20 Apr 20
@eladar saidI don’t look at NYC data. I look at the state data, of which NYC is the lion’s share anyway. I’m perfectly content to rely on Cuomo’s PowerPoints to know what’s going on in NY. He’s likely to have much better data than the City anyway.
Do you want an explanation of what they are doing or are you satisfied with your present understanding?
20 Apr 20
@sh76 saidCuomo averages the last three days. It means his numbers will be a bit larger than what is actually happening if the numbers are in the decline.
I don’t look at NYC data. I look at the state data, of which NYC is the lion’s share anyway. I’m perfectly content to rely on Cuomo’s PowerPoints to know what’s going on in NY. He’s likely to have much better data than the City anyway.
20 Apr 20
@deepthought saidWhat would you consider validation?
Public Health England attempted to validate a large collection of this type of test and found that they were all useless. Have these tests been validated?
20 Apr 20
@sh76 saidOn some of the tests Public Health England looked for the presence of anti-bodies in blood samples taken last year which could not possibly have anti-bodies to SARS-CoV-2 and the tests returned a positive result. They tested people who had previously tested positive for the presence of the virus and got negative results. Which means that the tests lacked both specificity and sensitivity. Validating tests means looking to see if they get the right results in cases where you know the results with some acceptable rate of error.
What would you consider validation?
The BBC article below mentions that Public Health England has not found a test they regard as adequate and talks about the problems with what anti-body tests actually demonstrate.
https://www.bbc.co.uk/news/uk-52335210
20 Apr 20
@deepthought saidIt’s really a question of what you’re looking to do. If you’re looking to declare a person safe, then sure, precision is critical. But if you’re looking to measure prevalence the. 80 or 90% accuracy will do.
On some of the tests Public Health England looked for the presence of anti-bodies in blood samples taken last year which could not possibly have anti-bodies to SARS-CoV-2 and the tests returned a positive result. They tested people who had previously tested positive for the presence of the virus and got negative results. Which means that the tests lacked both specificit ...[text shortened]... e problems with what anti-body tests actually demonstrate.
https://www.bbc.co.uk/news/uk-52335210