[NoPie]

I understand how public can misunderstand this phrase but scientifically it is clear and justified.

Sometimes “no evidence” means that we haven't found compelling evidence yet. Some people are desperate and want all the studies to be done immediately and evaluate the risk if the theory is real or not.

But for scientists the desire to reach certain outcome is actually counterproductive as it can introduce bias. Slower and less passionate process can lead to better results.

For example:

1) We had no evidence of covid being airborne and then we found this evidence.

2) We had no evidence that masks help and then we found no evidence.

Two different theories, two different outcomes. Covid is airborne and it changes our understanding (however, we gradually realized that it is impossible to limit the spread and all the measures ultimately were useless). And that masks most likely had very little effect.

The reporters could write better for lay public explaining that “no evidence” means that currently we don't have evidence but it could be found later or that “no evidence” is actually that we have a lot of evidence that is indicating in some other direction and the chance of new evidence that rejects those findings is smaller but still could happen.

[Sporktacular]

"We had no evidence that masks help and then we found no evidence."

Actually we did from the decades surgeons wore masks to prevent the spread of airborne diseases. Once we knew Covid was one of them, then even if the protection was one way, it would have reduced infections if everyone wore one.

[timr]

> Actually we did from the decades surgeons wore masks to prevent the spread of airborne diseases.

Funny you should say that. There were randomized controlled trials on exactly this question prior to 2020. Guess what the outcomes were?

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4480558/

(Notably, this review is from 2015...it is not subject to the ridiculous politics of Covid.)

> However, overall there is a lack of substantial evidence to support claims that facemasks protect either patient or surgeon from infectious contamination. More rigorous contemporary research is needed to make a definitive comment on the effectiveness of surgical facemasks.

Also, rather (in)famously, a review saying the same thing was censored from the web in spring 2020, because...reasons.

While I'd personally love to see extensive, rigorous investigation of this question, simply repeating "masks work", or "people didn't do it correctly" (your current argument) when all of the current high-quality evidence suggests otherwise doesn't inspire confidence in those of us who actually use the scientific method.

[Sporktacular]

"All of the high quality evidence"

If the claim masks don't work didn't come from the same people who said isolating didn't work, I might take them seriously. I might not just see it as motivated reasoning or another lame effort to discredit or cast suspicion on authorities.

If it didn't come from the same conspiracists who see nefarious censorship everywhere or people who only see their personal rights being infringed over our social obligations to each other, I might take them seriously.

Or I could take them seriously if they understood the risk of confirmation bias by cherry picking preferable information when credible contradicting studies or meta-studies exist (like this one that states "The preponderance of evidence indicates that mask wearing reduces transmissibility per contact by reducing transmission of infected respiratory particles in both laboratory and clinical contexts." https://pubmed.ncbi.nlm.nih.gov/33431650/ ). Or the risk of socially promoting that one-sided certainty.

Or if they just understood the precautionary principle that if in the face of competing evidence, we could potentially all benefit from taking the more cautious approach at the risk of minor personal inconvenience, I might take them seriously. But I don't because they aren't serious people. And they aren't even slightly interested in applying the scientific method.

And they don't realise they are in the loud, unreasonable minority who have a megaphone they would never have had before the Internet, and they don't feel obliged to use that power responsibly. With time and education I hope they will dwindle in number, or at least shut up a bit.

[ifyoubuildit]

I was curious what an llm might think about this comment:

> Please rate the following comment from 1 to 5 on how rational it is, how emotional it is, and whether or not the author is making a strong or weak argument:

> Rationality: 3/5 - The author presents a reasoned argument supported by a reference to a scientific study. However, the argument is somewhat undermined by a lack of direct engagement with specific counter-arguments and a generalizing tone about those who hold opposing views.

Emotionality: 4/5 - The comment is emotionally charged, especially in its dismissive tone towards those who disagree with the author's perspective. The language used ("lame effort," "shut up a bit") indicates a strong emotional investment in the topic.

Strength of Argument: 3/5 - The argument is moderately strong. It relies on a credible source and logical principles like the precautionary principle and the risk of confirmation bias. However, it is weakened by broad generalizations about the opposing side and a lack of specific rebuttals to their claims.

That lines up pretty well with how I perceived that. There is a lot of emotion and broad generalizations in these conversations. Is anyone changing their minds about any of this after 4 years of digging into their positions?

[Sporktacular]

"Is anyone changing their minds about any of this after 4 years of digging into their positions?"

LLMs probably don't understand that not everyone is intellectually honest, that conspiracists who ignore a preponderance of expert evidence to insist on their own positions, generally feed on the attention. I'd say that in countering that, sarcasm, caricature and calling attention to the absurdity of their arguments can be more persuasive to others than watching two non-experts debate specific points while carefully select papers ad nauseam. I'd guess others might even identify with the emotion of a commenter's frustration that after years of evidence, discredited positions that unnecessarily risked people's lives continue to be retold.

And I'd hope that others would recognise the expert consensus that masks are effective at scale in flattening the curve, buying time to develop vaccines and hence saving lives, and see the importance of reason over a choice of linguistic flourish. Maybe since an LLM can't understand how sick of the post-truth era we've become you'd want to weigh in on the discussion itself.

[peyton]

My understanding is that flattening the curve was an argument in favor of social distancing and movement restrictions. It predates mask mandates. A properly trained LLM would miss your masking reasoning. See

https://www.nytimes.com/article/flatten-curve-coronavirus.ht...

The bigger problem is an LLM doesn’t have access to things that aren’t written down. Although it may see a bunch of publications popping up declaring “pain is the fifth vital sign” and a bunch of new pain management CME courses, it cannot see the pharmaceutical reps showing up at doctors’ offices to sell Oxycontin.

[timr]

The Masks4all paper (your link) is neither credible, nor is it a study. It should never have been published.

It is a mashup of a literature review with a bad methodology, and a simulation that adds no new information to the debate. It excludes and/or minimizes randomized clinical trial data when that data doesn't support the desired narrative.

For those who might read this later, the Cochrane collaboration published a high-quality review of masking literature:

https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD...

[Sporktacular]

That's one paper out of dozens, feel free to pick another.

“Many commentators have claimed that a recently-updated Cochrane Review shows that ‘masks don’t work’, which is an inaccurate and misleading interpretation,” Dr. Karla Soares-Weiser, the editor-in-chief of the Cochrane Library, said in a March 10 statement. https://www.factcheck.org/2023/03/scicheck-what-the-cochrane...

It's not particularly conclusive, whereas many other sources point to a small increase in protection. Which is important at scale.

[NoPie]

>> That's one paper out of dozens, feel free to pick another.

It is not. It is the review of all available evidence (studies).

[timr]

I picked one. Did you read it, or did you just read the factcheck.org link?

[NoPie]

>> If the claim masks don't work didn't come from the same people who said isolating didn't work, I might take them seriously.

It didn't. The claim came from Cochrane group.

So, are you taking it seriously now?

[Sporktacular]

Yes. And the reasonable criticisms of it too. And the people I'm referring to continue to misinterpret the claim to twist it to their own ends.

Individually masks reduce infections. En masse, even more so.

[classichasclass]

Are you actually suggesting we should have surgeons unmask in an operating theatre and see if the rate of post-op infections go up? I'd like to see that get through an IRB. This is the same mentality that says every current vaccination should be compared against saline placebo.

[NoPie]

Actually we could. Maybe we won't do such studies because they are irrelevant because avoiding spitting into the surgical opening is a reason good enough. But then we cannot claim that masks during the surgery prevent the spread of airborne viral diseases.

But if we needed to guard for them and the good evidence is lacking, then not testing would be unethical.

Yes, even covid vaccine today could be compared with placebo, for example, in children. Europe never mandated covid vaccine for children and today in the UK they cannot even get the vaccine unless in a risk group. The US however recommends covid vaccine for children without the evidence that it makes any difference today. It definitely should be tested in trials before such recommendations.

I don't think that IRB would reject such studies. At the start of pandemic everybody was saying that doing human challenge trials by infecting healthy volunteers would be unethical. And yet the UK did them. The red tape takes time and I can understand that during pandemics we may need to act quickly and cannot test everything. But in principle we can and do need to all kinds of trials to obtain proper evidence.

[classichasclass]

As you say, COVID was an outlier because of the urgency of the situation and the newness of it. We don't have either situation with OR hygiene, and if we're wrong and OR masks actually are doing something preventive, then we would be doing harm to the patients involved relative to the inconvenience to the surgeon.

The vaccine question was about vaccines in general, much as RFK Jr is talking about doing. Again, with the weight of long experience on how well they prevent diseases in mind, it would be unethical to expose a kid to that by giving them a placebo shot. Measles is pretty benign but not totally so. See a case of SSPE in your career and you'll never forget it.

[NoPie]

If we already have strong evidence that the current vaccine is effective, then I agree, we don't need to to do another placebo controlled studies. But very often in medicine we don't have any evidence besides our beliefs.

[Sporktacular]

No, I think they're more saying that until studies prove with certainty that surgical masks work with 100% effectiveness, doctors should not feel the inconvenience of having to wear them.

[classichasclass]

The specific claim quoted was for surgeons, though. I don't see how this can be tested without having them mask off in the OR.

[Sporktacular]

True, but I don't see them proposing something so irresponsible by a surgeon. Just everyone else.

[indymike]

> Actually we did from the decades surgeons wore masks to prevent the spread of airborne diseases.

I remember asking my dad (who was a doctor who performed surgeries - mostly c-sections and appendectomies - often) about masks, and his answer in the early 80s was interesting:

“It’s mostly to prevent me from getting spit into open wounds and incisions when I'm talking or I have to sneeze or cough. Bacteria is a real problem, and the mask stops that.”

I never really thought much of that until recently.

[classichasclass]

But you can see this go both ways. Ortho surgeons in total procedures have full air filtering due to the large amount of bone dust that's liberated, some of which are very fine particles. No one wants to breathe that in.

Meanwhile, I'll be in TB clinic shortly with an N95 mask on. I've yet to convert my TB test in 17 years.

We absolutely rely on surgical masks not to contaminate the field. But they don't have one way valves, so if we trust the airflow one way, it's logical to trust them for airflow going the other.

[generalizations]

>> “It’s mostly to prevent me from getting spit into open wounds and incisions when I'm talking or I have to sneeze or cough. Bacteria is a real problem, and the mask stops that.”

> so if we trust the airflow one way

Pedantic, but it matters: that's not airflow, that's droplets and spit.

[isolli]

Surgeons don't wear masks to prevent the spread of airborne diseases. They wear them to prevent spittle going into an open wound and to protect themselves from blood splatter.

If they wore masks to prevent airborne diseases, they would wear them when meeting patients, not just during the operation.

[sobani]

The difference is during a regular meeting, the skin is uncut. The skin is a major protection against diseases and the body has a bunch of mechanisms at its regular openings (nose, ears, etc) to protect you.

When you cut through someones skin, you bypass one of the major first lines of defense. Therefor surgeons reduce the risks, for a similar reason why clean their tools before use.

[u32480932048]

One of my pet peeves is the continued use of ambiguous phrases like "masks work" or "masks help" when there are effectively two different mechanisms (inhalation/exhalation).

I suspect the prolonged mass confusion over such an elementary topic will be one for the Science Communicator books.

[titzer]

IMHO the issue is that masks work but masking as a social phenomenon can suffer a number of problems, like non-compliance, partial compliance, poor fit, children, misuse, reuse, damaged masks, cloth masks, needing to eat, and all kinds of factors. So scientific experiments in a controlled setting with one or two people show that masks work as a mechanism, but getting everyone to go along with and practice good masking etiquette might not work so well. So population studies show masks working much worse than their mechanism would suggest they can. Especially when there is intentional non-compliance and protests motivated by culture war battles.

Arguing with people who just spout "masks don't work" and then intentionally are non-compliant (and encourage others to) is like arguing with motivated idiots.

No, masks work, period. Masking only works if people freaking do it.

edit: spelling

[agos]

on the flip side, so much of the Covid discourse was about these fabled policies which would do great things (but required 100% compliance), followed by indignation/panic and hysteria/angry mobs when it turned out that 100% compliance is hard.

[Sporktacular]

We won't need 100% effectiveness and 100% compliance. If they only work at 10% effectiveness, then over a network of thousands, they will measurably save lives.

Studies have shown they work (such as https://pubmed.ncbi.nlm.nih.gov/33483277/), despite the 'ah, yes but only [insert unsupported caveat here]' of some.

[timr]

> No, mask work, period.

...on a mannequin, in a lab, or when used as a filter between two hamster cages. Perhaps. But there have now been plenty of negative studies in hospitals, which really starts to beg the question: if you can't get effective compliance in a hospital, where are you going to get it?

Just to be clear, I was with you right up until the part I quoted. It's fine to say that mechanistic studies show something to be true, but it's totally wrong to leap to the conclusion that these mean anything. If I tell you that you're 100% certain to lose weight if you stop eating, that's True ("Starving works. Period."), but it's not meaningful.

Every failed drug ever tested worked in a laboratory before it went to clinical trials.

[Sporktacular]

Except masks do work. Here:

https://pubmed.ncbi.nlm.nih.gov/33431650/ - "We recommend that public officials and governments strongly encourage the use of widespread face masks in public, including the use of appropriate regulation." https://pubmed.ncbi.nlm.nih.gov/33370173/ - "Evidence suggests that the potential benefits of wearing masks likely outweigh the potential harms when SARS-CoV-2 is spreading in a community". Mentions the network effect. https://pubmed.ncbi.nlm.nih.gov/32497510/ - "Face mask use could result in a large reduction in risk of infection" https://www.mayoclinic.org/diseases-conditions/coronavirus/i...

"...on a mannequin, in a lab, or when used as a filter between two hamster cages. Perhaps." No, you are coping hard. The studies above are not done by idiots.

"if you can't get effective compliance in a hospital, where are you going to get it?" Nice pivot, but the subject is whether they work, which they do. Compliance is a different question. That's like arguing that condom effectiveness is low when people don't use them properly. And because the figure is low, let's just conclude condoms are ineffective so people should stop using them altogether.

"Every failed drug ever tested worked in a laboratory before it went to clinical trials." And some actually worked in trials. Your job is to explain which is the better analogy. That unanticipated practical considerations exist does not invalidate every conclusion you dislike.

However, even if masks work slightly, or not at all, even if there are dozens of studies saying so, there is enough conjecture to warrant wearing them anyway just in case. If two equally qualified mechanics disagree on whether to change your brakes, you change your damn brakes. If you don't that's stupid. If you don't and then drive a busload of people, that's criminal.

But masks do work, and they're cheap and barely an inconvenience. The problem is everyone thinks they're smarter than an epidemiologist. Thanks Internet.

[timr]

I'm not getting in an argument about this. You don't understand what you're reading, and you're cherry picking papers based on a poor understanding of data quality. I will say that your first link is the Masks4all (Jeremy Howard) paper, and it is not a reputable scientific publication. It doesn't follow a valid methodology for a literature review, and leaves out/minimizes major RCTs that don't support their pre-determined conclusions.

The second link is not a study, or a meta-review, and contributes nothing. The fourth link is not a scientific paper at all.

The third link is the WHO summary of masking data in mid-2020. It covers the same ground as the Cochrane review (below), but re-weights the data somewhat arbitrarily to achieve the stated conclusions.

The Cochrane review is the gold standard summary of the evidence for public masking, considers all published data, and includes/excludes/weights data based on a rigorous standard for statistical and experimental quality. Please read it.

https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD...

[Sporktacular]

If masks only work one way and one infected person is in a train car with 10 uninfected people, then they all get protection if they all wear masks.

[samtho]

The problem with the phrase “No evidence” outside of a scientific context is that it sounds like it’s dodging the question, when, as you pointed out, it just means we don’t have data to support that relationship. The other way this is used is to try to “prove” a negative by communicating that we cannot establish a link, and may be accompanied by the expert stating there is evidence for the opposite case. Text bites rarely have context other than this “no evidence” statement.

For a layperson, they feel like they are having the wool pulled over their eyes and are not given a truthful answer. For example:

    Interviewer: Can you say for certain that X causes Y?
    Scientist: We have no evidence that X causes Y.
    Interviewer: But can you say for certain that Y is NOT caused by X?
    Scientist: There is no evidence of that relationship.

The scientist is trying to choose their words carefully because they are operating on the principle of only communicating what the data is telling us. The interviewer wants a clear, definitive answer. Both parties become frustrated.

I do agree with the article that this is poor communication because the scientist sounds like a lawyer and is hiding something rather than stating the facts of the situation. I’m sure there are better ways to communicate this that communicate either “we don’t know yet” and “we do not believe there is a link” instead this line.

Side note: another related and interesting distinction is the difference between unproven and disproven. If you saw a headline that a major hypothesis was “unproven”, it means only that there is no data to support it. Disproven is what the layman often thinks this means which is actively finding evidence to the contrary.

[blub]

There’s different kinds of “masks” and different kinds of “work”

FFP3 are the standard for protecting against airborne disease and they work as intended. That’s literally one of the purposes they’re designed for. FF2/N95 also work quite well, which was known for a long time, including from SARS and MERS studies.

Work can mean that they prevent transmission or reduce the risk. It might have been worth wearing a surgical mask, as I remember reading that it did reduce risk enough for it to be recommended in Germany once the establishment got over their mask procurement predicament.

Finally, original Covid was not as contagious as current Covid, which is one of the if not the most contagious airborne disease.

Based on e.g. SARS it would have been rational to recommend FFP2 masks, as was done in e.g. South Korea. However, incompetent Westen governments utterly failed to make such masks available to the population, so they tended to err on the “no evidence masks work” side.

[bluGill]

Just because you have not looked for evidence does not mean it doesn't exist.

I have no evidence that my neighbor has a water softener - but that evidence does exist and I could obtain it in several ways (ask him; break into his house). Maybe he doesn't have one, maybe he does - but either way the evidence exists if I cared to look for it.

[indymike]

> Sometimes “no evidence” means that we haven’t found compelling evidence yet.

Most of the time when no evidence is used, it's to drive political or business activity. For example:

“There is no evidence that Asbestos is unsafe.”

I wish that news media would just be honest, or tell us when research is ongoing so people people were not panicking over something that is simply just not known.

“Researchers have good reason to believe that masks will help prevent the spread of COVID-19, but research is ongoing at _______ to determine if COVID-19 is airborne, and how effective masks are in preventing the spread of COVID-19.”

Unfortunately, the truth is boring and outrage and panic drive clicks.

[andrewla]

That's one sense in which it is used, but the other sense is the opposite -- "no evidence that covid vaccines cause people to become magnetic" is not a "yeah, we haven't checked for the evidence yet, but we'll get right on it".

The point of the article is because this can cut both ways it is impossible to distinguish between the shades of meaning, and the phrase should be discarded entirely.

[linuxftw]

Sometimes 'no evidence' also means there is evidence, they're just ignoring it. Like myocarditis.

[hnfong]

I think "ignoring it" is unnecessarily imputing bad faith.

But ignorance is a thing, even for authorities.

There's a difference between a person/institution having no evidence of X, and no evidence of X in general. So, "who" doesn't have evidence, and how much effort they took to attempt to find evidence before they declared it non-existent, is a relevant question when "no evidence" claims are proposed.

It's funny, because when a person or institution comes out and says "there is no evidence" of whatever (without a context), it would be discredited if any obscure person in the world has such evidence, even if not widely published. It would, IMHO, be much better to say that "we have looked into <all the reasonable sources and literature> and found no evidence" instead.

[acdha]

Where by “ignoring” you mean “carefully studying and doing comparative risk analysis”? That risk was reported early on, and has been very well covered over time but each round of studies has shown that catching COVID while unvaccinated is substantially riskier.

[linuxftw]

Public health officials summarily denied any association early on. The only serious risk they acknowledged was propylene glycol allergy.

[acdha]

Do you have any evidence to support that assertion? I first learned of the issue from the scientific community tracking the data collected by the public health community.

Here’s an example of what that looked like in June 2021, covering developments in May, just 5 months after the first country in the world had approved the vaccine (UK, 2020-12):

https://www.yalemedicine.org/news/myocarditis-coronavirus-va...

Similarly Israel’s public health agency’s report was covered in June 2021:

https://www.reuters.com/world/middle-east/israel-sees-probab...

[linuxftw]

Here's a quote from the very Reuters article you linked:

> The European Medicines Agency (EMA) said last week that heart inflammation after receiving the Pfizer vaccine had been no cause for concern as such incidents were similar rate to those in the general population.

The CDC and other agencies also continue to heavily downplay the risk as 'mild myocarditis.'

[ceejayoz]

"No cause for concern" and "denied any association" are not the same thing.

[mrguyorama]

Both of those are 100% accurate claims.

So what are you actually upset about?

[acdha]

Yes, and that’s accurate. People have carefully monitored it, but the risk is very low and much lower than getting COVID. That doesn’t fit any definition of “ignoring” in the English dictionary just because antivaxxers would desperately love to have something they weren’t wrong about.

[broast]

The risk of myocarditis may be higher for an unvaccinated person than a vaccinated person which makes the "downplaying" much more nuanced, doesn't it?

[darkerside]

Also scientifically justified, "Scientists investigating whether COVID is airborne". Either of those headlines leads people down a path.

Also, I think the whole question here is, who decides what qualifies as "compelling" evidence?

[stevenAthompson]

>Also, I think the whole question here is, who decides what qualifies as "compelling" evidence?

Qualified experts. Any other answer eventually turns into political gibbering and social media nonsense.

[gdubs]

No evidence sounds definitive. "We _lack_ evidence" would be clearer, for instance, because it at least suggests that their _could_ be evidence, but it hasn't been found yet.

[RyEgswuCsn]

__lack__ evidence sounds like you have already made up you mind but is just searching for confirmation.

I'd say a better expression would be "we don't know yet whether ...", e.g.: "we don't know yet whether Covid is airborne".

[mrguyorama]

No, actually, definitive would be "We have evidence to the contrary".

[somewhereoutth]

Except of course that masks do help, and we did find evidence. However many would like to justify their inability to accommodate the slightest inconvenience necessary to help the people around them - a clear case of starting with the premise ("I don't like wearing a mask") and then perceiving any data through that view.

[NoPie]

That's not accurate. Obviously there are many different studies with different conclusions. But by “evidence” we usually mean the total conclusion from all of them. In medicine it is usually done via metareviews and even then evidence can be graded to different levels and quality.

So, in short, Cochrane review shows that we don't have a good quality evidence that masks were effective. And the low quality evidence indicates that masks either had no or very little effect. Some people try to quote one or two studies out of context but that's not helpful because we need to take the totality of evidence into account.

It is possible that once we obtain high quality evidence, these conclusions will be overturned. Surprisingly there is very little interest in doing such studies.

[somewhereoutth]

Total conclusion from all of them would be a high bar, given that such a large section of society had become convinced that they do not work for reasons far removed from any facts.

Lots of people died.

[NoPie]

What do you mean by high bar? Cochrane group are specialists who do exactly that and provide gold standard of evidence. Sometimes the evidence goes against accepted wisdom. That's how science work.