Hacker News Reader

Comment by regularfry | Hacker News Reader

https://www.cdc.gov/mmwr/volumes/69/wr/pdfs/mm6947e2-H.pdf isn't subtle. You can make the argument that it might not be the wearing of masks specifically that's responsible for the difference here, but the weaker argument - that it's worth being the sort of area that will introduce mask wearing because whatever is going on is correlated with it - still works.

View on HN

Exactly. It's hard to see how masks themselves could have a negative effect. If they have a positive effect caused by people being more conscious of the current situation and keeping a greater distance, or not going out at all because having to wear a mask is a burden, so what? Worst case is they are neutral all things considered.

It may be hard to see how masks themselves have a negative effect, but it's easy to see how mask wearing as a medical intervention in the real world could. All it would take is people being willing to stand slightly closer together or take slightly more risks because they believed masks were effective protection.

I don't like this argument b/c what I have seen is that anti-mask people are generally covid deniers and take risks regardless.

With that said, the medical professionals telling people to wear masks always say in addition to distancing and hygiene.

As a counterpoint and to reiterate the parent comment's point, masks are a stark visual reminder that you need to practice other measures like social distancing and going out less. I'd be willing to bet that masks increase likelihood of practicing safe pandemic measures due to a "top of mind" effect rather than decrease them due to a false sense of safety.

That's why the message must be clear: even if it protects a single wearer just 20%, it can provably have huge benefit to everybody, but only if everybody does it (1) while all that doesn't mean that other measures could be avoided (2), especially if R raises in spite of one single measure.

1) https://news.ycombinator.com/item?id=25301380

2) The Swiss cheese model of how defences, barriers, and safeguards may be penetrated by an accident trajectory (see the figure):

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1117770/

https://www.wsj.com/articles/how-the-swiss-cheese-model-can-...

Very interesting! I'm a bit confused about the data/graph that shows their main conclusion, anyone know why:

They use a 7-day rolling average and general consensus is that covid has a 3-14 day incubation period. Combining these two would indicate that the mask's effect should be gradually becoming stronger starting ~4days after introducing and reaching peak effect after 21 days in the graph. The graphs show a strong effect after 4 days and remain constant after that. Does that undermine the conclusion or am I seeing this wrong?

They use a linear regression while infection is an exponential process in my understanding, doesn't this undermine the regression?

Reply on news.ycombinator.com