Tags

, , , , , ,

Norm’s note: you can follow Dr RollerGator on Twitter, HERE. I want to draw your attention to two pieces written by the Gator, but I want to first caution you that neither of these pieces is making any claim about the efficacy, or lack thereof, of the mRNA vaccines. Rather, the upshot of the pieces is to demonstrate that the business of statistics can all too easily — if sometimes by design, then also sometimes despite the best and most honest of intentions — obfuscate rather than clarify a situation we are trying to understand. Thus, for example, in the piece titled, Lies, Damned Lies, and Vaccine Statistics, RollerGator makes the following assertion: “The real world data has shown that the death rate among the vaccinated, if infected with COVID, can be 3 to 5.7 times higher1 than the death rate of the unvaccinated.” And on the basis of the real world data being parsed, that is exactly right. However, as RollerGator goes on to show, it would be a mistake to conclude from those numbers that the vaccines are not highly effective. You can, as he demonstrates, in a situation where vaccinations are actually reducing infection rates, and precisely because they are reducing infection rates, have for the vaccinated cohorts a rising “infection fatality rate.” So these interesting reads are not so much about the vaccines as such, as about the potential intractability of the statistics being used to inform decisions, both private and public.

Do, then, see these two companion pieces:

Lies, Damned Lies, and Vaccine Statistics

&

Lies, Damned Lies, Vaccine Statistics, and Eric Topol

Oh, and another piece that caught my eye and that I thought really interesting and that dovetails nicely with the foregoing is this one by Mathew Crawford:

Gell-Mann Amnesia, Dissonance, and Resolution


Update (8/19/2021 @ 9:32 am): In an article written by David Fuller that initially badly represented (and yet continues to misconstrue) the upshot of Dr.  RollerGator Phd’s Lies, Damned Lies, and Vaccine Statistics, the following clarification by the Gator has been included, albeit not until after Fuller had had his feet held to the fire over the course of a week or so for his misrepresentation, and I quote that clarification here on the off chance that my remarks might also have been misleading (although I do not think they miss the mark entirely):

Dr Rollergator has sent me his response to this section, pasted below as I offered.

“My article had two central premises. The first premise is that public communicators, including journalists, credentialed medical professionals, and public facing government officials, regularly fail to maintain an important distinction between COVID vaccines preventing COVID infection and COVID vaccines reducing hospitalization and death rates after infection with COVID.

The second premise is that, at the time, it was claimed that the vaccines had ~97% protection against death after infection. Those numbers were coming from the largest study in the world at the time of 6.5 million people observed meticulously via government surveillance, and only related to preventing infection. Those who never get COVID obviously do not become hospitalized or die with COVID. The same study did not contain enough information to determine if for ages under 65 there existed any reduction in hospitalization or death after infection.

David’s Medium article confirms both premises. The Bad Stats twitter account agreed with both premises explicitly and Gideon Meyerowitz-Katz confirmed that the largest study in the world at the time of 6.5 million people observed meticulously via government surveillance, did not contain enough information to determine if for ages under 65 there existed any significant reduction in hospitalization or death after infection, and therefore could not support the claim of ~97% protection from death if infected.

David demonstrated the first premise by failing to maintain the difference between preventing infection, and reduction in hospitalization or death after infection. David’s examples, including graphs, ER doctors, Virginia data, are results of preventing infection and do not maintain the distinction. My article is not in contradiction with any existing evidence. It simply requires one maintain a significant and important distinction. David’s inability to maintain that distinction should not make me eligible for a Pulitzer Prize.

Lastly, I did not make any point estimates and consider them the true value, and I didn’t make any argument “against efficacy.” Those are a complete invention of Meyerowitz-Katz.”