supprimer
  3. il y a 3 heures

    Right on cue, this drops into my inbox 🤔🤔🤔🤔

    Afficher cette discussion
    Supprimer
  4. il y a 3 heures

    My other tip: follow lots of experts. For me, that means , , , , , , and others. They don’t always agree! And that’s *good*. This stuff is complicated.

    Afficher cette discussion
    Supprimer
  5. il y a 3 heures

    We already know from comprehensive research in other countries that the share of people who've had Covid in even hard hit countries, is around 5%. Claims that differ significantly from that require extraordinary evidence

    Afficher cette discussion
    Supprimer
  6. il y a 3 heures

    My tip for anyone, fellow journalist or otherwise: weigh any surprising new claims against the balance of evidence already out there on the issue in question.

    Afficher cette discussion
    Supprimer
  7. il y a 3 heures

    Yesterday's Manchester paper is a particularly acute case, as the claims in that study concern a critical issue that people will use to justify policy — how many people in the UK may have already had Covid.

    Afficher cette discussion
    Supprimer
  8. il y a 3 heures

    It's absolutely vital that as journalists we do all the necessary checks before reporting on highly sensitive issues like this.

    Afficher cette discussion
    Supprimer
  9. il y a 3 heures

    Few weeks ago media reported studies saying air pollution levels had big impact on Covid death rates. Problem 1: studies hadn't been peer reviewed. Problem 2: air pollution & pop dens are correlated. Over at SMC, experts pointed out flaws:

    Afficher cette discussion
    Supprimer
  10. il y a 3 heures

    And here's a detailed, point-by-point take-down of the same paper by , including an explanation that even peer review isn't enough to ensure a study's findings are watertight:

    Afficher cette discussion
    Supprimer
  11. il y a 3 heures

    A site I'm finding increasingly useful is the Science Media Centre (), where domain experts present feedback to papers circulating in the media. Here are and others pointing out the flaws in that Manchester paper

    Afficher cette discussion
    Supprimer
  12. il y a 3 heures

    Thread: Critical assessment of scientific papers by the media has never been more important than during the pandemic That new Manchester study saying 25% of UK has HD Covid *was* peer reviewed, but has already been comprehensively debunked by many leading epidemiologists.

    Afficher cette discussion
    Supprimer
  13. a retweeté
    13 mai

    In France, it’s currently possible for a woman simultaneously to be fined €165 for being in public with a face covering and €135 for being in public without a face covering

    Afficher cette discussion
    Supprimer
  14. il y a 21 heures

    Covid data is extremely poor quality and often doesn’t mean what you think it does, exhibit #5817

    Supprimer
  15. a retweeté
    13 mai

    the UK government has now stopped publishing international COVID-19 mortality comparisons but you can still find them at the

    Supprimer
  16. a retweeté
    13 mai

    Today both France and Spain (two badly hit countries) have come out with seroprevalence numbers. (France is a model, paper linked below. Spain was representative sample/survey of 60K; I don't have the paper yet). Both are ~5 percent. If that holds up, that is very very bad news.

    Afficher cette discussion
    Supprimer
  17. a retweeté
    13 mai

    Indeed, folks need to be much more careful with the "Get ready for cases to EXPLODE in [state/county]!" types of headlines They tend not to recognize that behavioral and regulatory changes are often incremental, and the data is noisy—and they have often turned out to be wrong.

    Supprimer
  18. a retweeté
    13 mai

    We’ve got a superb line up of on the Media Show to discuss ‘patriotic journalism’ in a time of crisis and the new rock starts of the media, data journalists & 4pm

    Supprimer
  19. 13 mai

    Edit: analysis of data was by / , not

    Afficher cette discussion
    Supprimer
  20. 13 mai

    This means govts can’t just switch the economy back on to full power overnight. People will need to feel completely safe before they start piling back into restaurants etc, regardless of when rules permit that. And some behaviours may simply *never* return to normal.

    Afficher cette discussion
    Supprimer
  21. 13 mai

    Again, this matters because same will be true as restrictions ease. Govts may like to think they can just switch activity on/off, but people will act independently. Some will move around more despite no loosening of restrictions; others will stay home weeks after lockdowns end.

    Afficher cette discussion
    Supprimer

Le chargement semble prendre du temps.

Twitter est peut-être en surcapacité ou rencontre momentanément un incident. Réessayez ou rendez-vous sur la page Twitter Status pour plus d'informations.

    Vous aimerez peut-être aussi

    ·