gwern comments on Too good to be true - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (119)
That's not a reply to what I wrote.
Yes, that's what a bias is. A systematic tendency in one direction. As opposed to random error.
And before that, they were using phages despite apparently pretty shaky evidence it was anything but a placebo. That said, pointing out the systematic bias of Russian science (among many other countries, and I'm fascinated, incidentally, how the only country you're defending like this is... your own. No love for Korea?) does not commit me to the premise that phages do or not work - you're the one who brought them up as an example of how excellent Russian science is, not me.
How many are there now? Shouldn't you have looked that up?
Difference in rates is prima facie evidence of bias, due to the disagreement. If someone says A and someone else says not-A, you don't need to know what A actually is to observe the contradiction and know at least one party is wrong.
Yes it is.
And naturally, you have not looked for anything on the topic, you just doubt it.
Strawman. No country engages in 'abandonment of normal medicine' - if you go to China, do you only find acupuncturists? Of course not. The problem is that you find acupuncturists sucking up resources in dispensing expensive placbeos and you find that the scientific community is not strong enough to resist the cultural & institutional pressures and find that acupuncture doesn't work, resulting in real working medicine being intermeshed with pseudomedicine.
Fortunately, normal medicine (after tremendous investments in R&D and evidence-based medicine) currently works fairly well and I think it would take a long time for it to decay into something as overall bad as pre-modern Western medicine was; I also think some core concepts like germ theory are sufficiently simple & powerful that they can't be lost, but that would be cold comfort in the hypothetical cargo cult scenario ('good news: doctors still know what infections are and how to fight epidemics; bad news: everything else they do is so much witch-doctor mumbojumbo based on unproven new therapies, misinterpretations of old therapies which used to work, and traditional treatments like acupuncture').
I was referring to your other comment.
That's the one I know most about, obviously. I have no clue about what's going on in China, Korea, or Japan.
Look, it doesn't matter if phages work or don't work! The treatment, in favour of which there would be strong bias, got replaced with another treatment, which would have been biased against. Something that wouldn't have happened if science systematically failed to work in such an extreme and ridiculous manner. I keep forgetting that I really really need to spell out any conclusions when arguing with you. It's like you're arguing that a car is missing the wheels but I just drove here on it.
Besides, the 90%+ proportion of positive results is also the case in the west
(also, in the past we had stuff like lobotomy in the west)
So why do you think your defense would not apply equally well (or poorly) to them? What's the phage of China?
Oh wow. What a convincing argument. 'Look, some Russians once did this! Now they do that! No, it doesn't matter if they were right or wrong before or after!' Cool. So does that mean I get to point to every single change of medical treatment in the USA as evidence it's just peachy there? 'Look, some Americans once did lobotomy! Now they don't! It doesn't matter if lobotomies work or don't work!'
You didn't drive shit anywhere.
That's on a different dataset, covering more recent time periods, which, as the abstract says, still shows serious problems in East Asia (compromised by relatively small sample: trying to show trends in 'AS' using 204 studies over 17 years isn't terribly precise compared to the 2627 they have for the USA) with the latest data being 85% vs 100%. And 100% significance is a ceiling, so who knows how bad the East Asian research has actually gotten during the same time period Western numbers continue to deteriorate...
Unless A contains indexicals that point to different things in the two cases.
(Maybe Asian acupunturists are better than European ones, or maybe East Asians respond better to acupuncture than Caucasians for some reason, or...)
( I'm not saying that this is likely, just that it's possible.)