COVID-19/Coronavirus Information and Support Thread (see OP for useful links)

  • Thread starter baldgye
  • 13,234 comments
  • 551,067 views
x3ra
False negatives are interesting but discussed even less than false positives.
Click to expand...
I think this is untrue. If I read the report right false negatives take up a page and a half of the report while false postives get less than a page. Either way, they're more dangerous than false positives.

x3ra
So are you saying that the number of false positives doesn't have anything to do with the number of tests taken? Because that's not what the govt website says.
Click to expand...
Where does it say that? I can only see it discussing false positives as a percentage of the number of positive tests, not negative tests or total tests taken.
 
Last edited:
x3ra
False negatives are interesting but discussed even less than false positives.

So are you saying that the number of false positives doesn't have anything to do with the number of tests taken? Because that's not what the govt website says.
Click to expand...

To be fair, it was clear what you meant to refer to in your post.

Whether it's technically FPR or not is a quibble that hardly matters given how wide of the mark you were with your conclusion.
 
While this absolutely makes sense, I didn't really think about it:

Association of Daily Wear of Eyeglasses With Susceptibility to Coronavirus Disease 2019 Infection

In short, if you're a daily wearer of eyeglasses you may have less of a chance of being infected with COVID than those who don't wear glasses. It's a pretty limited study, but since wearing goggles apparently helps prevent contracting COVID I have to imagine eyeglasses offer some increased protection, no matter how minor.
 
Outspacer
Must admit in that case that I didn't know how the term FPR was defined, so in my previous post I was using it as he was (not that it changes the math).

Is there a specific term for the ratio of false positives to tests?

Edit: As far as I can find out, the best advice is to check what a particular paper says FPR is! Other than that, wikipedia says that False Positive Ratio is (typically) the ratio of false positive results to tests that should have come out negative. That makes a lot more sense to me as it can deal with the situation where there are zero true positive results.
Click to expand...
Yeh, I think the 'paper' is going by a different definition of FPR:

https://assets.publishing.service.g...9_Impact_of_false_positives_and_negatives.pdf

Either way, though, the number of positive results during July was still way lower than 2% of all tests done (i.e. it was around 0.6% IIRC). So even if every single positive was a false positive, the 'operational FPR' cannot be greater than this number (unless the entire country suddenly switched to the same, less reliable tests).

The point is, though, that no matter which way you slice it or define FPR, the argument that the 'second waves' currently underway across Europe are somehow not real and/or are simply artefacts relating to unreliable test results is nonsense.

-

Today is the first day of term for Scottish universities and thus the students are back. Fortunately, there is no face-to-face teaching this semester, so hopefully the health risk to staff will be minimized.

I was surprised to learn, however, that my friend at Heriot-Watt University in Edinburgh will be resuming her lectures. I'm not sure why some universities and departments are doing face-to-face teaching while (most) others are not.
 
Last edited:
Touring Mars
Yeh, I think the 'paper' is going by a different definition of FPR:

https://assets.publishing.service.g...9_Impact_of_false_positives_and_negatives.pdf

Either way, though, the number of positive results during July was still way lower than 2% of all tests done (i.e. it was around 0.6% IIRC). So even if every single positive was a false positive, the 'operational FPR' cannot be greater than this number (unless the entire country suddenly switched to the same, less reliable tests).
Click to expand...

Their definition demonstrated by "100,664 tests ... If the operational false positive rate was 0.4%, 400 of the 1,570 positive tests would be false positives" seems to be what I used in my mathy post, except that essentially I was using net FPR (false positive ratio minus false negative ratio) and taking it as far towards the offered "1%" as possible.

Touring Mars
The point is, though, that no matter which way you slice it or define FPR, the argument that the 'second waves' currently underway across Europe are somehow not real and/or are simply artefacts relating to unreliable test results is nonsense.
Click to expand...

Indeed. I've added percentage positive tests comparison to my post as well to cover the "increased testing" argument, and as you'd expect that still comes out as a notable increase in true cases (1340% if net FPR is 0.6%, or 340% if net FPR is 0%).

It is possible that net FPR is negative with more false negatives than positives, which would mean the % increase isn't as high. However, that would mean there are more true cases than reported, and we didn't get it down as far as we thought during July.
 
Last edited:
x3ra
that's not what the govt website says.
Click to expand...
giphy.gif
Touring Mars
Yeh, I think the 'paper' is going by a different definition of FPR:

https://assets.publishing.service.g...9_Impact_of_false_positives_and_negatives.pdf

Either way, though, the number of positive results during July was still way lower than 2% of all tests done (i.e. it was around 0.6% IIRC). So even if every single positive was a false positive, the 'operational FPR' cannot be greater than this number (unless the entire country suddenly switched to the same, less reliable tests).
Click to expand...
It appears to be using something it terms "operational false positive rate" to mean "proportion of all tests conducted that are incorrectly positive", and the numbers used are insane.

Firstly the paper cites a sensitivity of 95%, which means that no more than 5% of the positives are falsely positive. With 1,570 positive tests, that gives a ceiling of 79 false positives and 1,491 true positives.

To then posit that the "operational false positive rate" could be* 0.4% of 100,664 tests, or 402 falsely positive tests, is incomprehensible. These are data points that are almost an order of magnitude apart - and gives the test a sensitivity of 74% - and the latter guess suggests that a third of all conducted tests are subject to contamination.

It's odd that they don't make a similar guess for false negatives, however at a cited specificity of 95%, that suggests that there are 4,955 people tested as negative who are actually positive (and possibly more, for the same reasons given for false negatives... though I doubt six times as many, as with their false positive guess)...


... and that's worse. A lot worse.


And it is a stab in the dark number with no basis elsewhere in the paper
 
Famine
It appears to be using something it terms "operational false positive rate" to mean "proportion of all tests conducted that are incorrectly positive" ...
Click to expand...

Pretty much all the definitions I can find say that FPR is the ratio of false positives to true negatives, the inverse of specificity (and FNR is the ratio of false negatives to true positives, inverse of sensitivity).

When positive results are around 1% of tests, using all tests instead of true negatives doesn't give significantly different results for FPR calcs.

Given that, an FPR of 0.4% would indicate a specificity of 99.6%, not a sensitivity of 74%, no?

(Not saying these are realistic numbers, just trying to get the terms straight).
 
Last edited:
I'm now a 'COVID Safety Supervisor' because the guy who was supposed to be the supervisor today has had to go home and begin two weeks of self-isolating because his wife has been contacted by the Track and Trace people :ill:

I'm not sure what I have to do other than to be in the building between 9-5pm, and be an emergency point of contact... though I've no idea what that actually involves!
 
Touring Mars
I'm now a 'COVID Safety Supervisor' because the guy who was supposed to be the supervisor today has had to go home and begin two weeks of self-isolating because his wife has been contacted by the Track and Trace people :ill:

I'm not sure what I have to do other than to be in the building between 9-5pm, and be an emergency point of contact... though I've no idea what that actually involves!
Click to expand...

I hope you get one of these:

images
 
Touring Mars
I'm now a 'COVID Safety Supervisor' because the guy who was supposed to be the supervisor today has had to go home and begin two weeks of self-isolating because his wife has been contacted by the Track and Trace people :ill:

I'm not sure what I have to do other than to be in the building between 9-5pm, and be an emergency point of contact... though I've no idea what that actually involves!
Click to expand...
Just do like security and sit at the desk watching your phone till someone comes up and bothers you. ;)
 
Just had a call to say that another member of staff is now isolating, so I've got their supervisor slots now too...

I think I can see where this is going....
 
It would be interesting to find out where the outbreaks are.@Danoff

Wonder if they follow the Gompertz curve ? Most places do.

In Europe it's basically run it's course. You will of course get the odd hiccup, but it's basically gone.

Ef3hBWbUEAAWXVK
 
Touring Mars
Just had a call to say that another member of staff is now isolating, so I've got their supervisor slots now too...

I think I can see where this is going....
Click to expand...
With great power comes great responsibility..... or do you just have the responsibility and zero power?
 
x3ra
In Europe it's basically run it's course. You will of course get the odd hiccup, but it's basically gone.

Ef3hBWbUEAAWXVK
Click to expand...

Do you honestly think we won't notice that that chart is more than a month out of date? Really? :banghead:

I suppose I shouldn't be surprised since it's been about that long that you've been stuck in this loop of... "basically over" ... "increased testing" ... "all false positives" ... "basically over" .................................................
 
Danoff
Colorado cases are taking off right now. Headed for a 3rd peak. Looks like the US is on the rebound too.
Click to expand...
I honestly think people are getting tired of the whole situation and relaxing a bit on the "rules". Every day I see people either not wearing masks, or not wearing them properly.
 
R1600Turbo
I honestly think people are getting tired of the whole situation and relaxing a bit on the "rules". Every day I see people either not wearing masks, or not wearing them properly.
Click to expand...

I was watching the NFL Redzone channel yesterday and I would say less than half of the coaches were like you say, not wearing them properly if they had them on at all. And that's after the league sent out a memo last week advising them to properly wear their masks. At least Andy Reid's face shield didn't fog up this time.
 
Jezza819
I was watching the NFL Redzone channel yesterday and I would say less than half of the coaches were like you say, not wearing them properly if they had them on at all. And that's after the league sent out a memo last week advising them to properly wear their masks. At least Andy Reid's face shield didn't fog up this time.
Click to expand...
As far as I've seen, I think coaches have been allowed a lot more leeway with not always wearing masks, like the players. While I don't watch the NFL, in the NBA, the head coaches usually have no masks or take off their masks to yell plays. I assume it may be a similar, acceptable exception made in the NFL for easier communication.
 
Jezza819
I was watching the NFL Redzone channel yesterday and I would say less than half of the coaches were like you say, not wearing them properly if they had them on at all. And that's after the league sent out a memo last week advising them to properly wear their masks. At least Andy Reid's face shield didn't fog up this time.
Click to expand...

McLaren
As far as I've seen, I think coaches have been allowed a lot more leeway with not always wearing masks, like the players. While I don't watch the NFL, in the NBA, the head coaches usually have no masks or take off their masks to yell plays. I assume it may be a similar, acceptable exception made in the NFL for easier communication.
Click to expand...

I just assume they all get tested every 3 days. I mean if 17,000 Chiefs fans can get tested to attend a game, surely the NFL is doing this. Right?
 
Joey D
Most major countries in Europe are experiencing a surge in cases right now. I don't think it's over, or even close to being over. You will see waves like this until a vaccine comes out.
Click to expand...

The continued use of dramatic and misleading language used by the UK government and the media is becoming ingrained in peoples minds... a 'surge' would indicate we're seeing the situation returning to March/April levels, but this is far from the case.. it's not a surge, it's not even a wave, though it might just be a ripple...

Screenshot-2020-09-20-at-02.58.17.png


This relatively tiny increase in cases is easily explained by the natural increase in prevalence of flu like illnesses seen every year as we go in to autumn and winter.

Deaths 'with covid19' are practically irrelevant - 11 in the UK yesterday and mentioned in 1% of all deaths in the UK for w/e Sept 11.

Even in Spain and France, which are quoted as being 3 weeks ahead of the UK in this 'terrifying 2nd wave', deaths are nowhere near March/April levels... and despite the terrifying messages in the media, only 9% of Spanish hospital beds are occupied by Covid19 patients, with only 1,400 in ICU.

gk6aHdb.png

(Deaths in France are significantly lower than than Spain)

Despite quoting what's happening in Spain and France as one of their reference points, the charts presented yesterday by the UK governments medial advisors painted a significantly more dramatic potential outcome for the UK.... creating more fear in the population.

The data does not support their actions - Boris, Whitty and Handcock can't (or won't) even answer basic questions on false positive tests... they are either too stupid to do the basic maths, or they are choosing to avoid an answer as they know it doesn't support their position.

They are already responsible for thousands of deaths back in March and April, following their decision to empty hospitals of old people and send them to nursing homes, taking Coivd19 with them to slaughter other old and vulnerable people. They are responsible for the current ineptitude of the NHS, which has become totally focused on Covid19 and is not carrying out critical work as a result. Their decisions are politically based. They have no strategy for dealing with this - there's no end point, just hanging on to the vain hope a vaccine comes quickly... and that any side effects from a vaccine that clearly won't have been properly tested will land after their period in government is over.

The impact of the restrictions on the operation of the NHS, businesses and the economy will be far more destructive to society over the medium/long term than Covid19... unsustainable National debt (£2tn and counting), massive unemployment, bankruptcies, massive back log of critical testing (cancer screening, for example) and treatment in the NHS which will lead to tens of thousands of premature deaths... all this without considering the longer term change to they way we live, and the destruction of our civil liberties and proper democratic process.

Protect the vulnerable (the old and those with existing serious contributing conditions) but let the rest of the population get back to life as normal... the current approach is like using a hammer to try and kill a fly.
 
And now we're in for this for another six months. That's a death sentence for hospitality and the arts.
 

Latest Posts

Back