Forgot your password?
typodupeerror
Stats The Internet

The History of 'Correlation Does Not Imply Causation' 223

Posted by Soulskill
from the does-not-imply-godwin-either dept.
Dr Herbert West writes "The phrase 'correlation does not imply causation' goes back to 1880 (according to Google Books). However, use of the phrase took off in the 1990s and 2000s, and is becoming a quick way to short-circuit certain kinds of arguments. In the late 19th century, British statistician Karl Pearson introduced a powerful idea in math: that a relationship between two variables could be characterized according to its strength and expressed in numbers. An exciting concept, but it raised a new issue: how to interpret the data in a way that is helpful, rather than misleading. When we mistake correlation for causation, we find a cause that isn't there, which is a problem. However, as science grows more powerful and government more technocratic, the stakes of correlation — of counterfeit relationships and bogus findings — grow larger."
This discussion has been archived. No new comments can be posted.

The History of 'Correlation Does Not Imply Causation'

Comments Filter:
  • Maybe (Score:5, Funny)

    by mgrivich (1015787) on Tuesday October 02, 2012 @03:34PM (#41529643)
  • However, as science grows more powerful and government more technocratic, the stakes of correlation — of counterfeit relationships and bogus findings — grow larger."

    Well is science growing powerful finds all these false correlations? Or these correlations always existed and now we know enough to say they were false. Anyway correlation is not causation.

    • Who says the correlations are false? Relationships besides A->B do exist.

    • by Lord Bitman (95493) on Tuesday October 02, 2012 @04:04PM (#41530003) Homepage

      The correlations are NOT "false". The relationships between the numbers are (almost always) NOT "conterfeit".

      "Correlation does not imply causation" means exactly that. If the sky is dark and people are carrying around umbrellas, this does NOT imply that darkness causes umbrellas, or that umbrellas cause darkness. The causal relationship between two numbers is not determined by how often one number changes at the same time as another.

      To put it another way: correlation is an *observed* behaviour, causation is a *tested* behaviour.

      • by vlm (69642)

        To put it another way: correlation is an *observed* behaviour, causation is a *tested* behaviour.

        Nice, but how bout correlation is a math formula, on the other hand causation has a ten page philosophical wikipedia page and even though Hume died like 300 years ago this year people are still arguing about it, with the exception that on the internet everyone agrees correlation isn't it, which I guess is at least some progress.

        • by Kjella (173770) on Tuesday October 02, 2012 @06:40PM (#41531931) Homepage

          Nice, but how bout correlation is a math formula, on the other hand causation has a ten page philosophical wikipedia page and even though Hume died like 300 years ago this year people are still arguing about it, with the exception that on the internet everyone agrees correlation isn't it, which I guess is at least some progress.

          The reason causation has a ten page philosophical page is that on the macro scale everything is a result of and happen in conjunction with a gazillion butterfly effects that were either present or absent, in fact the physical article is quite short. Imagine for a murder every detail that happened in both the murderer's and victim's life who lead them there, they're all causally necessary but we put the blame on the killer. Not the policeman who forgot his bulletproof west at home or the kids who teased the murderer in third grade or the parents for conceiving him. It tries to give weight and quality to those causes that depends on the state of mind, say attempted arson is a lot more "dominating" relative to poor fire safety than an accident even if the fire is physically identical. And it all depends on how well the person could predict or control the chain of effects set in motion.

          Causality is easy. Causal responsibility - which by the way is not just about assigning blame, but also things like credit - is very hard. For example, you are hanging off a cliff and another person clings to you but you can't hold on. Either you kick him off so he falls to his death or you both fall to your deaths. In no case is there a question of physical cause and effect, but would you philosophically cause his death by kicking him off or was he dead either way? What if you can hold on another minute, is that murder? Five minutes? Fifty years? I mean he's human, he's eventually going to die - you're not really changing the outcome. What if you're 100% sure you can pull yourself up, but only 99,9% sure you'll both fall to your deaths and a 0,1% chance that you through superhuman strength will pull you both up? It's hard not to get philosophical.

    • by jellomizer (103300) on Tuesday October 02, 2012 @04:17PM (#41530157)

      Correlation may not lead to causation... However it tends to give a clue on the causation.

      For example a Correlation between the number of tattoos vs. the number of Motorcycle accidents.
      Well ink in your skin doesn't cause you to get in an accident. However people who are more apt to taking risks will more likely get a tattoo. People who take more risks get into accidents more.

      In terms of policy, you want to reduce motorcycle accidents, telling people you need to stop getting tattoos will not be effective. However with this correlation you may get results by posting motorcycle safety information at the tattoo parlors.

      But using Correlation != causation as a way to short circuit an argument isn't that effective. Because if your goal is to dig for the truth or a solution, the correlation is important, and if the correlation seems reasonable to create the causation it is worth further investigation.

      • This.

        My shrink LOVES to pull the "Correlation is not causation" trite, but when I take a new pill and symptoms pop up, and they cease after I quit taking it, I don't care what your ideologies are. If the effect plays out like this, you need to prove to me that this correlation is not causation, not the other way around.
      • by PhotoJim (813785) <jim@phot o j im.ca> on Tuesday October 02, 2012 @05:37PM (#41531317) Homepage

        My statistics professor called these "lurking variables". Something might exist to cause both elements in a correlative relationship, but if it's not being considered in the analysis, the analysis of the correlation will be misleading. Yours is a great example.

    • Well is science growing powerful finds all these false correlations?

      Think of it this way: When you find a correlation, there are four possibilities: A causes B, B causes A, C causes A and B, or chance. Repeating experiments helps strengthen the correlation, which diminishes the probability of chance. Further experiments varying those parts of A and B that can be controlled help distinguish among the remaining three and help identify C.

    • by ceoyoyo (59147)

      The summary is making a common mistake (particularly common among people who like to parrot "correlation does not imply causation!" in arguments. Correlation does not imply causation has nothing to do with whether a correlation is real or not. A true correlation implies that there IS a causal link, it just doesn't specify which of three general types it might be (reverse, forward or common).

  • by sackvillian (1476885) on Tuesday October 02, 2012 @03:35PM (#41529661)
    In what sense, exactly does science grow more powerful? In my experience, sciences grows more expensive, less funded, more hyped, less understood, and overall less heeded.
    • by flaming error (1041742) on Tuesday October 02, 2012 @03:43PM (#41529769) Journal

      > In what sense, exactly does science grow more powerful?

      Space Stations. Tsunami warning systems. Earthquake warning systems. Cochlear implants. Big Dog. Spirit & Opportunity. Curiosity. Exoplanets. Higgs Boson.

    • Science grows more powerful as an explanatory tool as we grasp more of the world around us. Philosophical or logical power.

      If you have premise p which is "science", the set of things you can derive or contradict from that grows quite rapidly.

    • by Nyder (754090)

      In what sense, exactly does science grow more powerful? In my experience, sciences grows more expensive, less funded, more hyped, less understood, and overall less heeded.

      Cosmic & Gamma rays dude. Don't you read Marvel Comics?

    • by scamper_22 (1073470) on Tuesday October 02, 2012 @03:54PM (#41529901)

      We're the most scientific society the world has ever known.
      Science holds more power today that at any point in history.

      Now you can say, it is 'not true science' in as much as people can say Saudi Arabia or Iran is not 'true Islam'.

      Some abstract notion of science or religion as *truth*.

      But back in reality.

      People who say science is their guide are at the most powerful in history. Regular people walk around saying 'we need independent scientific bodies to set healthcare, education policy. People readily accept the truth from scientific panels without much understanding of the actual science. Institutions of science are well funded by government as is the education systems (Relative to most other times in history).

    • You mean the more questions we ask, the more questions we find we need to ask? Toke another one buddy. This is what is called progress.
  • by fahrbot-bot (874524) on Tuesday October 02, 2012 @03:40PM (#41529717)

    ... and is becoming a quick way to short-circuit certain kinds of arguments.

    ... Correlation does not imply causation.

    • by Tackhead (54550) on Tuesday October 02, 2012 @04:02PM (#41529975)

      ... and is becoming a quick way to short-circuit certain kinds of arguments.

      ... Correlation does not imply causation.

      The decline in classical education standards is, however, a causal factor behind the shift from references to the "post hoc ergo propter hoc fallacy" towards references to the phrase "correlation does not imply causation".

      • by ceoyoyo (59147)

        Yes, a decline in the teaching of dead languages and vague logical fallacies, and the increase in teaching of formal statistical techniques and their logical interpretation, is a causal factor in a shift from references to a vague Latin phrase to a (probably equally misunderstood but more precise) English one.

    • by Cinder6 (894572)

      It does imply causation. It doesn't prove it, which is why saying that cause is there is irresponsible. Correlation merely states that one thing might cause another, so further study may be warranted.

      • by Cinder6 (894572)

        I suppose I should mention I'm using the definition of "imply" that means "suggest"; as has been mentioned below me, "imply" has different meaning in mathematics.

      • by ceoyoyo (59147)

        Nope. You're closer to the truth than the average Slashdotter, but not quite there.

        Correlation absolutely implies causation. It doesn't specify the precise form of causation. If there is a correlation between two things, there is one of three general types of causal relationship between those two things.

  • by argStyopa (232550) on Tuesday October 02, 2012 @03:42PM (#41529751) Journal

    Correlation doesn't PROVE causation.... ...but it bloody well DOES suggest it, at least in the course of our daily lives.

    The reason this phrase is so catchy is that it's counter-intuitive, and easily proven to be true. People love to use it as a "gotcha" phrase, PRECISELY because in regular life correlation does in fact usually imply causation.

    In fact, correlation is used by most scientists to begin the hypothesis process. A power plant is built on a river, and the river starts drying up - most people would begin their analysis by checking on the power plant, and not the population of honeybees.

    Your kid is alone in the kitchen. The cookie jar is (now) empty. Does his presence CONCLUSIVELY PROVE that he ate the cookies? Of course not, and a wise parent would find other evidence to draw a conclusion. But the correlation of their places in time and space, as well as a known predilection for cookies means that correlation strongly suggests an avenue of investigation (you're probably not going to start figuring out what happened by pursuing some other entirely different course).

    It's the sort of empty-headed 'gotcha' phrase that's so popular and so often used without real thought behind it.

    • The main problem that the phrase attempts to solve these days is with data-mining. If you have a huge dataset, and start pulling random trends out of it, there's a decent likelihood that some of those trends will correlate. But you have no real evidence that one caused the other - the correlation is just as likely to be random chance. You didn't start by finding a problem which you need to find a cause for: you started with 'What's going on?' You found that out, but it's tempting to start assigning prob

      • by Hentes (2461350)

        Exactly, this is a very big problem with social "sciences" that are almost purely empirical and the only theory they know (in the good case) is statistics. Without a theoretical foundation it's almost impossible to detect when correlation is misleading: the 95% rule is not a replacement for a scientific model. When a physics experiment yield results suggesting faster-than-light particles scientists knew that it was likely an error because of the field's theoretical background. These checks aren't present in

    • by flaming error (1041742) on Tuesday October 02, 2012 @03:54PM (#41529899) Journal

      Correlation suggests only Correlation. It doesn't suggest causation, but as you noted, it does suggest areas for further investigation. The relationship may or may not turn out to be directly causal.

      • by ceoyoyo (59147)

        Correlation doesn't suggest causation, it "proves" it. Correlation implies causation. Just not the type.

        If there is no causal relationship between two things, there is no correlation between them either. There may APPEAR to be a correlation, but there isn't actually one.

    • by zolltron (863074)

      The reason this phrase is so catchy is that it's counter-intuitive, and easily proven to be true. People love to use it as a "gotcha" phrase, PRECISELY because in regular life correlation does in fact usually imply causation.

      I agree, and this cannot be overstated. I worry that the use of this phrase is almost more dangerous than the mistaken belief that correlation does imply causation.

      To be precise, in most of the examples people love to trot out, correlation does imply causation, just not direct causation. A and B might be correlated because they are both caused by the same thing. While a correlation between obesity and TV watching doesn't imply that TV watching causes obesity, the correlation is good evidence that one cau

    • by arth1 (260657)

      Your kid is alone in the kitchen. The cookie jar is (now) empty. Does his presence CONCLUSIVELY PROVE that he ate the cookies?

      Um, where's the correlation here?

      Presence of a kid in the kitchen correlates inversely to number of cookies in jar?
      Why, then the conclusion is obvious - a dearth of cookies causes kids.

    • Your kid is alone in the kitchen. The cookie jar is (now) empty. Does his presence CONCLUSIVELY PROVE that he ate the cookies? Of course not, and a wise parent would find other evidence to draw a conclusion. But the correlation of their places in time and space, as well as a known predilection for cookies means that correlation strongly suggests an avenue of investigation (you're probably not going to start figuring out what happened by pursuing some other entirely different course).

      The kid there at the ti

    • by jemenake (595948)

      Correlation doesn't PROVE causation.... ...but it bloody well DOES suggest it,

      ... or it could suggest that there's a third cause, right? Like when Steven Leavitt mentions that people used to think that ice-cream caused polio because of some correlation. Turns out that the correlation was due to the fact that, when it got hot in the Summer, people would: 1) eat ice-cream and 2) go swim at the local swimming hole (where they'd get polio).

      Now, I take it that your point (about the "suggestion" of causation) is that it gives us reason to pursue, further, investigation into a possibly l

    • by poity (465672)

      It only suggests a simultaneous occurrence. If B happens when A happens, it may be that A causes B, but it could just as likely that an unknown C is the cause of both A and B. We can only say that there is a relationship, and that an observed change in A can help us better predict a change in B, nothing more.

    • Correlation doesn't PROVE causation.... ...but it bloody well DOES suggest it, at least in the course of our daily lives.

      The reason this phrase is so catchy is that it's counter-intuitive, and easily proven to be true. People love to use it as a "gotcha" phrase, PRECISELY because in regular life correlation does in fact usually imply causation.

      No, you have it exactly backwards. Causation usually implies correlation(*). But there are lots of correlations that are not in any way causally related, such as a) the decline in piracy and increase in global temperatures in the past two centuries; b) the rabbit population in Australia and the performance of the London stock exchange for the past century; or c) the monthly per capita consumption rate of ice cream and the monthly per capita rate of drowning deaths for seaside locations north of 40 degree

    • by Kjella (173770)

      I think you've drawing the conclusion the wrong way, because we have a vague idea of the causal relationship the correlation is collaborative evidence. There's lots and lots of things that correlate that my mind dismisses because it's absurd or there's obviously an underlying cause for it. I would say my investigation of the missing cookies would be far more causal - "Why are they gone? Probably because someone has eaten them. Who likes cookies? Who had access to the cookies? Who do I know has been in their

  • by EvilNTUser (573674) on Tuesday October 02, 2012 @03:45PM (#41529799)

    The people who mindlessly deny the possibility of causation are worse than those who compare everything to Hitler.

  • by pla (258480) on Tuesday October 02, 2012 @03:53PM (#41529887) Journal
    and is becoming a quick way to short-circuit certain kinds of arguments

    The real problem here comes from people using that as a "short cut" to an actual argument.

    On the one hand, we've done a great job at getting them to grasp that correlation does not imply causation. Now, we need to get people to understand what does - Necessary and Sufficient.

    Next time someone uses that as a catch-phrase to shoot down a correlation as meaningless, ask them:
    Does B require A? Necessary.
    Does A lead to B? Sufficient.
    QED, A causes B (or vice-versa).

    Of course, my choice of the word "meaningless" there carries its own problems - Using correlation vs causation as a rhetorical shortcut to actual logic glosses over the fact that (statistically significant) correlations can have meaning (just that they don't "mean" causation). FWIW, The vast majority of modern medicine involves dealing with correlations rather than causes - "depressed people have low serotonin, prozac increases available serotonin", "people with high cholesterol have more heart attacks; lipitor reduces cholesterol". You can often use a correlation, as long as the two sides actually do link via some unknown variables. When they don't, however - Well, pirates don't prevent global warming because adding more pirates to the world doesn't somehow put us back before the industrial revolution.
    • and is becoming a quick way to short-circuit certain kinds of arguments The real problem here comes from people using that as a "short cut" to an actual argument. On the one hand, we've done a great job at getting them to grasp that correlation does not imply causation.

      Emphasis mine. Correlation DOES suggest causation though as many here have already argued. It just doesn't prove/denote/equal it. Or to put it more in Slashdot terms correlation =/ causation. But it does imply, that's usually the basis of the first step in investigation.

    • Next time someone uses that as a catch-phrase to shoot down a correlation as meaningless, ask them:
      Does B require A? Necessary.
      Does A lead to B? Sufficient.
      QED, A causes B (or vice-versa).

      B is a flooded basement.
      A is an overflowing washing machine.
      N is a crack in the foundation allowing in water.

      Multiple paths can lead to B. So, even if B doesn't require A, A can cause B. But N can cause B too.

    • by Artifakt (700173)

      If we add enough pirates, and they are temporarily successful enough at stopping international trade, rading coastal settlements, and general rape and pillage, wouldn't putting civilization back before the industrial revolution, except in isolated pockets far inland, be a distinct possibility?

  • Dating advice (Score:5, Insightful)

    by 3ryon (415000) on Tuesday October 02, 2012 @04:01PM (#41529961)

    I had the phrase "Desired: A woman who understands that correlation does not imply causality..." in my dating profile.

    I married the woman who replied. Yes, I am surprised that worked as well.

    • by arth1 (260657)

      I go one step further, and require a basic understanding of Lorentz transformation.
      And don't all of you girls here run down my mailbox, now...

    • So, what traits did you correlate with that trait? ;)
  • TFA does a pretty good job of explaining why. Here's something I'd like to add: no, correlation does not imply causation, in the strict mathematical meaning of "imply"; in mathematical parlance, "A implies B" means that if A is true, B will always be true as well, and of course "X is positively correlated with Y" does not mean "an increase in X causes an increase in Y." But there's another meaning of "imply," and, like the common confusion about the meaning of the word "theory" in creationist arguments,

    • by ceoyoyo (59147)

      TFA is wrong. Correlation implies causation. The phrase, which is inaccurate as so many sound bite phrases are, is trying to say that A being correlated with B does not imply that A causes B. It DOES imply that there is a causal relationship between A and B, it's just not specific as to the type.

  • Here's the link to page 1 for the extremely laz—editors: [Filter error: That's an awful long string of letters there.] [slate.com].

  • But it gets the best odds in Vegas.
  • We know fossil fuel use is on the rise. We know the earth is getting warmer. So you HAVE to see the FACT that people are using more fuel to run their air conditioners precisely BECAUSE it's hotter these days. Warming causes increased energy usage. duh.


    That was a joke son, I'm not trolling....
  • by PopeRatzo (965947) on Tuesday October 02, 2012 @04:24PM (#41530281) Homepage Journal

    The most recent popularity of "correlation does not indicate causality" is the result of the rise of anti-intellecutalism and anti-reason. It's something that stupid people say to try to sound smart, and to deny data.

    Correlation is not proof, but if you see replicable continual correlation, ignoring it is dumb.

    It comes from people who try to use an 18th century view that Science "creates facts", instead of "creates models that either are supported by observation or are not supported by observation". Correlation is just another observation.

    It's one of the big favorites of the anti-intellectual Right and climate change deniers.

    • It comes from people who try to use an 18th century view that Science "creates facts", instead of "creates models that either are supported by observation or are not supported by observation".

      That or "science creates models, but I can't see how these models are so useful in the daily lives of those around me, so I refuse to endorse borrowing money from China and Japan to fund creating these toy models."

  • I always respond to that this way: "But causation does imply correlation. Since we can't directly see causes (if we could, we wouldn't be investigating looking for them) and we need something that we can see to tell us where to start looking, correlation is as good a starting point as we're going to get.".

  • The argument is "Cum Hoc Ergo Propter Hoc" as described in Latin. The similar phrase "Post Hoc, Ergo Propter Hoc" or "after this therefore because of this" dates back to 1704, according to Merriam Webster. I would assume "Cum Hoc Ergo Propter Hoc" is of similar age and origin.

    • This is actually the most reasonable definition of "causation" for the following values:

      "Hoc"= state of the Universe at a given instant

      "Post Hoc"= state of the Universe at (given instant + infinitesimal interval)
  • by Beeftopia (1846720) on Tuesday October 02, 2012 @04:40PM (#41530503)

    Statistics can ONLY show the degree of correlation. Statistics can never show causation. So, all you're ever going to get from statistics is correlation.

    That reality escapes many.

    References:
    1) Although [statistical] regression cannot prove causation, no statistical method can do that, [harvardlawreview.org]

    2) Epidemiological studies can never prove causation; that is, it cannot prove that a specific risk factor actually causes the disease being studied. Epidemiological evidence can only show that this risk factor is associated (correlated) with a higher incidence of disease in the population exposed to that risk factor. The higher the correlation the more certain the association, but it cannot prove the causation. [cornell.edu]

    The reality is that statistics can ONLY show you whether there is a correlation or not, and how strong it is. Then it requires other methods to suggest whether there is a causative relationship.

    • by ceoyoyo (59147)

      Nonsense. Statistical analysis of observational data can show correlation, which implies a causal relationship but can't specify the nature of that relationship. Statistical analysis of experimental data, where you manipulate a variable in a controlled way, can show correlation with the manipulation, which, if the experiment is done properly, narrows down the choices of causal relationship to one.

  • by Cyno01 (573917) <Cyno01@hotmail.com> on Tuesday October 02, 2012 @04:58PM (#41530797) Homepage

    I would argue that correlation absolutely IMPLIES causation, but does not PROVE causation.

    • Careful there.

      Implication is conditional [wikipedia.org], but that is the only difference between implication and proof.

      A = correlation
      B = causation

      "A imples B" is the same as "B or not A" (see the linked article). So your first clause is the same as "there is causation, or there is no correlation". Then, if we grant that there is correlation, it follows that causation is proven, which contradicts the second clause of your statement.

      I think what you meant is that "correlation is evidence of causation". This is different fr

  • No, but only because of Betteridge.
  • First, correlation: If A then B is more likely than if not A.

    Then Causation: If first A then B, all the time.

    It is pretty clear that these two statements do not have any implication. There is nothing in Correlation's definition at all about 'first', which determines whether A causes B or B causes A. In addition, it is clearly worded to avoid the definitive 'all the time' which is necessary for causation.

    Basically, Causation causes Correlation, but not the other way around. It is exactly as likely

  • by Thyamine (531612) <thyamine@o f d ragons.com> on Tuesday October 02, 2012 @05:17PM (#41531069) Homepage Journal
    It is absolutely true that correlation does not imply causation, but people seem to use it (especially on here) as if it magically refutes everything. Usually more so when they don't want it to be true, or just don't want to deal with it.

    I hit you with a bat.
    You are bleeding on the ground.
    'But.. but.. correlation does not imply causation.. Maybe I started bleeding spontaneously...'
  • The funniest part is where the author claims that throwing out the phrase stops arguments in their tracks. I guess he doesn't spend much time around here.
  • by Greyfox (87712) on Tuesday October 02, 2012 @07:12PM (#41532189) Homepage Journal
    We can still burn witches for being left-handed, right?
  • by anwyn (266338) on Tuesday October 02, 2012 @07:50PM (#41532505)
    absence of evidence is not evidence of absence.

    This reply usually confuses them enough to go away.

1 Dog Pound = 16 oz. of Alpo

Working...