Heading into winTier ..... the one we hoped it would be over before...

[dancemom]

New thread Tierers ...

but I think the ONS is based on a not huge sample in Scotland (and wales/ NI) , so they've always said trends are a bit more uncertain in the devolved nations.

This is worrying @ResilienceWanker if ONS figures do not extrapolate well in Scotland due to sample size. You'd think ONS could fix that 
because these figures HAVE been used by Gov to justify restrictions.
I distinctly remember it being said in Scottish parliament in early January how well our protections over Christmas had done because we had lower rates according to ONS 

• 7,195 new cases of COVID-19 reported.
• 16 new reported deaths of people who have tested positive
• 11 people were in intensive care yesterday with recently confirmed COVID-19
• 1,041 people were in hospital yesterday with recently confirmed COVID-19
• 4,433,961 people have received their first dose of a COVID-19 vaccination, 4,154,182 have received their second dose, and 3,415,942 have received a third dose or booster

Our cases are now 20% higher than England's using NS's preferred measure of rates (as previously used to demonstrate our rates being 20% lower than England's)

@imstilljenny2

What’s her explanation?

Fewer people in hospital with covid. The numbers have swung about over the past week. A difference of 50 on two or three days. Maybe not all the health boards have been reporting daily? It's still a small fraction of the overall number in hospital but it was inching down then it suddenly went up by about 5% one day and increased by approx 20/25% over a weekish.
Which is WEIRD.
(In my opinion)

Wouldn't lower sample sizes just be reflected in wider confidence intervals for the ONS sampling? They're are a little wider than
England's, but smaller than Wales and NI, which makes sense as that's the order of population sizes. Even considering the error bars there's still a clear upward trend in Scotland compared to a downward trend in the other UK nations, so I'm inclined to believe it (especially as the trend seems to be confirmed in the daily figures, and even ZOE was saying Scotland was increasing recently).

But yes, 🤣🤣🤣🤣 at Chris Musson using her own -dodgy- statistical presentation to make the differences look worse. I wonder if anyone will ask her about this (at which point it will obviously not be a competition and they're infantile for even raising it). It's the trend that's most telling I think, rather than the snap shot in time. Looking purely at the data, our additional restrictions appear to be making infection rates worse somehow!

Why is there greater uncertainty over the numbers at the moment? It says so on the screenshot.

@StarryEyeSurprise, the only thing that I can see in the screen shot that mentions uncertainty refers to the trend, not the numbers. So if the numbers for a nation have been bouncing around for the past few weeks, the trend is uncertain. So this applies to Wales and NI, but not to England (where the downward trend is clear) or to Scotland (where the upward trend is clear).

It really does look as though NS’ handling of all this has made things worse.

I said at Christmas it would become clear whether BJ or NS was proved correct and it is now pretty.l clear.

It's beside each graph for each country - the others have greater uncertainty from the 16th Feb and Scotland from the 17th Feb. So, somethings changed which has resulted in greater uncertainty across the four nations.

@wouldbegood The deaths are lower per population in Scotland, if you look at Travelling Tabby ( which @Y0ucannot) referred to earlier.

@StarryEyeSurprise I really don't think there's anything unusual about the latest few days of modelling being subject to some uncertainty. This uncertainty is still incorporated into the figure as the grey lines. From the ONS report:
"These early estimates are provisional and are subject to change as we receive more data, but they have undergone sufficient quality assurance to ensure that they are based upon an acceptable number of test results received up to the end of the reference week."

Yeah, I've never seen it on the graphs before though. Not massively important when we're possibly heading towards WW3, just wondered what it referred to.

Oh and @StarryEyeSurprise I don't know which bit of Tabby you're looking at, but excess deaths are currently higher in Scotland than any of the other UK nations (which I seem to recall was reported recently). You might be able to argue that COVID deaths are lower (although that's subject to quirks like levels of testing, and early on we were doing a fraction of England's testing), but since dead is dead 'overall death' is usually considered the gold standard measure.

@StarryEyeSurprise

Yeah, I've never seen it on the graphs before though. Not massively important when we're possibly heading towards WW3, just wondered what it referred to.

Totally agree that case numbers don't matter any more and there are far bigger things to be worrying about than COVID, at home and abroad!

@Y0uCann0tBeSer10us

Oh and *@StarryEyeSurprise* I don't know which bit of Tabby you're looking at, but excess deaths are currently higher in Scotland than any of the other UK nations (which I seem to recall was reported recently). You might be able to argue that COVID deaths are lower (although that's subject to quirks like levels of testing, and early on we were doing a fraction of England's testing), but since dead is dead 'overall death' is usually considered the gold standard measure.

Total covid deaths is what I was referring to .

@Y0uCann0tBeSer10us

Wouldn't lower sample sizes just be reflected in wider confidence intervals for the ONS sampling? They're are a little wider than England's, but smaller than Wales and NI, which makes sense as that's the order of population sizes. Even considering the error bars there's still a clear upward trend in Scotland compared to a downward trend in the other UK nations, so I'm inclined to believe it (especially as the trend seems to be confirmed in the daily figures, and even ZOE was saying Scotland was increasing recently).

But yes, 🤣🤣🤣🤣 at Chris Musson using her own -dodgy- statistical presentation to make the differences look worse. I wonder if anyone will ask her about this (at which point it will obviously not be a competition and they're infantile for even raising it). It's the trend that's most telling I think, rather than the snap shot in time. Looking purely at the data, our additional restrictions appear to be making infection rates worse somehow!

Yes, I think you're right re the error bars. I just remember that a few weeks ago they mentioned the trend (in the devolved nations) was uncertain, and I assume that is just because the error bars overlapped, so although it looked like numbers were going in one direction generally, they couldn't be certain as the margin of error was greater than the difference in figures week on week. So not so much that they were sampling fewer people in Scotland proportionately, but that because the population is smaller, fewer people WERE sampled in total, leading to bigger error bars, so less confidence in any statistical difference week on week.

But at Chris musson. I do kind of think that it's NOT a competition... But it is annoying that we've had restrictions based on these figures for so long... And when they are "good" figures it's "because we've been being safe" and when they are "bad" figures it's random, but a Reason to keep restrictions that bit longer. I hope the public inquiry has a big long session about statistics and their (mis) use throughout this, frankly.

Sorry, o forgot the party line that NS is our great and good leader and has made everything better, thank you

• 5,484 new cases of COVID-19 reported.
• 11 new reported deaths of people who have tested positive
• 11 people were in intensive care yesterday with recently confirmed COVID-19
• 1,093 people were in hospital yesterday with recently confirmed COVID-19
• 4,434,404 people have received their first dose of a COVID-19 vaccination, 4,154,924 have received their second dose, and 3,419,928 have received a third dose or booster

Wednesday 1093 in hospital with covid
Thursday 1041 in hospital with covid
Friday (today) 1093 in hospital with covid

@mibbelucieachwell

Wednesday 1093 in hospital with covid Thursday 1041 in hospital with covid Friday (today) 1093 in hospital with covid

Yeah, I have serious reservations over these numbers.

Hmmm. Yes, all looking a bit weird. Mind you, the UK figures today have exactly the same last 3 figures as yesterday, so I'm donning a tin foil hat and calling it all out as a conspiracy today, frankly

• 5,498 new cases of COVID-19 reported.
• 0 new reported deaths of people who have tested positive (Noting that Register Offices are generally closed at weekends)
• 19 people were in intensive care yesterday with recently confirmed COVID-19
• 1,175 people were in hospital yesterday with recently confirmed COVID-19
• 4,436,113 people have received their first dose of a COVID-19 vaccination, 4,157,209 have received their second dose, and 3,430,747 have received a third dose or booster

80-odd more hospitalisations over the weekend, and not a kick in the arse off of a doubling of ICU cases. Not great is it! Every other nation's hospitalisations are in decline.

• 7,497 new cases* of COVID-19 reported.
• 21 new reported deaths* of people who have tested positive
• 18 people were in intensive care yesterday with recently confirmed COVID-19
• 1,191 people were in hospital yesterday with recently confirmed COVID-19
• 4,436,511 people have received their first dose of a COVID-19 vaccination, 4,157,929 have received their second dose, and 3,434,063 have received a third dose or booster

• From 1 March 2022, the national case definition has been updated to include reinfections. This means the headline number of COVID-19 cases and deaths now includes cases based on an individual’s first positive test as well as possible reinfections, where an individual tests positive 90 days or more after their last positive test.