8 ms·

Well I know several people for whom the Paleo diet has worked, and also alt medicine. I don't know why it works but it's unscientific to ignore the evidence sim

by analog 14y ago

Well I know several people for whom the Paleo diet has worked, and also alt medicine. I don't know why it works but it's unscientific to ignore the evidence simply because you don't understand it.

Nursie 14y ago
Anecdotes != evidence. It's unscientific to elevate your anecdotes above actual studies because you subjectively feel there's something there. This is why we have studies. I don't know why it works but it's unscientific to ignore the evidence simply because you don't understand it. Classic. Homeopathy enthusiasts always wheel this out and it's hilarious. "Just because you don't understand how it works you shouldn't ignore it!". I don't understand how it works because it doesn't! And this has been shown over and over again! I have no idea about the paleo diet, but if it seems to do good things for people we should first establish whether that is true and then find out how it's true. Alt med always misses step 1.
- analog 14y ago
  I'm sorry but anecdotes are evidence, just with a very small sample size. You can keep saying "it doesn't work" as many times as you like but you'll still be wrong.
  - Nursie 14y ago
    I'm sorry but anecdotes are evidence, just with a very small sample size. OK, anecdotes are not good evidence, and cannot be used as evidence because they suffer from nearly all the problems that a good study accounts for and real evidence gathering processes account for and mitigate. Anecdotal data suffers from confirmation bias, from selection bias, it has a small enough sample size to be meaningless, and is as a result worthless. Better? You can keep saying "it doesn't work" as many times as you like but you'll still be wrong. [Citation Needed]
    - analog 14y ago
      Right, but if you've tried doctors and they haven't been able to help, and if you have heard of a treatment which anecdotally works, and has no reported side effects, then why wouldn't you try it. If it works for you then you still have a sample size of one but this would be good enough.
      - Nursie 14y ago
        That depends on what the condition is, how trustworthy the anecdotes are and how much someone was trying to charge me for it. For instance in the case of terminal cancer, some people get ripped off by charlatans to the tune of hundreds of thousands of dollars. The anecdotal data around these assholes is all overwhelmingly positive and claims no side effects, because people who are ripped off and then die tend not to talk about it but the very few who survive for reasons unknown shout a lot. The reason you might not try it is because you'd be bankrupting the family you're going to leave behind, or funneling good-will donations to a conman. There are numerous examples of alt-med 'cures' that are dreadfully harmful (black salve, sodium hypochlorite enemas). This idea that CAM is fine to try if you're out of options because it has no side effects is itself very dodgy. Hell, even acupuncture carries some elevated risk of infection because you're pushing needles through the skin. But if it was something minor, cheap and there really were no side effects, sure - why not? --edit-- But telling if there really are no side effects does require some basis in provable fact, or you're just necking unknown quantities of unknown stuff, which may not be a great plan.
        analog 14y ago
        "That depends on what the condition is, how trustworthy the anecdotes are and how much someone was trying to charge me for it." This is true for any medical treatment. You weigh up the costs, risks, benefits and then you decide if you want to go ahead with it. There are also numerous examples of normal medicine cures that have been dreadfully harmful. Thalidomide and SSRIs to name a couple off the top of my head,
        14y ago
dalke 14y ago
Let's say that 10% of the people who switched to a paleo diet found that it "worked". (Without knowing what "worked" means, I'll assume it's "lose weight.") If you talk to those people for whom it worked, then great! there may be evidence there. However, let's say that of the people who decided lose weight through other means (a low calorie diet, more exercise, gastric bypass, whatever), 10% were successful. With this additional information, it's hard to conclude that choosing a paleo diet makes a strong difference. In other words, looking at the success cases doesn't give a conclusion, only a possible path for followup investigation. It's unscientific to say that something "works" by only looking at the successful cases. You also need to track the failure cases (those who tried paleo but where it didn't work), and the success and failure cases for other methods. The same for medicine. If someone tries X (aspirin, herbs, whiskey, zinc, penicillin, etc.) for the flu then gets better within a couple of days, then it doesn't mean that X actually "worked". There could be other reasons for why the person got better, like the person's own immune system. (For the record, none of those 'medicines' will work against the flu, though a few of them will help relieve some of the symptoms.) That's why these anecdotal accounts, while true and even heart-felt, don't suffice to make a scientific conclusion.

by analog 14y ago

Well I know several people for whom the Paleo diet has worked, and also alt medicine. I don't know why it works but it's unscientific to ignore the evidence simply because you don't understand it.

Nursie 14y ago

Anecdotes != evidence. It's unscientific to elevate your anecdotes above actual studies because you subjectively feel there's something there. This is why we have studies. I don't know why it works but it's unscientific to ignore the evidence simply because you don't understand it. Classic. Homeopathy enthusiasts always wheel this out and it's hilarious. "Just because you don't understand how it works you shouldn't ignore it!". I don't understand how it works because it doesn't! And this has been shown over and over again! I have no idea about the paleo diet, but if it seems to do good things for people we should first establish whether that is true and then find out how it's true. Alt med always misses step 1.

analog 14y ago
I'm sorry but anecdotes are evidence, just with a very small sample size. You can keep saying "it doesn't work" as many times as you like but you'll still be wrong.
- Nursie 14y ago
  I'm sorry but anecdotes are evidence, just with a very small sample size. OK, anecdotes are not good evidence, and cannot be used as evidence because they suffer from nearly all the problems that a good study accounts for and real evidence gathering processes account for and mitigate. Anecdotal data suffers from confirmation bias, from selection bias, it has a small enough sample size to be meaningless, and is as a result worthless. Better? You can keep saying "it doesn't work" as many times as you like but you'll still be wrong. [Citation Needed]
  - analog 14y ago
    Right, but if you've tried doctors and they haven't been able to help, and if you have heard of a treatment which anecdotally works, and has no reported side effects, then why wouldn't you try it. If it works for you then you still have a sample size of one but this would be good enough.
    - Nursie 14y ago
      That depends on what the condition is, how trustworthy the anecdotes are and how much someone was trying to charge me for it. For instance in the case of terminal cancer, some people get ripped off by charlatans to the tune of hundreds of thousands of dollars. The anecdotal data around these assholes is all overwhelmingly positive and claims no side effects, because people who are ripped off and then die tend not to talk about it but the very few who survive for reasons unknown shout a lot. The reason you might not try it is because you'd be bankrupting the family you're going to leave behind, or funneling good-will donations to a conman. There are numerous examples of alt-med 'cures' that are dreadfully harmful (black salve, sodium hypochlorite enemas). This idea that CAM is fine to try if you're out of options because it has no side effects is itself very dodgy. Hell, even acupuncture carries some elevated risk of infection because you're pushing needles through the skin. But if it was something minor, cheap and there really were no side effects, sure - why not? --edit-- But telling if there really are no side effects does require some basis in provable fact, or you're just necking unknown quantities of unknown stuff, which may not be a great plan.
      - analog 14y ago
        "That depends on what the condition is, how trustworthy the anecdotes are and how much someone was trying to charge me for it." This is true for any medical treatment. You weigh up the costs, risks, benefits and then you decide if you want to go ahead with it. There are also numerous examples of normal medicine cures that have been dreadfully harmful. Thalidomide and SSRIs to name a couple off the top of my head,
        14y ago

dalke 14y ago

Let's say that 10% of the people who switched to a paleo diet found that it "worked". (Without knowing what "worked" means, I'll assume it's "lose weight.") If you talk to those people for whom it worked, then great! there may be evidence there. However, let's say that of the people who decided lose weight through other means (a low calorie diet, more exercise, gastric bypass, whatever), 10% were successful. With this additional information, it's hard to conclude that choosing a paleo diet makes a strong difference. In other words, looking at the success cases doesn't give a conclusion, only a possible path for followup investigation. It's unscientific to say that something "works" by only looking at the successful cases. You also need to track the failure cases (those who tried paleo but where it didn't work), and the success and failure cases for other methods. The same for medicine. If someone tries X (aspirin, herbs, whiskey, zinc, penicillin, etc.) for the flu then gets better within a couple of days, then it doesn't mean that X actually "worked". There could be other reasons for why the person got better, like the person's own immune system. (For the record, none of those 'medicines' will work against the flu, though a few of them will help relieve some of the symptoms.) That's why these anecdotal accounts, while true and even heart-felt, don't suffice to make a scientific conclusion.