DIYbanter - View Single Post

FromTheRafters

expressed precisely :
On Tue, 30 Jun 2020 20:25:22 -0400, FromTheRafters
wrote:

Bob F used his or her keyboard to write :
On 6/30/2020 12:56 PM, FromTheRafters wrote:
brought next idea :
On Tue, 30 Jun 2020 12:31:38 -0500, Jim Joyce
wrote:

Oklahoma does have a statistic to be proud of, though. Over the past two
days, of the people who have been tested, 100% have been positive.
That's the first time we've seen that.

All that tells me is they are only testing people suspected of having
the virus.

It figures that you would get somethng like that from something not even
saying that.

I had a hard time believing that 100% number when Chris said it. There is
only one hit googling for it, referring to the Chris Hayes report. I
suspect he was punked on this one.

That from the same camp which says there are so many false positive and
false negatives that the whole idea of testing is useless, now believes
that 100 percent 'any result' is even posible.

It defies logic.

No it is logical on several levels.

Okay.

First if the tests were perfect, it is saying everyone in Oklahoma, at
least those in a random sample are really positive ... not at all
likely.

If things were different, then things would be different.

We have evidence that a small population can be positive for antibodies
or particles without having symptoms or have only mild symptoms. That
is, the ones who didn't get sick and/or die. Wasn't there a prison
population with around a thousand in the data set which resulted in a
large percentage of positives?

The larger the data set, the less likely this is to happen. The same
goes for the 'imperfect' tests with a large enough data set.

Second If the tests are flawed, it is ridiculous to assume a random
number generator comes up with 100% of anything unless they always
come back positive. Then it is not a test.

Okay, but that is what I said, you are not disagreeing. 100 percent
from real world testing is unlikely. Jumping to an intentional skewing
by selective testing is not logical when so many other things can cause
it. Not that you are necessarily wrong, but nothing in what he said
indicates bias in testing.

Third They only selected people to test who were likely positive
anyway (not a random sample including asymptomatic cases).

You didn't write "if". Can you show me the biased methodology they used
for selection?

Fourth and most likely considering the source, Rachels male persona
Chris was lying.

Yes, that is another likely reason for the 100 percent figure.