All this trouble to get around taking the vaccine when they could just walk to their grocery store pharmacy and get the vaccine.
2 Likes
Yeah, they said they would unblind everyone as soon as the booster is approved.
1 Like
Because itâs not? The intervention was the education and not masks. The mask use in other communities wasnât controlled. You canât decouple that education about the risks of covid from the mask intervention.
You can still use it to support your case, but itâs not particularly high quality evidence.
Fruit flies must be PISSED!
The mask use in other communities was observed. The difference in actual real-world mask wearing is known and is the variable which was compared with COVID seroprevalence. How they brought the level of mask wearing up is also an interesting question, but somewhat decoupled from this study which is a comparison of actual mask wearing levels with COVID seroprevalence outcomes.
What on earth are you talking about? You could have done any of the following:
- Give people in the treatment group a truckload of free masks.
- Tell people in the treatment group that itâs critical they wear masks.
- Pay people in the treatment group for wearing masks.
- Telling people in the control group that if they wear masks, they will lose weight.
- Some combination of the above.
As long as the intervention is randomized and leads to them increasing the likelihood of wearing masks relative to a control group (it did) without altering other risk-related behavior, then itâs an RCT on the effectiveness of masks. (And this is an RCT as the village level, not the individual level. Of course individuals know if theyâre wearing masks or not.)
A real critique that you could make is that thereâs some correlated factor driving the results. Something like, âWhen people are educated about the importance of wearing a mask, they experience an overall increase in how concerned they are about the general risk of COVID. So the treatment leads them to reduce their risk in OTHER ways, not just wearing a mask. And these unspecified OTHER ways are actually driving the decline in infection among the treated villages.â But if thatâs your critique, youâd need to specify what other behavioral changes the treatment villages are experiencing and how those changes biased the results.
I have no idea what you mean when you say that the mask use in other communities wasnât controlled. Of course it wasnât, but that seems irrelevant, since the entire point of the RCT is to assess the difference between the treatment groupâs observed outcomes and the counterfactual outcome they would have experienced absent the treatment. And the use of an âuncontrolledâ control group is the proxy for that counterfactual. I donât even know what a âcontrolledâ control group would mean - telling them that they canât use masks?
3 Likes
exactly? yâall donât seem to really grasp what a real RCT is. Regardless, thatâs kind of a nitty point and Iâll let it go.
Wow, heâs giving up an entire CFL professional career. Thatâs leaving dozens of dollars on the table!
10 Likes
No, I agree that itâs not a real RCT (but a study with this kind of sample size is worth paying attention to). What I disagree about is that this matters:
You donât need to decouple education from masks. You have the actual numbers on how much mask wearing increased. It doesnât actually matter how that was achieved, like maybe an education campaign, or maybe neural manipulation by extraterrestrials. Like I guess the argument would be that thereâs a âplacebo effectâ where people are just able to positive-psychic-energy COVID spread into being less even though masks donât actually work. I just donât believe thatâs physically possible.
1 Like
Itâs not difficult to think that communities educated on covid would be faster to isolate or more hesitant to gather in groups. This doesnât require âpsychic energyâ, which is more than a little insulting from you.
I guess thatâs true. They also observed what percentage of the time people observed social distancing requirements and it was larger in the educated group, but not radically larger. So a larger behavioural explanation is possible, yeah, but it seems sort of unlikely to me.
This is addressed on page 32 in a very compelling way:
- The educational/encouragement interventions were applied to all treatment groups, regardless of whether the treatment groups received cloth or surgical masks.
- The treatment effect on physical distancing (as a proxy for non-mask risk-reduction behavior) was similar for the cloth group and the surgical mask group. Both groups exhibited an increase in physical distancing with the treatment, which raises the concern of a non-mask correlated omitted variable.
- But even though the two treatment sub-groups exhibited the same change in physical distancing, they exhibited significantly different outcomes in terms of symptomatic infection. Thatâs reasonably good evidence that the infection outcome was driven by the mask wearing rather than other non-mask risk-reduction behavior that might have been induced by the treatment.
I am impressed with the responsible reporting on the science in this article, although perhaps a little less so after clicking on the author and finding out heâs a PhD in molecular and cell biology and not just a random journalist. It would be easy to sensationalize a study like that, but he doesnât and instead writes a pretty good take.
my kid got his shot this morning for the flu
3 Likes
haha and now everyone knows her fake birthday too.
I think we can bridge the gap a bit here by being a bit more nuanced about what is being measured. Since at best 42% of the people were wearing masks properly, itâs not exactly a measurement of the effectiveness of wearing a mask vs. not wearing a mask, which seemed to be CNâs umbrage with some of the interpretation of the results. Instead what is being measured is the effectiveness of a mask education campaign, which measurably increases mask use and reduces the spread of covid compared with no education campaign. Thatâs an encouraging if unsurprising result, but it also doesnât quite qualify as a quantification of the effectiveness of masks.
3 Likes
To be honest, given all the randoms who were just volunteering at vaccination events, it would not be at all surprising for someone to come out with a card that says âMadernaâ.
In this case, the clincher is that itâs twice with the same handwriting.
If she had good friends, theyâd only celebrate her fake birthday and never her real one.
If youâre measuring variables at the village level, youâre looking at the change in infection rate as a function of the change in mask-wearing. Thatâs quantifying the effectiveness of mask wearing, as long as you assume the different rate of change in mask wearing is random and not correlated with changes in other non-mask factors. And itâs possible that the educational component prompts some change in a correlated non-mask factor. But focusing on the educational aspect, rather than the mask aspect, doesnât explain why the results differ so much between the surgical mask groups and the cloth mask groups.
Itâs true that itâs not giving an estimate of how infection rates would change if villages went from 0% mask wearing to 100% mask wearing, but I donât understand that critique. Asking how X% increased mask usage affects infection rate seems quite relevant to the question of mask mandates (where youâd also increase mask rates from some non-0% level to some non-100% level). (You could even assume that the relation is linear and just calculate the estimated effect of going from 0% â 100% mask wearing yourself. But thatâs probably a pointless exercise because: a) itâs not likely to be linear, and b) the effectiveness of masks is probably also a function of other covariates that differ between the villages studied and the places you might be interestd in.)
I donât understand looking at this study and saying that it doesnât offer high-quality evidence for the real-world effects of increased mask usage.
2 Likes
I am a science idiot but I have always felt pretty good with Ars Technical reporting on this stuff in a way a dummy like me can understand and also being accurate.
I was curious what the more scientifically inclined felt, and that is good to hear, at least for my personal consumption.