Read on Twitter
Hey @ianbremmer--you're using incorrect data as the denominator. To be fair, FL puts it out on the dashboard, but using the difference in the daily "Total People Tested" number is flawed, and pretty obviously so. Here's why... https://twitter.com/ianbremmer/status/1309644793450881024
First, here's your tweet and graphic. The graphic is clearly using the day-over-day difference in the Total People Tested metric, which is what @COVID19Tracking uses (last I checked), and which is wildly inaccurate when seeking out current testing. https://twitter.com/ianbremmer/status/1309647741308538884?s=20
That number measures only the number of people tested *for the first time ever*. So anyone ever tested--in April, May, June, whenever--will not show up here if they test again tomorrow. This excludes tens of thousands of tests every single day.
In fact, here are pics of today's dashboard numbers for Florida. If you just take the difference in the "Total" from yesterday to today (5,213,142 - 5,186,646), you get 26,496. Yet the total number of residents tested from the previous day was 66,708.
No wonder the % is so high.
No wonder the % is so high.
Think of what it means to remove everyone ever tested. Anyone tested tomorrow who was one of the 5.2M people ever tested before will not be included in the denominator. This removes tens of thousands of people every day, and makes the % positive skyrocket.
For a simple example of what this does, take a quick read of the email I sent to @COVID19Tracking about the discrepancy above (and requesting they use the other reported data in the state daily report). I provide a scenario of how this would look in the first half.
Here it is:
Here it is:
To their credit, @COVID19Tracking responded and agreed that this number will *not* capture any repeat tester--no matter when their first test was.
As far as I'm aware, no other state's positive percentage denominator (total tested) does this.
As far as I'm aware, no other state's positive percentage denominator (total tested) does this.
The worst thing about this is that it gets *worse* as time goes on, because more people continue to get excluded from the day's test count.
How many of the tens of thousands of people getting tested tomorrow will be excluded? A lot, when a quarter of the state won't count.
How many of the tens of thousands of people getting tested tomorrow will be excluded? A lot, when a quarter of the state won't count.
Florida puts out other data though in their daily state report--and its very easy to capture better data, which is why I switched to that a while back once I realized how crazy it is to use the difference in unique people tested. Here's the link:
http://ww11.doh.state.fl.us/comm/_partners/covid19_report_archive/state_reports_latest.pdf
http://ww11.doh.state.fl.us/comm/_partners/covid19_report_archive/state_reports_latest.pdf
You're at 1K likes and 257 retweets as I type this, but it'd be nice to get a clarification based on the above.
Folks can still disagree with @GovRonDeSantis decisions, of course (and no doubt many will), but they ought to do it with clear eyes in terms of the data.
Folks can still disagree with @GovRonDeSantis decisions, of course (and no doubt many will), but they ought to do it with clear eyes in terms of the data.
