Beyond that, as the link tm Taleb’s blog posted by a commenter below shows, if anything, the study shows a positive effect for mask wearers.
> Limitation: Inconclusive results, missing data, variable adherence, patient-reported findings on home tests, no blinding, and no assessment of whether masks could decrease disease transmission from mask wearers to others.
I would not be quick to promulgate or depend on the study.
It seems obvious by now that universal mask policies have not been very effective at controlling the transmission rate when compared to other variables: known and unknown. Maybe it is a little bit effective but that has not been conclusively shown yet. Where I live, the Ro value dropped below 1 before masks were required and shot well above 1 months later. Compliance is high, probably as high as any public health policy
As for studies that do meet standards, they are hard, not easy, to dismiss. That work tends to get meta-analyzed and synthesized, not brushed off, at least by the reader community at large. I wish more of it got replicated, too, but that's a different thread.
Meanwhile, we still have to make decisions. But not entirely without scientific guidance. We can reason from studies of how the disease spreads and how potential countermeasures work. We can argue by analogy to what science has revealed about apparently similar phenomenon, like flu. Those aren't scientific arguments so much as other kinds of arguments hopefully informed by science. Pending better science as such, that's what we've got.
It seems reasonable to me to think that, whatever the scientific contribution of the study, the fact that it doesn't test the primary claimed benefit of masks suggests that it is not the last word on the matter. When I say 'I would not be quick to promulgate or depend on the study', I am not dismissing the study, but recognizing its limitations, as do the authors.
Once these factors are removed its unlikely that mask wearing provides any value beyond security theatre.
But that's the point.
We know that distancing works. We know that hygiene works. We don't yet know whether masks work or not, but we do struggle to find benefit when research is done (and this includes for source control -- there's plenty of research on that).
When scientists claim "masks work, we know they work, and we know they work well" they risk feeding a narrative that we can go about our normal day to day life so long as we wear a mask. That's untrue. We still need to distance, we still need scrupulous hygiene routines.
They also feed into a narrative that people not wearing masks are causing harm, which leaves those people open to harassment and attack.
Check out his Twitter feed for more.
No paper can ever show that. You cannot prove a negative (except in extremely bounded cases, of which the mask proposition is not one).
This is why statistical studies attempt to provide evidence supporting an alternative hypothesis. If you cannot accept the alternative hypothesis, you must accept the null. Studies do not attempt to disprove the alternative hypothesis, and accept it otherwise.
This study attempted to find statistically convincing evidence that masks provided personal protection, and failed to do so. Does it mean that they have proven that masks don't work? Of course not. There are many reasons why they might have failed to detect a real effect. But it does imply that whatever effect is being hypothesized is smaller than the power of the study they used to detect it.
For some reason this has led to media concluding that face masks do not protect you but that's not the same.
Another important effect of wearing a face mask is that it may protect other people from being infected by you but that effect was not part of the study.
Stripped of all the political connotations it's actually kind of an interesting debate for scientists to have.
https://news.yahoo.com/fauci-shoots-down-flawed-hydroxychlor...
Rep. Blaine Luetkemeyer (R-Mo.): "it was peer reviewed"
Fauci: "It doesn't matter. You can peer review something that's a bad study."