COVID19 Factors We Should Consider/Current Events

[Shiva Kaul]

Noah, while you "generously" draw whatever conclusions you desire, the authors are, in fact, fixing their flawed paper. (As mentioned in that Twitter thread, and also Science magazine.)

The field of statistics isn't broken. Your ability to discern good statistical arguments from bad ones - that is definitely broken.

[Buddy Rich]

Hot Mic

YouTube

[Mark Rippetoe]

I spent far too much time finding those links and then arranging my thoughts for that post. I don't think that he'll deliver a well thought out reply.

That's physically impossible.

[Noah Ebner]

Noah, while you "generously" draw whatever conclusions you desire, the authors are, in fact, fixing their flawed paper. (As mentioned in that Twitter thread, and also Science magazine.)

From the Science article above: "Bhattacharya says he is preparing an appendix that addresses the criticisms. But, he says, 'The argument that the test is not specific enough to detect real positives is deeply flawed.'

They aren't "fixing" shit, they're addressing the criticisms to tell you stats wonks why what you think is a big deal actually isn't. BIG difference.

Also from the same article above:

Bhattacharya and Bendavid have also collaborated with Neeraj Sood, a health policy expert at the University of Southern California, to do a similar study in Los Angeles county. They used the same antibody test on 846 people selected by a marketing firm to represent the county’s demographics. In a press release issued this week, they estimated that roughly 4% of the county’s adult population has antibodies to the virus—as many as 300,000 people. (Sood told Science that 35 subjects tested positive.)

EXACT SAME RESULTS. I'm still right and you're still wrong. Rip's smart enough to see through your antics w/r/t this particular question, but please stop trying to confuse people with 'fancy' statistics. It's not a good look.

[mangen]

Attachment 7291

George,

Thank you for the link to the peer reviewed paper containing a statistical confidence interval that is practically significant. Seems we should all get behind Sweden and follow her lead.

Regards - Mike

[Nockian]

All-cause excess mortality does indeed seem to be higher right now:

25,000 Missing Deaths: Tracking the True Toll of the Coronavirus Crisis - The New York Times

I see figures like the above, and then the lockdowns seem reasonable. Then I look at unemployment numbers, and can't help but wonder if the trade-off is worth it.

What's a principled way to evaluate each alternative?

The principle is that man is the rational animal and is capable of determining and choosing his own degree of risk taking. Not only can he choose, but that he must be free in order to choose. Using force against people to keep them in semi-incarceration prevents them from making rational decisions about their lives and hence is a direct threat to their lives. Covid is no different to any other kind of risk factor in life. The Governments task here is to test, trace, track and isolate those who have the virus-the blunt force of isolating everyone is neither moral nor practical.

[Mark Rippetoe]

Mike,

Sometimes things can just be funny, without peer-review.

Rip

[Noah Ebner]

George,

Thank you for the link to the peer reviewed paper containing a statistical confidence interval that is practically significant. Seems we should all get behind Sweden and follow her lead.

Regards - Mike

Personally, I'd LOVE to get behind Sweden...

[Shiva Kaul]

They aren't "fixing" shit, they're addressing the criticisms to tell you stats wonks why what you think is a big deal actually isn't. BIG difference.

Folks: let us revisit Noah’s beliefs after the revised analysis is posted.

EXACT SAME RESULTS.

The conclusions of one paper do not affect the quality of the arguments in another.

'fancy'

‘Rigor’.

[spacediver]

What is this claim based upon?

I am fairly certain that Prof. Johan Giesecke said that there were mandatory restrictions on nursing homes.

But the idea that fully functioning adults of an advanced age should somehow not be allowed to make their own risk evaluations is crazy.

Attachment 7291

If you watch the unherd interview, Johan Giesecke laments that they didn't do enough to protect the nursing homes.