Posted on 04/24/2020 2:26:02 PM PDT by Bruiser 10
“A Chinese antibody test had a 1.4% error rate; I assume the Santa Clara study was using a test with a similar error rate. Therefore, the finding that 1.5% of the people in Santa Clara tested positive is within the margin of error.”
That’s not how the math works. If the SC study had the same error rate, it’s 0.015 (of people) x 0.015 (test error), or 0.000225 for the max error, either side. Multiply by population for estimated actual numbers.
[Then you add CI based on sample size]
“The necessary conclusions are not acceptable to politicians who now have near absolute control over the population or to pharmaceutical companies whose alumni staff the CDC and the the FDA.”
Yes. This is criminal, what they are doing. They need to pay, dearly, for what they are doing to our country.
Actually, it was flawed. Applying heavy duty statistics does not change the fact that the positive rate was right there in the margin of error—making the entire study meaningless. And then taking that result and extrapolating it to “find” the seropositivity rate is even higher than they measured? GIGO.
And well designed and executed? You have got to be kidding me.
“Let’s say .01% of a population has had CV and I do a random test of 100,000 with a 1% false positive and 1% false negative rate.
I will find that 1000 people had CV when in fact only 100 did.”
Good God, your math is bad. And your statistical knowledge. First, .01% is 0.0001. Times 100,000 is 10.
But if you had 100 positive out of 100,000, and the error rate was 10%, 10% of 100 is 10. So 100±10 is your likely range. Now add CI to it, based on sample size.
Seriously.
“CDC are the ones who are distorting the numbers.
They are ordering Coroners nationwide to list Corona as the cause of death even if people are killed in car accidents, along with a whole list of other causes, IF they had the Corona virus when they were killed. Car accident victims.”
Yes. They did this when it became apparent that Covid deaths would be close to flu deaths. They made everyone start cooking the books. And gave bounties(!) for Covid classified deaths. I.e., hospitals get reimbursed more for a pneumonia patient who dies, when they add Covid diagnostic code.
The CRC knew it weeks ago, when the predicted numbers didn’t come close to what was happening on the ground. And now they’re pushing “2nd wave!!!” to keep the panic going and discourage lifting the lockdowns.
This is criminal, and needs to STOP!!!!
“Actually, it was flawed. Applying heavy duty statistics does not change the fact that the positive rate was right there in the margin of error—making the entire study meaningless. And then taking that result and extrapolating it to “find” the seropositivity rate is even higher than they measured? GIGO.”
You are wrong. Error rate applies to the measurement method, not the results. Please understand that the error rate is multiplied by the number of samples, not the final predicted % who have it. You are mixing up population with sample data. No can do.
1.5% error rate is actually extremely good. For any measurement method. 50% would be quite bad. In Medical Device Validations, which I know quite well, 10% error rate is generally acceptable. That’s for implantable devices, like artificial heart pumps, when failure can mean death.
Yes, I’m fairly proficient at statistics, sample sizes, error measurement, etc. In fact was accepted as an expert witness in the subject in the DC appeals court.
Your making a basic mistake, seriously.
And the SC test has been widely reviewed by many, many experts, worldwide. It’s good. Likewise similar tests with recent released results. Pretty much every test they do now, randomly sampled, says the same thing. They are coming out each day. Vast undercounts in previously infected.
Heck, we knew it when the data from the Diamond Princess came out.
We’ve been played. Covid is not that bad.
No, my friend. This virus doesn't concern or frighten me as it has done to so many; but what it has done to the American psyche, what it has done to our society and, of significant concern to me, what it has done to American men is quite disturbing and not a little, well, sad. As I have previously noted, if the Chinese released this as a biological weapon, they didn't do so well. But if it was intended to be a psychological weapon, it was applied brilliantly ... the Chinese appear to know us better than we know ourselves.
Many thanks for this enlightening and entertaining repartee, my FRiend in the truly spectacular state of Tennessee. Let's keep it going as the opportunities arise. Cheers! from the Sierra foothills, part of California's shrinking conservative oasis.
To be frank I believe this is a beta test, not some accidental release.
Nice try, AH. 100,000 x 1 percent is 1000.
You test 100,000 people who don’t have CV with a test that has a 1 percent false positive rate and you get 1000 false positives.
I think you are mistaking the type of error going on here. This is not an error of magnitude of measurement, but a type 1 error—detecting a positive result when it is really negative. This has nothing to do with the actual measurements. The fact is, they are detecting 1.5% positive results, when the false positive rate of the assay is 1.4%. In other words, the 1.5% positive results are indistinguishable from the 1.4% positive results one would expect from the intrinsic error rate of the assay. If you were to take blood samples from summer 5 years ago—samples that cannot possibly represent people exposed to Covid-19—and assay them, you would not be able to distinguish between those and the samples measured in the Stanford study (which was NOT well designed) because you would still see around 1.4% positive results (type 1 error). Now, if I were to take several groups of blood samples and run them through the immunoassay, I could then quantitate a false positive error rate, which would hover somewhere around 1.4%, but would vary between sample sets.
With this kind of testing, the only way you can be certain that you are getting real results is if the number of positive results rises above background.
I highly suspect that many of the false positives result from the fact that antibodies tend to be cross-reactive—they recognize antigens other than the antigen they are supposed to recognize. People get colds every winter, and some of them are caused by coronaviruses. Those antibodies could very well recognize the Covid-19 antigen used in the immunoassay, simply because coronaviruses have certain structural similarities.
First thing you have to know is whether you’re dealing with the benign strain or the potentially lethal strain, or as is most likely, some combination of the two.
The CDC tested street bums in San Francrapsco and 66% were positive. They were in shelters and not hospitalized, and many of them had comorbidities. Obviously that population is not colonized by the strain that’s predominant in New York and New Jersey.
The benign strain is the vaccine, I believe. When a bio-weapon is made, the counter to it must be developed concurrently or else you have no way to make it selective for your enemy.
Of course he does!
He has what you would call divine OCD!
We have to remember that hes the one who will not let a bird followed the ground without his permission and tracks on a continuous basis the number of hairs on my head and everyone else’s for that matter.
He cares hes watching hes keeping track and he has a plan.
He actually had a plan before we even knew we had a problem and that applies to every problem that we have especially the ones that we know nothing about.
That is what many of this nation’s original founders believed and stake their lives on.
They believe that this Almighty God vested us with a certain level of free will And personal liberty to govern our own lives without fear trusting in him to preserve us and trusting ourselves to act in accordance with whatever wisdom we had along with faith in his faithfullness.
There’s only one being that is the originator of lies and who derives great joy pleasure and power from the spreading of lies and misinformation. That being is the moron and full of the universe but he dominates everyone with his lies and attempts to dominate each and every 1 of us.
this is a demonstration of his power to deceive and confuse and it’s coming to a halt sooner rather than later.
Your comments on what happens when governments allow the the Chinese Commies to come into their country brings to mind what happened to Italy.
Our approach to deal with this plague was highly influenced by the death and sickness due to the CCWV ( Chinese Communist Wuhan Virus) reports coming out of Italy.
Most Americans including this one were not aware that that the Italian socialist government had actually allowed the Chinese communists to purchase thus remove Italian ownership of a large segment of their fashion industry producers. Then replace the Italians working at those companies taken over by the Chinese and have them manned by thousands of Chinese nationals allowed to be brought in to work there.
When that CCWV breakout occured in Wuhan the politburo closedoff Wuhan but did not warn the Italian Socialist government or the Pope with whom the Chinese government had good relations with but certainly violated when the socialists should have been informed and allowed to deny those Chinese workers travel to and from China particularly for their New Year while this was occuring. Which Trump did with the exception of only allowing American citizens returning..
Those statistics being reported always indicated the casualties were aged Italians no focus was permitted on China or even if the victims were from China and the emphasis was on the speed of contagion. While doing so also ignored the direct connection those producers have with the Chinese Communist government which owns them.
We’re letting them corner the swine market.
If you look at the deaths graph on the last page of NYC’s own statistics
You will notice that, for the last week, the “probable” deaths (no positive COVID test) exceed the “confirmed” deaths (had positive COVID test)
With the expansion of testing ability, NYC has NO EXCUSE for not giving each suspected COVID death a test.
NYC is counting lots of non-COVID deaths as COVID deaths.
We need to be open for business.
Here’s an addendum to my previous reply about the questionable Stanford study. I just found this today; I am not the only person who has extreme reservations about that study.
Apparently, a lot of experts question the study. As I noted before, I doubt it would ever get past peer review.
The antibody tests are wrong and might as well be random, not joking
“Apparently, a lot of experts question the study. As I noted before, I doubt it would ever get past peer review.”
I’m revising my opinion on it, to “maybe good.” I’ve been playing with Bayesian statistics on their sample sizes and alpha/beta error rates. I honestly thought it would support their CI, but, as far as I can tell, it doesn’t. That troubles me, but I am definitely not a Bayesian expert.
I’ve read a couple very good pieces using the more simple (but still difficult!) binomial analysis. Seems the jury is still out. The experts I read, pretty neutral people, said “maybe it’s good.” The error rates, that is. Seems it’s not as simple as using the initial posted +/- error rate. I closed the pages, as it was just for my own curiosity, but if I get time, I’ll see if I can find and post here.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.
