Masks help. But we should be careful not to over-interpret correlations like this. These uncritical takes regarding the efficacy of masks are very easy to poke holes in. Such arguments are likely to vindicate those who are skeptical of masks, rather than convincing them. https://twitter.com/EricTopol/status/1320027008286388224

@EpiEllie explains it better than I could in her linked threads concerning ecological studies, where a population-level exposure is correlated with a population-level outcome. https://twitter.com/EpiEllie/status/1320007430441472000

This is my own attempt at visualizing the issue of treating mask-wearing as if it is the sole explanatory factor for present outbreak size by state (measured as % of people knowing someone with COVID symptoms).

Obviously, there are other factors that influence mask wearing at a state level, as well. But I suspect having experienced a severe outbreak of COVID-19 in the past is a strong contributor.

I would also caution against uncritically promoting correlations that seem to confirm pre-existing biases. Again, it's easy to poke holes in and is unlikely to convince anyone skeptical of masks. https://twitter.com/_cingraham/status/1319658169077952512

COVID death rate/100,000 is probably a more important cause of differences in present state-wide mask wearing than who the state voted for in 2016.

https://www.statista.com/statistics/1109011/coronavirus-covid19-death-rates-us-by-state/

It's possible that partisanship could help explain some outliers, like Vermont's high mask wearing and extremely low COVID death rate. But on the other hand, Vermont is in the same region as many of the states with the worst death tolls (NY, NJ, MA, CT).

And besides, the variables of interest are measured at a very coarse level (state) using weak proxy variables (% who self-report mask use, % who know someone with COVID-19 symptoms). Measurement error is a big concern.

Yes, the correlation in this survey is interesting. But let's not read into it too much more than that. It's far from the best evidence we have that masks are effective.