4 ms·
> hodgkin's lymphoma from The Vax. Rather than censorship, a better way of combatting this sort of rumor would be a government agency, for example, one that wo
by kleton 2y ago
> hodgkin's lymphoma from The Vax.
Rather than censorship, a better way of combatting this sort of rumor would be a government agency, for example, one that would be put in charge of food and drugs, collecting and publishing k-anonymized incidences of disease annotated by such things as what drugs and risk factors groups of individuals were subject to. This would allow such claims to be evaluated empirically.
- _hyn3 2y agoAt least until such 'food and drug' agency was inevitably overwhelmed by lobbyists, patent royalties, and a revolving door between industry and government. Then it'd be worse than nothing at all.
- bandyaboot 2y agoThis is all well and good, but let’s not pretend it would be particularly effective.
- fhub 2y agoCost of making up a rumor : 2c Cost of proving a rumor is false : $$$ Cost of disseminating that proof to say 90% of people fooled: $$ Cost of persuading those 90% that your proof is more reliable than the rumor: $$$$$
- lesuorac 2y agoAnd of course if you make a rumor that everybody else's vaccines cause autism you can profit off it by making an autism-free vaccine.
- spondylosaurus 2y agohttps://en.wikipedia.org/wiki/Brandolini's_law https://en.wikipedia.org/wiki/Brandolini's_law
- tzs 2y agoAdd: Cost of updating the rumor just enough so the proof it was false no longer quite applies: 1c Also, in addition to dollar costs there are time costs which similarly favor the rumor maker over the rumor debunker.
- drdaeman 2y agoSuch agency will have hard time building trust (especially given it's a government agency - modern governments aren't particularly good at building trust), and it could lose it all over a single honest mistake, or (as we're talking about conspiracies here) some sort of a poison pill dataset. Neither does censorship, of course - if anything, it fuels the conspiracy. What works is education, in general. The better the quality and versatility of education, the higher are the chances an alarm would ring in person's head when he reads some controversial statement. But quality education is an extremely hard and costly problem.
- kleton 2y agoThere exist several US federal bureaus, such as the census and labor statistics ones, that have built and maintained widespread trust in publishing accurate tabulations. It is not an impossible problem.
- drdaeman 2y ago> such as census Uh... There are plenty of people who seem to think census results are rigged in one way or another. E.g. I think I've seen that the infamous Heritage Foundation (I know, I know) has some beef with The Census bureau over some under-counting or something like that. If anything, I'd love to see more zero-trust approaches to government, where possible (voting, statistics, etc). No idea if that's going to work, but probably worth a try.
- consteval 2y agoThese are not widely trusted by the American public, at least in my own anecdotal experience. I've been told numerous times that the Bureau of Justice Statistics lies because they're the government. Typically, in relation to modern politics. From what I've seen, distrust for the government is at an all-time high, but trust in randos with podcasts is also at an all-time high.
- mrguyorama 2y agoMost Americans cannot read at a high school level. How many do you think even know what "k-anonymized" means? A spreadsheet of "cross every variable with every other variable" also does nothing but produce ample fodder for people with zero stats training to conclude that a medicine prescribed for X, causes X. Most people do not have 20 years education in anything. For an average human, you need several stats classes before you can reliably make good conclusions from data, and reliably understand how and why bad data is flawed. There is not enough time in an average human life for everyone to be meaningfully good at stats. Not letting experts be experts just literally does not scale.
- kleton 2y agoYou actually wouldn't need any stats classes to do this particular analysis, the sample sizes are quite large. You only need to know four values, the size of the groups of all vaccinated or unvaccinated, and the numerators, the number of incidents of cancer in both groups, to determine if there is a difference in the incidence rate between the two populations.