COVID Data Follies: Vaccination Rates, Relative Risk, and Simpson’s Paradox

Link:https://marypatcampbell.substack.com/p/covid-data-follies-vaccination-rates

Video:

Graphic:

Excerpt:

On Monday, December 6, 2021, I gave a talk with the title “COVID Data Follies: Vaccination Rates, Relative Risk, and Simpson’s Paradox”, to the Actuarial Science program at Illinois State University (thanks for the t-shirt, y’all!):

You may have heard statistics in the news that most of the people testing positive for COVID, currently, in a particular location, or most of the people hospitalized for COVID, or even most of the people dying of COVID were vaccinated! How can that be? Does that prove that the vaccines are ineffective? Using real-world data, the speaker, Mary Pat Campbell, will show how these statistics can both be true and misleading. Simpson’s Paradox is involved, which has to do with comparing differences between subgroups with very different sizes and average results. Simpson’s Paradox actually appears quite often in the insurance world.

I will embed a recording of the event, copies of the slides, the spreadsheets, and the links from the talk.

Author(s): Mary Pat Campbell

Publication Date: 8 Dec 2021

Publication Site: STUMP at substack

Israeli data: How can efficacy vs. severe disease be strong when 60% of hospitalized are vaccinated?

Link: https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-vs-severe-disease-be-strong-when-60-of-hospitalized-are-vaccinated

Graphic:

Excerpt:

These efficacies are quite high and suggests the vaccines are doing a very good job of preventing severe disease in both older and young cohorts. These levels of efficacy are much higher than the 67.5% efficacy estimate we get if the analysis is not stratified by age. How can there be such a discrepancy between the age-stratified and overall efficacy numbers?

This is an example of Simpson’s Paradox, a well-known phenomenon in which misleading results can sometimes be obtained from observational data in the presence of confounding factors.

Author(s): Jeffrey Morris

Publication Date: 17 August 2021

Publication Site: Covid-19 Data Science