Saturday, January 20, 2018

"It doesn't happen 100% of the time, therefore it doesn't happen at all"

There's a very strange idea some people have that if there's a claim that "thing X causes disease Y", that claim is refuted by finding even one single counter-example.

I recently watched a guy, a chain smoker, say that he knows lots of people who have smoked their entire lives and have not got lung cancer, therefore smoking is not the cause of lung cancer. Due to the context and the tone of voice I'm convinced that he was being completely serious, and that he genuinely believes that. I don't think he was joking, or trolling, or anything like that.

That is, in fact, a somewhat common sentiment among some people, and chances are that you have heard it too (or at least have heard of it indirectly).

And it's not like they are saying "I don't believe the science behind it" or something like that. It's always "I know people who haven't got it, therefore smoking doesn't cause it".

Of course it doesn't work like that. You could just as well say "I know lots of people who have got influenza and didn't die of it, therefore influenza doesn't kill people." Or "I know lots of people who have been obese their entire life and they do not have diabetes, therefore obesity does not cause diabetes."

Of course "smoking causes lung cancer" is just a kind of shorthand for "smoking significantly increases the risk of lung cancer". Obviously it doesn't mean that if you smoke one cigarette you are 100% certain to get lung cancer. It's a risk, and the risk increases the more you smoke. Your chances of getting eventually lung cancer increases significantly compared to non-smokers (and this risk is really, really higher.)

Sometimes you hear the same argument in a kind of reverse form. In other words: "I know people who have never smoked in their entire lives, who got lung cancer." (You could just as well say "I know people who have never been obese who have diabetes.")

Of course it doesn't work like that either. The claim is not "only people who smoke get lung cancer". Again, it's about the risk: Non-smokers have a much lower probability of getting lung cancer (but this probability isn't zero.) Smokers have a much higher probability (but it isn't 100%.) Likewise obese people have a higher chance of developing diabetes than fit people (but this doesn't mean that fit people never, ever get diabetes. It's just much less likely.)

In either case the sentiment is just idiotic, and I have hard time understanding why people would think like that.

No comments:

Post a Comment