Really great article by statistician @nataliexdean on how low-quality data are severely affecting the covid effort. nature.com/articles/d4158…
"One of the best ways the world has to get a clear view of COVID-19 is going underused. It’s time to exploit the power of random sampling."
Basically, a lot of the times, we have absolutely no clue what's going on with covid because of problems like undercounting or because our information is weeks out of date.
Are we undercounting because we've run out of tests and covid is surging out of control? Is it because lots of cases are mild so people aren't coming in to get tested?
We eventually figure it out when people either do (or don't) start overwhelming the hospital system or dying, but by then our understanding is weeks behind the virus.
Monitoring waste water helps a bit but "waste water can’t pinpoint who in a community is getting infected and who is getting sick." This means we can't pinpoint our efforts to respond.
A major problem with random sampling is people might opt out of it and end up biasing the sample. I think we should cross this bridge when we get to it. Maybe it will only be a small number so why not try?
I also think if politicians on all sides put a real effort into encouraging people to do it for their country, then it might be less of an issue.
I've been wondering why we aren't doing more random sampling for a while so I'm glad to see efforts like this to bring the issue to the attention of a wider audience.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
