Additional analysis on Kansas mask data: Does the size of the first wave (case levels before June) have an impact on the change in cases between August 11th - November 20th? (1/x)
https://twitter.com/covidtweets/status/1330734755634089986?s=20
Using population adjusted case numbers, below are the correlations between first wave cases, mask mandate, and change in cases after August. There is a significant negative correlation between first wave size and change after August. (2/x) 
The correlation between masks and change is positive, meaning mandate counties had bigger changes. However, this is likely confounded by first wave size, as three counties with bigger first waves are non-mandate counties. (3/x)
In fact, just looking at the counties without significant first waves, we find that counties with and without mask mandates had similar levels of increase in cases since August, with mask counties still doing slightly worse. (4/x)
Finally, comparing counties with significant first waves (4 of 22 counties) to those without in our data, we see what really is happening... Those with bigger first waves see a 2.8x increase, while others see an 8x increase in their cases since August. (5/x)
What do these all mean? Mask mandates have no effect on case trajectories, and the best predictor of how much cases increased since August is how many cases counties had in April-May. (6/6)
• • •
Missing some Tweet in this thread? You can try to
force a refresh
