Kevin McConway

@kjm2

Emeritus professor of statistics ; Milton Keynes inhabitant; choral singer; perpetually puzzled person, interested in anything interesting.

Milton Keynes, England
Joined May 2007
17 Photos and videos Photos and videos

Tweets

You blocked @kjm2

Are you sure you want to view these Tweets? Viewing Tweets won't unblock @kjm2

  1. Aug 13

    These are periods, well, really, the only periods in the data, when deaths were rising fairly fast. So what's the reason? Artefact? Different reporting or certifying behaviour? Any thoughts? 5/5

    3 replies . 2 likes
    Show this thread
    Show this thread
    Undo
  2. Aug 13

    ...that's not true for 3 periods: right at the start of the pandemic (13-20 March 2020), then from mid-Oct 2020 to mid-Jan 2021, then again recently (from early July 21). Then the ONS figures are still higher (except for 2 weeks), but much less so, generally well under 10%. 4/n

    1 reply . 1 like
    Show this thread
    Show this thread
    Undo
  3. Aug 13

    So they should align, and avoid problems of late reporting or registration. (I'm ignoring the latest numbers where those issue do interfere.) As is well known, the ONS figures are generally higher than the dashboard ones for the same week, by typically about 30 or 40%. But... 3/n

    1 reply . 1 like
    Show this thread
    Show this thread
    Undo
  4. Aug 13

    I'm comparing the ONS weekly registrations in England and Wales, of deaths where Covid-19 is mentioned on the certificate, but classified by date of death, and the weekly totals of the dashboard 28-day figures for E&W, also classified by date of death. Same date classificn. 2/n

    1 reply . 1 like
    Show this thread
    Show this thread
    Undo
  5. Aug 13

    I've been looking at ONS death registration data and at the counts of Covid deaths on the 28-day definition on the dashboard, and noticed something that must have been seen and explained before, but I haven't seen it. Any thoughts? I dare say it's obvious. 1/n

    4 replies . 1 retweet 2 likes
    Show this thread
    Show this thread
    Undo
  6. Retweeted
    Jun 23

    Now we are seeing a return to non-Covid health stories, it’s time for press officers, journalists and anyone else to use RealRisk! (with links to our videos)

    4 replies . 37 retweets 103 likes
    Undo
  7. Retweeted
    Jun 8

    It's been announced that implementation of the central GPDPR digital database will be delayed until 1 September. So there's some time to reconsider aspects. We encourage all stakeholders to use this time wisely, so we get a system that benefits patients and is trustworthy.

    4 retweets 4 likes
    Undo
  8. Retweeted
    Jun 2

    The statistical approach to the 2020 summer exams, and the broader legal and ethical import on the use of comprehensive systems in public administration, is the subject of our next meeting, on Tuesday 8th June, starting at 2:30pm (UK time)

    1 reply . 3 retweets 2 likes
    Show this thread
    Show this thread
    Undo
  9. May 27

    Definitely of interest to teachers, I'd say. If you can't be there at the start (e.g. because you're teaching) you might still want to tune in for later talks. It's free to RSS e-teacher members.

    On 8 June, 2.30-6.30, meeting on "Statistical accuracy sacrificed to gain public acceptance: Legal, ethical and communication lessons from the UK and Irish 2020 summer exams models", organised by Statistics & Law section and . Sign up at .
    Show this thread
    Undo
  10. Retweeted
    Apr 9

    Please take a look my latest blog in which I think carefully about the work of and Paul Ernest. How do you feel about these issues?

    NEW BLOG: Mathematics as an ethical practice
    2 replies . 11 retweets 20 likes
    Undo
  11. Apr 2

    (c) It *is* news that SAGE haven't agreed a UK-wide figure, but they've been saying for a long time that that don't think UK-wide R ranges are very meaningful or helpful.

    1 reply . 1 retweet 2 likes
    Show this thread
    Show this thread
    Undo
  12. Apr 2

    (b) It's actually quite good news - 1 is the highest they think it's likely to be so they are saying it's most probably smaller and infections still falling, despite schools reopening etc. 2/3

    1 reply . 2 likes
    Show this thread
    Show this thread
    Undo
  13. Apr 2

    I'd say this comes close to scare-mongering, . (a) Not news that SAGE say England's R could be as high as 1. The England range is 0.8 to 1.0 unchanged from last week. 1/n

    1 reply . 3 likes
    Show this thread
    Show this thread
    Undo
  14. Mar 31

    Guidance from and me for statisticians (and others?) engaging with the media on Covid-19 (or anything else) now appearing at . Sound as human as you can, keep off the jargon, realize that it might go wrong, and learn from what went wrong.

    1 retweet 8 likes
    Undo
  15. Retweeted
    Mar 31

    Guidance for any statisticians willing to engage with the media on COVID (and anything else), by and me. Basically, try and sound human, and be prepared for it all to go wrong

    7 replies . 65 retweets 166 likes
    Undo
  16. Retweeted
    Mar 17

    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: Statistics Regulator (Economy)

    1 reply . 13 retweets 11 likes
    Undo
  17. Retweeted
    Mar 16

    I've written a short piece about the precautionary principle and the AstraZeneca Vaccine issues

    8 replies . 56 retweets 136 likes
    Undo
  18. Retweeted
    Feb 20

    Comp! RETWEET THIS TWEET and one retweeter will win a copy of Born to be Mild and a set of Very British Problems books! (Comp will finish Sunday 21st Feb 2021 at midnight and the winner be chosen at random and informed on the week of Monday 22nd)

    121 replies . 1,754 retweets 779 likes
    Undo
  19. Retweeted
    Feb 19

    good analysis refuting misinformation about deaths in 2020

    20 replies . 74 retweets 161 likes
    Undo

@kjm2 hasn't Tweeted yet.

Loading seems to be taking a while.

Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.

    You may also like

    ·