CDC, COVID-19 and Math

Because they never published the stats. CDC warns antibody testing still too inaccurate to use for coronavirus-related policy decisions.

The math is the point, but the upshot is that about half of the positive antibody tests are false positives.

In other words, less than half of those testing positive will truly have antibodies.

It is crazy, that they are only telling us these stats now.

2 thoughts on “CDC, COVID-19 and Math

  1. You know, ever since this Covid-19 nonsense got started, I’ve been telling myself somebody needs to write a piece on Type I and Type II statistical errors, and ways to deal with them.

    I just wouldn’t trust myself to do it from memory, I’d have to dig out some books I’m not even sure I own anymore.

    Liked by 1 person

    • No one would read it. Or not the people who need to read it.

      When was the last time you saw anything statistical in the mainstream press that included uncertainty for the polling numbers? Hell, it can be difficult to determine what the sampling strategy for a given poll is. And that is pretty basic statistics. (A whole lot of polls today include tremendous self-selection bias, which is usually ignored.)


Comments are closed.