Wednesday, July 29, 2020

Interpreting numbers

Raw numbers of disease cases aren't always very informative, beyond some general "Holy cow, that's a lot of cases!" or "Is that a problem?" How many tests were done; how big is the population; and on and on. The details matter.

I wasn't expecting this detail: in reporting cases "they prioritize those positive cases, the negative cases are dragging behind a day or two or three." It makes sense, but it means that you have to use not the test numbers as given by the day of information release, but by the day the test was administered--which adds a little more complexity. And my observation of reporters suggests that they don't handle complexity well.

No comments: