Strong link between mask wearing and reduced COVID-19 spread

As reported by the Washington Post, Carnegie Mellon’s CovidCast provides strong evidence of the importance of wearing masks.

The R2 of this correlation is 0.73, which is extremely strong.

This evidence reinforces the importance of strong mask policies for public, indoor spaces.

5 Likes

Freedumb advocates are still the main problem, and secondarily, conspiracists and denialists. Of course, many are all of the above.

1 Like

It is a nice graph, but there is a risk that it is skewed by the fact that different states are in different phases of the pandemic.

The Northeast green grouping may be in the bottom right because it was hit hard earlier and has stricter laws, so more people there wear masks and more people there are now immune so fewer people will know others with the disease.

This may make the correlation appear greater than it is.

An interesting graph would be one that has the same horizontal axis and a different vertical axis:
“People who knew someone who had Covid in the past”

I suspect past experience with an acquaintance having Covid is a larger driver (than current knowledge) to consistent mask wearing.

Thanks for this. Much needed. Masks significantly reduce the intake of virus.

2 Likes

Another possible bias in this graph is population density.

If the graph had kept the same horizontal axis and had a vertical axis of “Population Lack of Density: Square Miles per Person,” the graph might look similar.

The areas with low population density (those Western states not on the coast) would still be upper left. The densely populated East Coast states would still be bottom right.

Would we conclude that mask wearing increases population density?

Yeah, more are surviving.

The bottom of the curve after the second peak is higher than the first peak.

Ah, a little comic relief. Thanks, Dale.

Hi Vance -

Are you aware of LA? San Francisco? San Diego? Seattle? Portland? Denver? Salt Lake City?

Western states have dense population centers just as Eastern states do.

Best,
Chris

Please read my posts again.

I said those low density Western states not on the coast.

Then look on the graph, and you will see the top left has S.D., N.D., Wyoming, Montana, Idaho. Those are low density Western states not on the coast.

I read things too quickly sometimes too.

Maybe you thought I meant East Coast. I did not write that and did not mean that.

If you are Eastern-centric, you might have jumped to defining “coast” as exclusively meaning “East Coast.”

“Under control”, unlike us who have “turned the corner” and gone around the bend.

This one’s from a single state, Tennessee:


The study is here

5 Likes

Thanks for clarifying my misreading. Hat tip.

That’s a very small sample size, don’t you think?

And why would your sample exclude lower density, lesser-impacted-early states like Iowa, West Virginia, and Arkansas?

Best,
Chris

1 Like

Why didn’t you raise that objection to the original analysis and calculation of r^2?

I was pointing out the shape of the graph remains the same with a different vertical axis. Please read my post again. I think you will understand.

There is correlation, however people in a low density/ low population state are likely to know a larger portion of the state’s population than high density/ high population. So why did they not plot cumulative cases in the last 28 days versus mask wearing?

It is interesting that nobody is below 60% and most daa is in the range 70 to 80% so it can be concluded that most americans are wearing masks in public.

Looking at the source COVIDcast Dashboard | DELPHI and reading the fine print, the people wearing masks is done by a facebook survey as are most of the other stats. So all of this is self reported and likely subject to biases.

1 Like

Interesting.

I am surprised Chris posted something so questionable.