What do you make of this?

Status
Not open for further replies.
Amity Island said:
Total covid19 tests done to date
19,583,360

Total covid19 positives found to date
423,236 (which hasn't been adjusted for any false positives -- which is his grievance)

Using a false positive average of 2.3% of the total number of tests done, not just 2.3% of the positive tests (Government's 0.8% low and 4.3% high averaged to 2.3%)
Click to expand...

Oh darn it. You're right, and I'm wrong! (Though I think there is also a kind of thing that's what I referred to as the false positive rate, it's not what we're talking about here.)

I'm still not sure it matters as much as he's making out. The ONS survey (where they do extra checking (I think they check each positive test again, or something like that)) is currently producing a rate of 1 in 500 or so. So that means their false positive rate can't be very much larger than 0.2% right now, and since they haven't changed their processes that much recently it can't be very much more than 0.02% (since they were measuring that a month or so ago).

So as the ONS survey numbers continue to increase (presuming they do) we can be reasonably confident in them (at least that they're measuring something actually meaningful, though not necessarily infectious people). And we have the other measures of hospital admissions and eventually deaths; I agree the numbers don't match up with what was seen in March/April but I think it's plausible that's explained by just the enormous differences in testing.

(The 19,583,360 tests done doesn't seem all that helpful. What would be more meaningful would surely be people rather than tests? That kind of data can be found in the annex of the weekly reports, 10 September to 16 September. The difference looks like a factor of roughly 3. Oddly it looks like about the same factor for total tests/people and positive tests/people.)
 
What he said does not make sense. Thats because it was nonsense. He is so poor with numbers that he cannot see it.

A tiny proportion of the tests that have been returned as positive will be in error. That has been expressed as a percentage. What he and other numerically illiterate people have done is to take that percentage and use it to give a measure of the "false positive" rate by applying it to the total number of tests carried out. The result is you get the nonsensical numbers you are referring to, @rebrascora. All he wanted to do was to try and make things look better than they are. Do not trust any numbers he comes out with. Double check everything he claims, you cannot take what he says at face value.
 
Overall I think all the false-positive angst is pretty dumb, in the face of facts.

First: No diagnostic is perfectly sensitive & specific - ie you always have some false positives and some false negatives. This is is true of every medical diagnostic & nobody says it means we should abandon medicine or whatever.

Second: The false positive rate for gold-standard coronavrius PCR tests must be very low. For example, here in Oz outside of Victoria the positivity rate is less than 0.02% in recent weeks which naively puts on upper bound on the false positive rate. In my state, there are *zero* cases being detected which tightens the upper bound further. Similar comments if you look at NZ.
 
I thought the ONS were testing random members of the population, not people who had symptoms or were contacts.
 
Docb said:
It would be easier to find the last "More or Less" on BBC sounds and have a listen where somebody far more articulate than me explains it. Maybe the episode one before last.

Essentially there are two ways of defining the false negative rate and unless you say which definition you are using then whatever you say is rubbish - as demonstrated by Hancock and that crazy interviewer who thought she knew what she was talking about but clearly showed she did not.
Click to expand...
Morning Doc B,

Thanks very much for your posts, they're really helpful.

So, I found the programme you mentoned, Wednesday 23rd Sept at the BBC radio 4 More or Less interview with David Spiegelhalter from the University of Cambridge (great find btw) had a listen, although it takes a bit of thinking to understand it.:D

Suprisingly, the crucial bit about the "9/10 tests being false claim" is confirmed to be true at 13.10mins (which is what the Paul Weston video is refering to I think). Spiegelhalter said the 9/10 false positives claim is the case "where a 1000 people are tested at random" in the community using a 0.8% false positve. But then goes on to say, but at pillar 2 this is not the case because in pillar 2, they are only testing those with symptoms.

Like you said, he does confirm that the politicians (the ones informing the public) have no understanding of the stats.

Doc B, So the question to solve this mystery seems to be, "have/are people in the community been tested at random?"

Doc B, do you happen to know what the answer to this question is?

The answer would be a combination of the random testing and the pillar 2 testing to get an overall truth.

Thanks Very Much

www.bbc.co.uk

BBC Radio 4 - More or Less: Behind the Stats - Downloads

Podcast downloads for More or Less: Behind the Stats
www.bbc.co.uk www.bbc.co.uk
 
Eddy Edson said:
Overall I think all the false-positive angst is pretty dumb, in the face of facts.
Click to expand...

That's my feeling, too. Yes, the numbers are probably slightly wrong, but we now have increases in hospital admissions, increases of people in ITU, and the beginning of increases in deaths.

Maybe the relative low numbers of deaths compared to cases isn't explained by the radically different testing regime now. (I'm quite partial to the viral load idea: that physical distancing, mask wearing, hand washing, means the average infected person gets a much lower amount of virus than they did back in March. That one feels to me more plausible than the population immunity, mutating virus, etc., ideas.)
 
@Amity Island. Yes, ONS are doing random tests in the community and these data are used to provide some of the stats. It will not, and cannot, provide data of test accuracy because as far as I can see you cannot get that from that sort of testing. I am not sure you could get a definitive answer from any of the routine testing. You would need a specific experiment to do that. I would imagine that was done as part of the test protocol development and that data will be around somewhere but probably not readily available. The best assumption is that the test would not exist if it gave silly answers.
 
Bruce Stephens said:
That's my feeling, too. Yes, the numbers are probably slightly wrong, but we now have increases in hospital admissions, increases of people in ITU, and the beginning of increases in deaths.

Maybe the relative low numbers of deaths compared to cases isn't explained by the radically different testing regime now. (I'm quite partial to the viral load idea: that physical distancing, mask wearing, hand washing, means the average infected person gets a much lower amount of virus than they did back in March. That one feels to me more plausible than the population immunity, mutating virus, etc., ideas.)
Click to expand...

For me, the conceptually simplest answer lies just in the huge increase in fatality rates with age, better protection of older people since the early days, and more testing of young 'uns. But dunno if the data actually backs that up.
 
Docb said:
@Amity Island. Yes, ONS are doing random tests in the community and these data are used to provide some of the stats. It will not, and cannot, provide data of test accuracy because as far as I can see you cannot get that from that sort of testing. I am not sure you could get a definitive answer from any of the routine testing. You would need a specific experiment to do that. I would imagine that was done as part of the test protocol development and that data will be around somewhere but probably not readily available. The best assumption is that the test would not exist if it gave silly answers.
Click to expand...
Thanks Doc B
 
There aren’t enough tests to randomly check the population in any event. And there is no way you can check if a positive is false from the available data. There are an unknown number of people who have the virus, but are asymptomatic . So this discussion is becoming picky about stats that aren’t based on reliable data.

This situation is one where pragmatism rules. If the number of cases increases, and the number of deaths increases, then do something about it. It’s as simple as that. Plus, of course, have an effective trace and test system, as operates in Wales, Scotland, and NI.

No amount of statistical analysis will help this, but for sure it can hinder it. It’s a bit like checking the tuning of the orchestra on the Titanic.
 
mikeyB said:
So this discussion is becoming picky about stats that aren’t based on reliable data.
Click to expand...
Hi Mikey B,

I agree, I do, but it's obviously important to know if say 9/10 tests are false positives, when it's the testing and not the deaths/hospital admissions as the primary reason behind removing peoples liberties and freedoms.

This looks like it's going to be a permanent change in the world, and it seems less and less about the virus and more and more about some kind of Green Agenda (e.g shop from home, eat out at home, holiday near home, work from home, study at home, speak to your g.p at home, stay away from hospitals etc etc) or some other agenda which is not being made public. How can it be that "they" are restricting human interaction, socializing, meeting family, making it almost impossible to travel, to sing, to dance, live freely and be human in the name of protecting health and reducing hospital admissions, when at the same time they are making millions off selling toxic tabacco, toxic alchohol, promoting gambling, where were the lockdowns for these things which are bigger risks to life than covid19? Why isn't this being put into the context of other life risks which far out weigh the risk of death from covid19. It all seems politically driven and not science driven and its all having a devastating impact on peoples mental health, wellbeing, jobs, businesses and freedoms.

We can't sing in a pub, but we can drink ourselves into an early grave and be admitted to hospital putting uncessary strain on the NHS and others that support you.

Mikey B, I was watching TV the other day, when I saw 2 people shaking hands, my immediate response was one of horror I thought, what the hell are they doing! It cannot be right to think like that about seeing someone shake hands?
Something is a miss.
 
Last edited:
I don’t disagree with anything in that response,@Amity Island. You might also ask why we are told to wear masks, keep our distance and tolerate 10pm closing of pubs and restaurants, when in the House of Commons the bars are open well after 10, masks are not compulsory, nor is the rule of six or 2m distancing. A fine example to the general population, I must say. The government doesn’t even comply with its own rules - why should we?
 
mikeyB said:
I don’t disagree with anything in that response,@Amity Island. You might also ask why we are told to wear masks, keep our distance and tolerate 10pm closing of pubs and restaurants, when in the House of Commons the bars are open well after 10, masks are not compulsory, nor is the rule of six or 2m distancing. A fine example to the general population, I must say. The government doesn’t even comply with its own rules - why should we?
Click to expand...
I don't like the direction all this is heading in and it seems to be in one direction at quite some speed...
 
Last edited:
mikeyB said:
I don’t disagree with anything in that response,@Amity Island. You might also ask why we are told to wear masks, keep our distance and tolerate 10pm closing of pubs and restaurants, when in the House of Commons the bars are open well after 10, masks are not compulsory, nor is the rule of six or 2m distancing. A fine example to the general population, I must say. The government doesn’t even comply with its own rules - why should we?
Click to expand...

Actually the thing about the bars in the Houses of Parliament really was false news, as pointed out by the speaker in the house this morning. As for the other things, it is a place of work and they do what is possible like any other place of work. Your other reflections on the quality of government, the lies, deception and general incompetence I have much sympathy with.
 
Docb said:
Actually the thing about the bars in the Houses of Parliament really was false news, as pointed out by the speaker in the house this morning. As for the other things, it is a place of work and they do what is possible like any other place of work. Your other reflections on the quality of government, the lies, deception and general incompetence I have much sympathy with.
Click to expand...
I was just going to say this too!
 
The only thing I can pick flies with @Docb 's post is his use of the invented term 'false news'. It's the UK and the UK Government etc that we are discussing whereas the terminology used as far as I'm concerned has only ever been applicable in the US to one specific individual.

What's wrong with the time honoured term 'Red herring' ?
 
I quite deliberately used the term "false news" instead of "fake news" for the reasons you suggest.

Red herring does not fit at all. A red herring is something that is true but irrelevant and refers to something diverting attention from a substantial point. The report about the bar opening times was false and it appeared in the news so false news seemed entirely appropriate.

Get out of that! 🙂
 
Docb said:
I quite deliberately used the term "false news" instead of "fake news" for the reasons you suggest.

Red herring does not fit at all. A red herring is something that is true but irrelevant and refers to something diverting attention from a substantial point. The report about the bar opening times was false and it appeared in the news so false news seemed entirely appropriate.

Get out of that! 🙂
Click to expand...

In that case, you should not have used the word 'news' after 'false'. You should have used 'false' instead of 'thing' after 'Actually the'.

I accept the ruddy fish.
 
Nowt worse for non pedants than a couple of pedants being pedantic. 🙂

OK, your rewording suggestion is probably more elegant but it looses the opportunity to have a bit of fun with the ideas of fake and false news.
 
Status
Not open for further replies.
Back
Top