Coronavirus in the United States - news and thoughts

We all relate or share our own experiences, and unfortunately there are folks like I describe everywhere, from the clubhouse to the outhouse... half the houses I see with a Trump flag in the yard also have a confederate flag accompanying it. Now, as you point out, that is only in places I've actually been lately, and that'd encompass from about Macon, Ga. over to the coast and down to about Lakewood Ranch. Oh, and what I see all over the country every day. Regardless, I get along with most everyone because actually converting people kind of seems above my pay grade... my observations are hardly unique.

 

We already knew about the recruitment method. Stanford told us that in advance. Their data still shows that only 5% of their population has been infected. Again. Read the Stanford report. They told us the demo skewed White Female. We knew that. It was not a "flaw". it was a known.

It is also an "open study" meaning more work is being done.

Seriously, every single hint of good news gets people upset. It is a weird thing to me.
Science should be our guide...but only if it brings us bad news.

(That is a general statement of this board lately and not aimed at you specifically Davis.)

I mean, Stat geeks on the west coast are being questioned by Stat geeks on the East coast, and apparently "number nerds taking to Twitter" (as per the article) but the West Coast geeks have good news, so the East Coast geeks must be right.

None of us know, so why don't we stop acting like we do?

 

"We report the prevalence of antibodies to SARSCoV-2 in a sample of 3,330 people, adjusting for zip code, sex, and race/ethnicity. We also adjust for test performance characteristics using 3 different estimates: (i) the test manufacturer’s data, (ii) a sample of 37 positive and 30 negative controls tested at Stanford, and (iii) a combination of both."

 

Unless they adjust for the endogenous selection, which is not a demographic issue and woud be extremely difficult to do without a better sample, it is an inappropriate sample on which to try to extrapolate. The fact that it is "good news" should not lower the threshold.

 

Nobody is complaining about the demographic issues. People are complaining about the convenience sample.

Unless they are completely changing their sample, that is unimportant. If they are, they shouldn't have reported this data as it is misleading.

It isn't because the news is good. The issue is that you seem to want good news enough to overlook the fact that the good news is not on solid statistical footing, meaning that it isn't news at all. If I told you that we have a zero percent infection rate because my house has no infections, you should rightfully dismiss that due to statistical issues. Not sure why you won't do this with such a big statistical issue here.

Many of us have enough background in statistics to be able to state the issues with the statistics here. If you ask me about the scientific basis of the actual test, I would be lost. However, I have background in statistics and know that this is not an appropriate sample on which to extrapolate onto any population other than "people likely to opt-in on Facebook to the study in Santa Clara." Trying to extrapolate the findings onto Santa Clara is wrong and I have very little doubt that the authors knew that, as this is undergraduate statistics stuff, which is why I question their ethics.

 

1. I havent overlooked anything. I read the report the day it was issued. I knew the limits of the report because the Stanford release spoke clearly about it.

2. It is some good news. It clearly tells us there is a higher infection rate...the number may be less than 85% or whatever...but a higher infection rate means lower mortality rate and more hopeful immunity.

3. And demographics have been an issue. Many complained because it focused on white women. The study claims to have adjusted for that.

 

Again, it is a limitation to say that you shouldn't extrapolate beyond Santa Clara. It is a flaw to extrapolate the population from which the sample was drawn, people willing to opt-in on Facebook in Santa Clara to the entirety of Santa Clara.

It doesn't tell us anything because the sample is not appropriate in which to make inferences about a greater population of people. Here is a statistically appropriate statement: "It tells us that there is a higher infection rate amongst those who opt-in to their study on Facebook in Santa Clara County, CA." However, any statement beyond that has no scientific or statistical basis. If you wish to argue that there is external validity, provide me the population on which you think it is appropriate to extrapolate from this study and we can discuss that.

 

Have you ever conducted research that called for sampling, theoretical modeling/study design integrity, and statistical analysis? There are several on board here who have. There are also several research gurus in the scientific community pointing out the flaws in this study. You continue to espouse the dubious claim that the study was not flawed despite facts and support to the contrary. And no - good news doesn't get people who question this study upset; it's a matter of irresponsible research. We're not all hoping for more bad news. We're hoping for good news that is reliable and valid.

As for flaws - here is one of the sources cited in the article I previously posted: Concerns with that Stanford study of coronavirus prevalence « Statistical Modeling, Causal Inference, and Social Science

Now before you jump on the term "potentially," understand that it is being ascribed to the term fatal, not flaws, as in "these flaws" are potentially fatal for the study."
In Pursuit of Real Coronavirus Numbers

 

https://www.washingtonpost.com/poli...61b968-7e89-11ea-8de7-9fdff6d5d83e_story.html

 

*sigh*

 

He has it backwards. We need to get access in exchange for grants. Predict have us access to the lab that there is circumstantial evidence that this virus came out of. That is the problem with his innate viewpoint that he doesn’t want to pay for anything

 

Go GATORS!
,WESGATORS

 

I have something that might just wake people up after all the boring discussions about statistics. There is a new symptom for children with coronavirus: pustules on the toes.

What are 'COVID toes' and why are dermatologists seeing an 'epidemic' of them in kids?

Yikes.

 

I cited a reported case early on where a heart attack victim, died from his second heart attack in a week, and he had multiple previous attacks, but was listed as a covid-19 death. Are the heart attacks being lost in the jumble? @mutz87 pointed out early on that the data sometimes has to be sorted later. Will we find that heart attacks were happening at normal levels, but they were categorized as corona deaths in haste?

 

Can the data not overlap? In other words, do doctors have to pick one or the other when identifying the cause of death?

Go GATORS!
,WESGATORS

 

This might seem small, but think about this with an open mind. Trump put together his Opening Up America Again congressional group. Every single Republican Senator is on it but one. At his Sunday evening press conference he was asked if it was because he held a grudge.

See his reply below. Trump thinks Coronavirus response is a good time and place to pay back a grudge. How can anyone think he puts the country above himself when he admits being a small, petty grudge holder is impacting who he includes in the effort against COVID-19.

There are more outrageous things he’s done as President. But, that’s about as crystal clear an acknowledgment of self over country as you can get.

“I’m not a fan of Mitt Romney,” he said. “You know, I have 52 Republican senators — ”


“He was a governor. You don’t want his advice?”



“Well, I just don’t think — you know, I’m not a fan of Mitt Romney. I don’t really want his advice,” Trump said.

 

Yikes is right. Looks painful but maybe not?

 

Only the best...smdh..but that read sounds like the character of most people T promotes.

 

Not sure.

Let's say a person is in a bad car wreck but has a heart attack as a result. The COD may be listed as a heart attack rather than a car wreck.

I have heard of cases where a shooting victim was found in the ocean. The death can be listed as a drowning if the lungs took on water. Even though the gunshot was the primary contributor.

I don't know if it is uniform by M.E. and Dr...or is it astate requirement? Not sure.

I definately feel that post pandemic due diligence will change many classifications.