4 ms·

Correlation does not imply causation. While lead exposure may have something to do with it, an issue like the crime rate is driven by a number of complex factor

Correlation does not imply causation. While lead exposure may have something to do with it, an issue like the crime rate is driven by a number of complex factors. Perfect example of why this type of analysis is ridiculous: http://en.wikipedia.org/wiki/Super_Bowl_indicator http://en.wikipedia.org/wiki/Super_Bowl_indicator

jackvalentine 13y ago
This is not a meaningful contribution to the conversation. If you'd read the comments and replies made before you posted perhaps you'd have been able to contribute something other than the trite "correlation does not imply causation" which may just be the most obnoxious phrase on the internet today.
- downandout 13y ago
  I strongly disagree with the article for one reason: Correlation does not imply causation. That is what I said. I realize that my comment flies in the face of the opinion of radical environmentalists. Does that make my opinion invalid?
  - jackvalentine 13y ago
    I think this post explains it without me repeating someone's words. https://news.ycombinator.com/item?id=7619054 https://news.ycombinator.com/item?id=7619054 Your "contribution" of repeating the phrase doesn't educate anybody, doesn't address the actual content of the article. It doesn't make your opinion invalid but it also doesn't contribute anything meaningful to the understanding of the article. It was also said several times before you deigned to enter the conversation and "educate" us about correlation.
    - downandout 13y ago
      It doesn't address the content of the article? Apparently you did not read the article. It specifically is trying to say that the correlation between crime rate and removal of lead from gasoline is attributable to it (e.g. caused by it). I strongly disagree with it. That's what I said.
      - midas007 13y ago
        There's plenty of more strongly correlated suggestions for the drop in crime than TEL: legalized abortion, drug trends, law enforcement, on and on. Common sense would put TEL at hypothesized causation reason # 10 or so.
        jtheory 13y ago
        There's a strange battle going on in this discussion, but honestly -- I dismiss a lot of quasi-scientific articles that put a lot of importance on correlation without addressing a ton of alternate explanations. This one, I can't dismiss so easily; I've read into the details of where this theory comes from via a few sources, and the correlations are rather more than just "this line goes up & then down where this other line also goes up & down". I'd suggest the naysayers read into this one a bit more before dismissing it (though I do understand the reaction).
  - pyre 13y ago
    Does that make my opinion invalid? Do opinions become somehow more correct just because they are contrary to the 'popular' view? Statements like: I realize that my comment flies in the face of the opinion of radical environmentalists. are not part of a constructive conversation. Did the parent post claim to be a radical environmentalist? Are you trying to passive-aggressively claim that the poster is a radical environmentalist? If neither of these are true, then what was the point of that statement?
    - downandout 13y ago
      Since my objection to the very premise of the article (that premise being that a reduction in environmental exposure to lead caused a drop in crime rates) seems to be his primary objection to my comment, yes, I am saying he or she is quite invested in radical environmentalism.
  - mixmastamyk 13y ago
    I don't care much about this article, but I'm reminded of pg's middlebrow dismissal comment.
mikeash 13y ago
Correlation does imply causation. It does not guarantee causation.
- Someone 13y ago
  Also, in this case, we have multiple cases of correlations. Throwing three sixes in a row wit a die isn't evidence that a die is loaded, but doing that 10 times out of 20 attempts? Highly suspicious. Of course, with a die, we can weigh it, find its center of gravity, cut it poem, etc. that makes it easier to 'prove' what's the cause of our extraordinary luck. Bu suppose we find that the die contains some mechanism that allows one to change its center of gravity. Then, we cannot be sure that all those '3 sixes in a row' throws were caused by a change in the center of gravity. Here, the case is more difficult, because we cannot cut open the object of study.
  - mikeash 13y ago
    Nitpick: throwing three sixes in a row is absolutely evidence that the die is loaded. It's not proof, but it's definitely evidence.
dfc 13y ago
You might enjoy this article: "Correlation does not imply causation: How the Internet fell in love with a stats-class cliché" http://www.slate.com/articles/health_and_science/science/2012/10/correlation_does_not_imply_causation_how_the_internet_fell_in_love_with_a_stats_class_clich_.single.html http://www.slate.com/articles/health_and_science/science/201...
- downandout 13y ago
  Cliche or not, in this case and many other cases, it remains true.
- etfb 13y ago
  That comment was such an epic burn, it contributed to global warming! ... Which is why, I suspect, the commenter you replied to won't believe it either. Sigh.
6d0debc071 13y ago
> Correlation does not imply causation. Okay, accepting that as a working assumption for the argument: What does? It's not things happening together that implies it, it's not things not happening together that implies it - negative and positive correlation respectively - so what is it? The only other explanation I can think of is an understanding of the underlying mechanism, but what's that understanding based on? It can't be understandings of underlying mechanisms all the way down - you have to acquire the data from observation at some point.
api 13y ago
Lead is well understood to be a neurotoxin. There's a clear mechanism of action here.
Mvandenbergh 13y ago
Correlation does imply causation, it just doesn't prove it. Multiple correlations over different data sets, in the presence of an extremely plausible mechanism - we know for a fact what lead exposure does to developing brains - is as close as we ever really get to proving a causative link in this world.

jackvalentine 13y ago

This is not a meaningful contribution to the conversation. If you'd read the comments and replies made before you posted perhaps you'd have been able to contribute something other than the trite "correlation does not imply causation" which may just be the most obnoxious phrase on the internet today.

downandout 13y ago
I strongly disagree with the article for one reason: Correlation does not imply causation. That is what I said. I realize that my comment flies in the face of the opinion of radical environmentalists. Does that make my opinion invalid?
- jackvalentine 13y ago
  I think this post explains it without me repeating someone's words. https://news.ycombinator.com/item?id=7619054 https://news.ycombinator.com/item?id=7619054 Your "contribution" of repeating the phrase doesn't educate anybody, doesn't address the actual content of the article. It doesn't make your opinion invalid but it also doesn't contribute anything meaningful to the understanding of the article. It was also said several times before you deigned to enter the conversation and "educate" us about correlation.
  - downandout 13y ago
    It doesn't address the content of the article? Apparently you did not read the article. It specifically is trying to say that the correlation between crime rate and removal of lead from gasoline is attributable to it (e.g. caused by it). I strongly disagree with it. That's what I said.
    - midas007 13y ago
      There's plenty of more strongly correlated suggestions for the drop in crime than TEL: legalized abortion, drug trends, law enforcement, on and on. Common sense would put TEL at hypothesized causation reason # 10 or so.
      - jtheory 13y ago
        There's a strange battle going on in this discussion, but honestly -- I dismiss a lot of quasi-scientific articles that put a lot of importance on correlation without addressing a ton of alternate explanations. This one, I can't dismiss so easily; I've read into the details of where this theory comes from via a few sources, and the correlations are rather more than just "this line goes up & then down where this other line also goes up & down". I'd suggest the naysayers read into this one a bit more before dismissing it (though I do understand the reaction).
- pyre 13y ago
  Does that make my opinion invalid? Do opinions become somehow more correct just because they are contrary to the 'popular' view? Statements like: I realize that my comment flies in the face of the opinion of radical environmentalists. are not part of a constructive conversation. Did the parent post claim to be a radical environmentalist? Are you trying to passive-aggressively claim that the poster is a radical environmentalist? If neither of these are true, then what was the point of that statement?
  - downandout 13y ago
    Since my objection to the very premise of the article (that premise being that a reduction in environmental exposure to lead caused a drop in crime rates) seems to be his primary objection to my comment, yes, I am saying he or she is quite invested in radical environmentalism.
- mixmastamyk 13y ago
  I don't care much about this article, but I'm reminded of pg's middlebrow dismissal comment.

mikeash 13y ago

Correlation does imply causation. It does not guarantee causation.

Someone 13y ago
Also, in this case, we have multiple cases of correlations. Throwing three sixes in a row wit a die isn't evidence that a die is loaded, but doing that 10 times out of 20 attempts? Highly suspicious. Of course, with a die, we can weigh it, find its center of gravity, cut it poem, etc. that makes it easier to 'prove' what's the cause of our extraordinary luck. Bu suppose we find that the die contains some mechanism that allows one to change its center of gravity. Then, we cannot be sure that all those '3 sixes in a row' throws were caused by a change in the center of gravity. Here, the case is more difficult, because we cannot cut open the object of study.
- mikeash 13y ago
  Nitpick: throwing three sixes in a row is absolutely evidence that the die is loaded. It's not proof, but it's definitely evidence.

dfc 13y ago

You might enjoy this article: "Correlation does not imply causation: How the Internet fell in love with a stats-class cliché" http://www.slate.com/articles/health_and_science/science/2012/10/correlation_does_not_imply_causation_how_the_internet_fell_in_love_with_a_stats_class_clich_.single.html http://www.slate.com/articles/health_and_science/science/201...

downandout 13y ago
Cliche or not, in this case and many other cases, it remains true.
etfb 13y ago
That comment was such an epic burn, it contributed to global warming! ... Which is why, I suspect, the commenter you replied to won't believe it either. Sigh.

6d0debc071 13y ago

> Correlation does not imply causation. Okay, accepting that as a working assumption for the argument: What does? It's not things happening together that implies it, it's not things not happening together that implies it - negative and positive correlation respectively - so what is it? The only other explanation I can think of is an understanding of the underlying mechanism, but what's that understanding based on? It can't be understandings of underlying mechanisms all the way down - you have to acquire the data from observation at some point.

api 13y ago

Lead is well understood to be a neurotoxin. There's a clear mechanism of action here.

Mvandenbergh 13y ago

Correlation does imply causation, it just doesn't prove it. Multiple correlations over different data sets, in the presence of an extremely plausible mechanism - we know for a fact what lead exposure does to developing brains - is as close as we ever really get to proving a causative link in this world.