With regard to COVID, we are awash in unreliable or even intentionally misleading empirical claims. A neurosurgeon from D.C. who understands statistics analyzes a popular claim that is being made on behalf of mask mandates.
I encountered a Washington Post figure making its way through social media, suggesting that not wearing masks is conclusively the reason for COVID spread:
If you look at the figure closely, the y-axis is “% who know someone with COVID-19.” Why on earth would any reasonable scientist or public health official use this “survey” variable to honestly study a scientific/public health effect? The reason is, it is not true.
I pulled the % wearing masks variable from the graph (in and by itself a dubious measurement), and correlated it with the Rt value (that you can pull from Rt.live). The RT value is the exponential coefficient that suggests growth (>1) vs contraction (<1) of the virus. I then plotted the results:
There is no correlation whatsoever. The “trend line” I plotted has a 0.02 correlation (1.0 means full correlation). When you take the outliers out (WY, RI, MS, DC, SD), the correlation drops to 0.01. In fact, some of the states with the worst RTs right now (NJ, CT, MA, RI) have the highest “mask wearing” percentages!
Furthermore, the graph is titled “Fewer covid-19 symptoms reported in states with higher rates of mask use” — which is completely wrong! The data do not graph symptoms, only the loose perceptions of individuals who “know” someone, who were surveyed.
The fact that this passes off as “journalism” or “science” is incredibly pathetic — and it is being “liked” on social media by people who should know better. The reasons for this are the obvious political implications.
Hope you can make a comment or two regarding this. People need to understand what is going on here!
This kind of fake science is rampant, and as our correspondent says, the motivation is political. The Washington Post is like the New York Times; no one should take anything it says seriously.
https://www.powerlineblog.com/archives/2020/10/do-masks-work.php
No comments:
Post a Comment