Seven-Day Moving Average – New Cases & New Tests by State
The spike continues for most of the southeast, with Florida looking really bad. In the last week, Florida’s total cases grew by 45%, from 97,000 to 141,000. The numbers there are so high, it skews all the other states, so I had to take them out of the linear growth chart for the region.
The only exception in the spikes is Virginia, which has reduced new infections significantly since their spike peaked a month ago. Below are charts for Florida and Virginia, with some commentary.
Before I get there, one major assumption on my part as I interpret this data is that testing needs to outpace cases. When testing is higher, and growing faster, than new infections, the state has a good grasp on the overall impact of the virus. It means more healthy people are getting tested and the overall percentage of positives is low. As new cases catch up to testing, it’s showing that more people are getting tested because they are sick. If the growth rate of cases is outpacing testing, then it’s getting out of control.
This assumption becomes important in causation vs. correlation. When I look at these charts, there is some obvious correlation between testing and cases, but only when there are spikes in cases. When the states had more strict guidelines in place, testing far exceeded new cases. In that case there is no correlation and certainly no causation would exist. When the cases start ramping up, which they are now, the testing ramps up equally. Are we getting more cases because there is more testing; or are we getting more testing because there are more cases?