Kevin McConway Profile
Kevin McConway

@kjm2

Followers
1K
Following
333
Media
23
Statuses
1K

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

Milton Keynes, England
Joined May 2007
Don't wanna be here? Send us removal request.
@PWGTennant
Peter Tennant has moved to Bluesky
1 year
No. This is not acceptable. The methods section is the single most important part of a scientific paper. If those details are relegated to supplementary materials, then it's not a scientific paper in a scientific journal.
115
480
5K
@d_spiegel
David Spiegelhalter
2 years
@Pouriaaa I am so pleased to see this! You and the whole team deserve the gratitude of the whole country
1
2
25
@HOS_ASA
ASA History of Statistics Special Interest Group
2 years
#OTD 1948 Douglas G Altman b (d 3 June 2018)🇬🇧Co-founder of the Equator Network. “One of the most influential medical statisticians of the past 50 years” he spearheaded massive reforms in medical research reliability & role of reporting guidelines. 1/6🧵
3
72
184
@absw
ABSW
3 years
Don't miss our next members' event with @kjm2 book now https://t.co/91km96YybA
0
3
3
@RSS_DatEthGov
RSS Data Ethics and Governance Section
4 years
The @NewtonGateway is running a 3-day virtual study group next week with 3 challenges from @anducnguyen @HardingProf and @StatsRegulation realating to how complex mathematical work is communicated and propagated. Still time to register and get involved: https://t.co/9JRSphfyzI
gateway.newton.ac.uk
This three-day virtual study group explored the challenges related to scientific misinformation/miscommunication in the media and the impact of the...
0
3
3
@kjm2
Kevin McConway
4 years
@ONS Graph of the ratio (just for England) of 'official' incidence estimates (new infections) from ONS to dashboard new cases 5/5
0
1
2
@kjm2
Kevin McConway
4 years
@ONS ..then the previous week, and despite the error margins, that must surely be a real and large increase (in early March). 4/n
1
0
0
@kjm2
Kevin McConway
4 years
@ONS I think this is probably because of reduced testing in the community, and I think it means dashboard trends (up or down) in new cases are not such a good indicator of infection trends as they used to be. The ONS estimates for the most recent week are almost a million higher.. 3/n
1
0
0
@kjm2
Kevin McConway
4 years
@ONS You'd expect dashboard cases to be smaller because of missing many asymptomatic infections, but until mid-Jan, ONS infection estimates used to be about 2 or 3 times the dashboard new case figures for the same period. They've been rising since, now about 8 times dashboard. 2/n
1
0
0
@kjm2
Kevin McConway
4 years
Similar pattern on incidence (new infections) in this week's @ONS infection survey results as last week. Total modelled new infections for latest week available (ending 9 March) from ONS is about 3.1m, with wide error margin. New cases on dashboard for that week about 380k. 1/n
2
0
2
@kjm2
Kevin McConway
4 years
@ONS ...it could mean that some of the recent increase in dashboard cases is because of more testing than real new cases (though I'm sure that most of it is likely to be real). 4/4
0
0
0
@kjm2
Kevin McConway
4 years
@ONS ...margin of error) and the dashboard has about 250k, so survey has 8 times as many. Not just rogue week - the proportion has been changing in that dirn for over a month. Presumably because people aren't being routinely tested &/or not reporting LFDs. On the bright(?) side.. 3/n
2
0
0
@kjm2
Kevin McConway
4 years
@ONS ..been about 2 or 3 times the dashboard confirmed case figures for the same period - not surprising they are bigger because the survey much more likely to pick up asymptomatic cases (etc). But for the latest week of modelled estimates, the survey has about 2.1m (with wide... 2/n
1
0
0
@kjm2
Kevin McConway
4 years
Today's @ONS Covid-19 survey results show something interesting in the incidence (new cases) estimates. They are always behind the positivity estimates for technical reasons - this week they go up to 26 Feb with modelled estimates up to 2 March. Historically they've usually.. 1/n
1
2
0
@alexvespi
Alessandro Vespignani
4 years
“Investing in random sampling can better prepare governments for the future” by @nataliexdean https://t.co/c3DHDrY8j4
1
59
160
@kjm2
Kevin McConway
4 years
Has something happened to the @Eurostar Android app? I tried to save some tickets from it to Gpay, nothing happened. So I tried to check on the Play Store whether there was an upgrade, Search for Eurostar revealed nothing, and links to app on Eurostar website said URL not found.
0
0
4
@thomasforth
Tom Forth
4 years
My strongest possible agree with this piece. And then even more times ten. The ONS has improved from good to excellent in the past 5-10 years. On regional statistics. The website. The accessibility. And it overachieved enormously during Covid.
Tweet card summary image
economist.com
It did more as its job became harder
6
24
179
@involveUK
involve
4 years
. @CamdenCouncil has agreed a new Data Charter, drafted by a Residents’ Panel https://t.co/Iganuho4iq The development of the Charter didn’t happen in darkened rooms, hidden in the Council buildings. #open #transparent #democracy
1
6
4
@kjm2
Kevin McConway
4 years
@ONS ...But vax rates a lot different too. In all English regions but two, according to the Govt dashboard for 29/12, between 50% and 60% of 12+ have had 3rd jab. Same for Wales and NI. SW England is 61%, Scot is 62%. London is 38%.
0
0
0