Originally posted by revenge_of_shocka_khan
View Post
I believe I have access to all the data I need to compile a US wide rate of hospitalizations. I'll know when I am back in town, but I will predict this: Including every state that reports hospitalizations will NOT prove your point and will make your point look weaker. Including them all will cause that graph to widen even further from the peak.
FL has been linear, for example, meaning no change up or down -- and it's considered by you to be a hot spot. Linear weekly hospitalizations will do nothing to the graph at all.
AZ hospitalizations are dropping from their small 5-day peak from three weeks ago. After that? Sharp decline. Including it will cause the graph to digress more, not increase.
TX is showing growth in hospitalizations, however not to the extent you think it is. The issue is that COUNTIES are hot spots, not states. The more counties you include, the more that graph is going to rise and fall. But we shall see when I am back.
That data set was specifically built because those counties were designated as "emerging" and/or existing hot spots. Adding in the 80% of the rest of the country -- the vast majority being rural -- is not going to strengthen your case.
Comment