- Admin
- #25,291
Offline
So, can you use your Math brain to help explain to those of us who are Mathematically challenged, how Simpson's Paradox applies to COVID and how it makes it look like we are treating it better and getting fewer deaths, but we really aren't? I'm sure you are right, but Statistics tends to be beyond my Math abilities.
The point is different states (and even subsets of states in many cases) are all in different points of their own outbreaks (and actually have differing amounts of population density and mitigation affecting their curves, but we can ignore that for the moment), so when you add up all the numbers as a country, you get data that is useful historically, but pretty much not useful at all as a modeling tool of what is coming in the future.
By looking at each area's outbreak individually (and associated hospital capacities, death/treatment rates, etc), you can model more accurately. I'm not (and I don't think the tweeter above is) saying that we aren't treating better - just that making predictions based on national data is somewhat flawed.