5 ms·

May I ask, completely sincerely, what approach you would settle on for trying to discriminate truth from fiction if you believed you couldn't get reliable infor

May I ask, completely sincerely, what approach you would settle on for trying to discriminate truth from fiction if you believed you couldn't get reliable information from reading 30-odd studies?

astrange 5y ago
Reading 30 studies certainly seems to be a bad way to learn anything unless you're conducting a meta-analysis of them. For one thing, you'd get tired. But just shoving 30 pieces of information into your head and forgetting some of them doesn't help you find out which of them were correct.
- nomel 5y ago
  > For one thing, you'd get tired. Suggesting that there's no way someone could have the mental stamina to read 30 papers, over the course of two years, is completely ridiculous, insulting, and invalidates your viewpoint, in my opinion.
  - astrange 5y ago
    I was talking about all at once. If it’s over years I would rather wait for a meta analysis, but it depends on the reliability of the early papers.
    - nomel 5y ago
      Why are you assuming they tried to read it in one sitting? Nothing remotely suggests that in the comment thread. All we know is that they read 30 papers sometime within the last two years or so. You chose a perspective, and just reinforced it, that they were incapable of extracting anything useful from reading those papers, which is insulting. It could be that they read a paper every couple of days, or perhaps when they became available. For some reason, it seems you've already decided that they're incompetent.
- jacquesm 5y ago
  Especially in a field in which you are not trained, do not have the background knowledge required to understand the various qualifications and established best practices in. That gap can be absolutely enormous, even for fields that are peripherally related. Specialization is there for a reason, the body of knowledge that we now command is way too large for any single individual to absorb to a substantial depth.
qqqwerty 5y ago
Isn't the correct answer to wait for more reliable studies? Which is what most "anti-ivermectin" folks were saying to begin with? If your primary goal is to have a miracle cure in hand that would allow you to push back against vaccine mandates, then this approach probably won't sit well. But if you are interesting in actually finding truth, than waiting for larger, more robust studies was the answer.
- csee 5y ago
  Why wait? Why not take ivermectin as a precautionary therapeutic given it is well tolerated with low risk of side effects? There's a chasm between the epistemic question of "does this drug really work?" and the pragmatic question of "does a risk-reward analysis under uncertainty say that I should take this mostly safe drug that might slightly reduce my covid severity?" FWIW, I wouldn't take ivermectin either as a prophylactic or therapeutic, the evidence just isn't strong enough, but I don't really fault anyone that would add it into the stack of treatments upon a diagnosis in the hopes of a small net benefit.
  - jacquesm 5y ago
    > given it is well tolerated with low risk of side effects That's not a given, that depends on the patient, the dose and a host of other factors. You may well end up harming someone significantly. You are in no position to proscribe any kind of medication to someone else unless you happen to be a licensed practitioner.
    - csee 5y ago
      "That's not a given" I never said it was a given. You're using another definition of the word "given" in order to argue. "You are in no position" Who said I was in a position to be making that decision? That's completely irrelevant.
    - kragen 5y ago
      While it's true that virtually any substance has a toxic dose, avoiding that dose is straightforward for drugs as well-known and widely-used as ivermectin. There is no risk of harming someone significantly. (There's also, I think, no chance of being effective against covid at safe doses.) It is true, but unfortunate, that in many societies only licensed practitioners are in a position to prescribe (or proscribe) most drugs. That's a policy that does some good but also an enormous amount of harm.
dekhn 5y ago
In short: papers lie, and grad school teaches you to read around the lies. Papers lie because people are competitive (you lose your job if you aren't publishing), want their narrative to be true (don't we all), and the people reading the papers professionally know how to interpret them. Pretty sad, but that's what we've ended up with. It was incentivized for.
kragen 5y ago
If the 30-odd studies are chosen adversarially to persuade you to believe something that isn't true, probably asking someone more trustworthy for relevant studies would be a good idea. If you're not familiar with the field the studies are in, it might be a good idea to rely on the opinion of someone who is; unfortunately, this means you have to judge whose opinion to rely on, which puts you in the position of trying to guess who's really familiar with the field. It's going to be hard for you to do better at this than just believing whatever is most popular. Scott Alexander wrote a post about this actually: https://slatestarcodex.com/2019/06/03/repost-epistemic-learned-helplessness/ https://slatestarcodex.com/2019/06/03/repost-epistemic-learn...

May I ask, completely sincerely, what approach you would settle on for trying to discriminate truth from fiction if you believed you couldn't get reliable information from reading 30-odd studies?

astrange 5y ago

Reading 30 studies certainly seems to be a bad way to learn anything unless you're conducting a meta-analysis of them. For one thing, you'd get tired. But just shoving 30 pieces of information into your head and forgetting some of them doesn't help you find out which of them were correct.

nomel 5y ago
> For one thing, you'd get tired. Suggesting that there's no way someone could have the mental stamina to read 30 papers, over the course of two years, is completely ridiculous, insulting, and invalidates your viewpoint, in my opinion.
- astrange 5y ago
  I was talking about all at once. If it’s over years I would rather wait for a meta analysis, but it depends on the reliability of the early papers.
  - nomel 5y ago
    Why are you assuming they tried to read it in one sitting? Nothing remotely suggests that in the comment thread. All we know is that they read 30 papers sometime within the last two years or so. You chose a perspective, and just reinforced it, that they were incapable of extracting anything useful from reading those papers, which is insulting. It could be that they read a paper every couple of days, or perhaps when they became available. For some reason, it seems you've already decided that they're incompetent.
jacquesm 5y ago
Especially in a field in which you are not trained, do not have the background knowledge required to understand the various qualifications and established best practices in. That gap can be absolutely enormous, even for fields that are peripherally related. Specialization is there for a reason, the body of knowledge that we now command is way too large for any single individual to absorb to a substantial depth.

qqqwerty 5y ago

Isn't the correct answer to wait for more reliable studies? Which is what most "anti-ivermectin" folks were saying to begin with? If your primary goal is to have a miracle cure in hand that would allow you to push back against vaccine mandates, then this approach probably won't sit well. But if you are interesting in actually finding truth, than waiting for larger, more robust studies was the answer.

csee 5y ago
Why wait? Why not take ivermectin as a precautionary therapeutic given it is well tolerated with low risk of side effects? There's a chasm between the epistemic question of "does this drug really work?" and the pragmatic question of "does a risk-reward analysis under uncertainty say that I should take this mostly safe drug that might slightly reduce my covid severity?" FWIW, I wouldn't take ivermectin either as a prophylactic or therapeutic, the evidence just isn't strong enough, but I don't really fault anyone that would add it into the stack of treatments upon a diagnosis in the hopes of a small net benefit.
- jacquesm 5y ago
  > given it is well tolerated with low risk of side effects That's not a given, that depends on the patient, the dose and a host of other factors. You may well end up harming someone significantly. You are in no position to proscribe any kind of medication to someone else unless you happen to be a licensed practitioner.
  - csee 5y ago
    "That's not a given" I never said it was a given. You're using another definition of the word "given" in order to argue. "You are in no position" Who said I was in a position to be making that decision? That's completely irrelevant.
  - kragen 5y ago
    While it's true that virtually any substance has a toxic dose, avoiding that dose is straightforward for drugs as well-known and widely-used as ivermectin. There is no risk of harming someone significantly. (There's also, I think, no chance of being effective against covid at safe doses.) It is true, but unfortunate, that in many societies only licensed practitioners are in a position to prescribe (or proscribe) most drugs. That's a policy that does some good but also an enormous amount of harm.

dekhn 5y ago

In short: papers lie, and grad school teaches you to read around the lies. Papers lie because people are competitive (you lose your job if you aren't publishing), want their narrative to be true (don't we all), and the people reading the papers professionally know how to interpret them. Pretty sad, but that's what we've ended up with. It was incentivized for.

kragen 5y ago

If the 30-odd studies are chosen adversarially to persuade you to believe something that isn't true, probably asking someone more trustworthy for relevant studies would be a good idea. If you're not familiar with the field the studies are in, it might be a good idea to rely on the opinion of someone who is; unfortunately, this means you have to judge whose opinion to rely on, which puts you in the position of trying to guess who's really familiar with the field. It's going to be hard for you to do better at this than just believing whatever is most popular. Scott Alexander wrote a post about this actually: https://slatestarcodex.com/2019/06/03/repost-epistemic-learned-helplessness/ https://slatestarcodex.com/2019/06/03/repost-epistemic-learn...