Covid 19 outbreak - arguing about Stats (again)

Use this board for general non-cycling-related chat, or to introduce yourself to the forum.
mjal
Posts: 28
Joined: 4 Jul 2011, 2:22pm

Re: Covid 19 outbreak - arguing about Stats (again)

Postby mjal » 24 Nov 2020, 3:56pm

Jdsk wrote:
mjal wrote:Your quotation implies that the words related are mine ; they are NOT and neither are the figures. They are part of a quote from a comment below the article in Nature ; the said quote was clearly delineated by me.

I'm sorry if I misquoted you. But I've just looked again and the section above that is in quotation marks. The section below has closing but opening question marks. That section has no quotation marks:

...

Here is a quote from the comments below the Nature paper :

""There were no positive tests amongst 1,174 close contacts of asymptomatic cases."

How is this possible? Even if all those contacts were indeed uninfected, where are the false positive tests? Even assuming just a 1% false positive rate, the chance of getting zero positives out of 1,174 is less than 1 in 100,000. In the "source data" section, there's an even more improbable claim: zero positive results out of 85,884 tests conducted in East Lake Scenic Area of Wuhan.

Isn't this clear evidence that the data being used is unreliable?"

...

That's why I thought that it was you speaking.

Jonathan


Jonathan,

Not pedantry (honestly) just precision...

The original comment below the article has quotation marks around the first sentence i.e. "There were no..." ; the commenter is quoting from the article. I add my marks so double marks at the beginning. The commenter then continues...How is this possible? No quote marks. He finishes speaking at...unreliable? No quotes. I add my marks.

Double at the beginning of my quote because of existing marks, single at the end ; everything in between my marks is not my speech. I did consider altering the paragraphing to try to clarify the extent of the quotation but decided against it.

Could/should quotation marks be routinely in bold in these circumstances? I also considered trying to show the quote in the usual forum panel but that would make it look as though that text was in a post...

Mike.

Jdsk
Posts: 5281
Joined: 5 Mar 2019, 5:42pm

Re: Covid 19 outbreak - arguing about Stats (again)

Postby Jdsk » 24 Nov 2020, 4:33pm

Thanks, I understand now.

One way of removing any ambiguity in this forum is to use the unattributed quote tags. But I suspect that some others might only think that appropriate for quotes of stuff from this forum.

My preference is to use both quotation marks and italics.

But the traditional solution was to put quotation marks at the start of each quoted paragraph, sometimes with another et at the end of each paragraph and sometime only at the end of the final paragraph.
https://en.wikipedia.org/wiki/Quotation_marks_in_English#Quotations_and_speech

Jonathan

PS: I've nothing against precision. : - )

mjal
Posts: 28
Joined: 4 Jul 2011, 2:22pm

Re: Covid 19 outbreak - arguing about Stats (again)

Postby mjal » 25 Nov 2020, 10:00pm

Jdsk wrote:
mjal wrote:I did muse about FPR in my long post and made the point that some virologists maintain that there are no "false positives" in the conventional sense and that any generated are due to contamination, clerical error etc. I then went on to say that I had seen a rate of 0.5% offered. I made no claim as to the accuracy of this figure.

Do you now accept Spiegelhalter's logic that the FPR in the UK this summer must have been much lower than this?

Jonathan


Apologies for delay in replying...health appointments am, long coastal walk in pm.

If the total positive rate is (very) low then the FPR must be (much) lower ; logical, no disagreement but...

I confess I did not watch the whole video ; does he mention false negatives at any point? If there was a significant number of these then the FPR could be higher...just musing.

https://www.nejm.org/doi/full/10.1056/NEJMp2015897

Quote from article:

"Antibody seroconversion was observed in 93%. RT-PCR testing of respiratory samples taken on days 1 through 7 of hospitalization were SARS-CoV-2–positive in at least one sample from 67% of patients."


If we assume that the antibody results are confirmatory of infection then the positive PCR figures are surprisingly low.

As ever, there is no clear answer to any of this but it is interesting...

Mike.

Jdsk
Posts: 5281
Joined: 5 Mar 2019, 5:42pm

Re: Covid 19 outbreak - arguing about Stats (again)

Postby Jdsk » 25 Nov 2020, 10:11pm

mjal wrote:... does he mention false negatives at any point? If there was a significant number of these then the FPR could be higher...

I don't understand that bit. What are you keeping constant and what are you varying... specificity, sensitivity, prevalence?

Jonathan

Psamathe
Posts: 12023
Joined: 10 Jan 2014, 8:56pm

Re: Covid 19 outbreak - arguing about Stats (again)

Postby Psamathe » 26 Nov 2020, 11:54am

Now we're all moving into a new Tiers System though I'd check what tier I'll be in and apparently I'm in "We're experiencing technical difficulties.". Fine.

But consistent with everything changing frequently, a subsequent check put me in "Sorry, we're experiencing technical difficulties".

So over a short period of time I've changed so I'm now with with added "Sorry". At least that makes it clear?

Ian

Jdsk
Posts: 5281
Joined: 5 Mar 2019, 5:42pm

Re: Covid 19 outbreak - arguing about Stats (again)

Postby Jdsk » 26 Nov 2020, 11:59am

But at least it isn't "I'm sorry that you feel that you're experiencing technical difficulties."!

; - )

Jonathan

User avatar
Mick F
Spambuster
Posts: 50687
Joined: 7 Jan 2007, 11:24am
Location: Tamar Valley, Cornwall

Re: Covid 19 outbreak - arguing about Stats (again)

Postby Mick F » 26 Nov 2020, 12:32pm

Just been looking at the list.
https://www.gov.uk/guidance/full-list-o ... rs-by-area

We in Cornwall are going into Tier 1.
Living here in the Tamar Valley, we can see Tier 2 Devon from our kitchen window.
Our local pubs will be opening on Wednesday, but the pubs in Tavistock will be closed ................ so where are the Tavistock pub-lovers going to go? Gunnislake! :shock:

Launceston is on the border of Devon, so they are going to get Devonians coming across too. I was hearing on the news a week or three back about the differences of Scotland and Wales and how people live in one place, and work in another. They were profiling Hereford as a case in point.

A very odd situation is at hand.
Mick F. Cornwall

Oldjohnw
Posts: 5805
Joined: 16 Oct 2018, 4:23am
Location: Northumberland

Re: Covid 19 outbreak - arguing about Stats (again)

Postby Oldjohnw » 26 Nov 2020, 12:33pm

Jdsk wrote:But at least it isn't "I'm sorry that you feel that you're experiencing technical difficulties."!

; - )

Jonathan


:D
Or "I've know Government websites for many years and they are always kind and helpful."
John

Psamathe
Posts: 12023
Joined: 10 Jan 2014, 8:56pm

Re: Covid 19 outbreak - arguing about Stats (again)

Postby Psamathe » 26 Nov 2020, 12:33pm

Jdsk wrote:But at least it isn't "I'm sorry that you feel that you're experiencing technical difficulties."!

; - )

Jonathan

And at least now it isn't changing. I've been in "Sorry, we're experiencing technical difficulties" for sometime now so things are no longer changing by the minute as rules & regulations have been doing since March.

Ian

Oldjohnw
Posts: 5805
Joined: 16 Oct 2018, 4:23am
Location: Northumberland

Re: Covid 19 outbreak - arguing about Stats (again)

Postby Oldjohnw » 26 Nov 2020, 12:37pm

We are having world beating technical difficulties. It is game changing. We will bash accuracy and efficiency. And get our non-technical mates to sort it at tremendous cost.
John

sjs
Posts: 931
Joined: 24 Jan 2010, 10:08pm
Location: Hitchin

Re: Covid 19 outbreak - arguing about Stats (again)

Postby sjs » 26 Nov 2020, 1:25pm

Oldjohnw wrote:We are having world beating technical difficulties. It is game changing. We will bash accuracy and efficiency. And get our non-technical mates to sort it at tremendous cost.


Probably Excel 97 is involved somewhere.

Jdsk
Posts: 5281
Joined: 5 Mar 2019, 5:42pm

Re: Covid 19 outbreak - arguing about Stats (again)

Postby Jdsk » 26 Nov 2020, 1:41pm

sjs wrote:
Oldjohnw wrote:We are having world beating technical difficulties. It is game changing. We will bash accuracy and efficiency. And get our non-technical mates to sort it at tremendous cost.

Probably Excel 97 is involved somewhere.

"Does contact tracing work? Quasi-experimental evidence from an Excel error in England"
https://warwick.ac.uk/fac/soc/economics/research/centres/cage/manage/publications/wp.521.2020.pdf

and press release:
https://warwick.ac.uk/fac/soc/economics/research/centres/cage/news/23-11-20-nhs_test_and_trace_works_new_research_finds_but_the_data_glitch_which_led_to_over_15000_late_referrals_caused_more_than_125000_additional_covid_19_infections

Jonathan

Psamathe
Posts: 12023
Joined: 10 Jan 2014, 8:56pm

Re: Covid 19 outbreak - arguing about Stats (again)

Postby Psamathe » 26 Nov 2020, 2:00pm

Jdsk wrote:
sjs wrote:
Oldjohnw wrote:We are having world beating technical difficulties. It is game changing. We will bash accuracy and efficiency. And get our non-technical mates to sort it at tremendous cost.

Probably Excel 97 is involved somewhere.

"Does contact tracing work? Quasi-experimental evidence from an Excel error in England"
https://warwick.ac.uk/fac/soc/economics/research/centres/cage/manage/publications/wp.521.2020.pdf

and press release:
https://warwick.ac.uk/fac/soc/economics/research/centres/cage/news/23-11-20-nhs_test_and_trace_works_new_research_finds_but_the_data_glitch_which_led_to_over_15000_late_referrals_caused_more_than_125000_additional_covid_19_infections

Jonathan

That is quite a shocking additional cases and deaths. That many additional cases with already sub-standard contact tracing sounds like a significant contribution to spread.

Don't these organisations run any sort of QA or double checking system. I appreciate the systems were thrown together in a mad rush but this Excel issue happened relatively recently quite some time after action on the pandemic was started.

Ian

Jdsk
Posts: 5281
Joined: 5 Mar 2019, 5:42pm

Re: Covid 19 outbreak - arguing about Stats (again)

Postby Jdsk » 26 Nov 2020, 2:15pm

I work in this area. A whole lot of available and offered expertise was not taken up. The model of data transfer and the absence of checking and fail-safe engineering staggered everyone.

Jonathan

Vorpal
Moderator
Posts: 18652
Joined: 19 Jan 2009, 3:34pm
Location: Not there ;)

Re: Covid 19 outbreak - arguing about Stats (again)

Postby Vorpal » 26 Nov 2020, 2:18pm

Data Glitch? I would call that blatant negligence.

Practically anyone who's ever had anything to do with large amounts of data could have told them that a spreadsheet was a crappy way to store data.
“In some ways, it is easier to be a dissident, for then one is without responsibility.”
― Nelson Mandela, Long Walk to Freedom