Notices
Results 1 to 46 of 46
Like Tree5Likes
  • 1 Post By skeptic
  • 1 Post By Strange
  • 1 Post By TridentBlue
  • 2 Post By scheherazade

Thread: Correlation and causation

  1. #1 Correlation and causation 
    Forum Radioactive Isotope skeptic's Avatar
    Join Date
    Nov 2008
    Location
    New Zealand
    Posts
    4,843
    One of the things that annoys the shit out of me, is when I present an argument based on a correlation, and the reply is "correlation does not equal causation". While that reply may be true, it is not an argument. Instead, it is a slogan, and argument by slogan is an invalid argument. To argue against a correlation, and make a credible claim that it does not equal causation, you need to go a bit further, and say why it does not equal causation.

    If we see a correlation, there are four possibilities. Imagine that we observe that A is correlated with B. What possibilities exist?

    1. A may cause B.
    2. B may cause A.
    3. Both A and B may be jointly caused by a third factor, C.
    4. The correlation is not correct, and we need better data.

    There is no fifth possibility. If you want to argue that a correlation is not causation, you need to point out that one of those alternate possibilities is, or is likely to be correct.

    Imagine for example, that a rooster crows at dawn. A correlation exists between the crowing and the sun rising. So, if I say that the rising of the sun causes the rooster to crow, and you want to dispute that, you need to say one of :
    A. The rooster causes the sun to rise.
    B. Both sun rising and rooster crowing are caused by a different, and third factor.
    C. The data is wrong, and the rooster does not crow at dawn.
    You cannot argue against my claim that the sun rising causes the rooster to crow by saying "correlation does not equal causation." At least, not if you want to be seen as a serious and rational debater.


    sarnamluvu likes this.
    Reply With Quote  
     

  2.  
     

  3. #2  
    Genius Duck Moderator Dywyddyr's Avatar
    Join Date
    Jan 2013
    Location
    Scunthorpe, UK
    Posts
    10,827
    Nope.
    If you're claiming that there IS causation then it's up to you to show it.
    It's not a default setting, ergo pointing out that "correlation does not imply causation" is simply saying that we're waiting to be shown that causation actually exists - it can't be assumed.


    "[Dywyddyr] makes a grumpy bastard like me seem like a happy go lucky scamp" - PhDemon
    Reply With Quote  
     

  4. #3  
    Forum Radioactive Isotope skeptic's Avatar
    Join Date
    Nov 2008
    Location
    New Zealand
    Posts
    4,843
    Dy

    Science does not work that way. There is no proof, just evidence. Scientific 'proof' works by falsifying material. Not by 'proving' a truth.

    If I pose an argument based on a correlation, that is evidence. If my debate opponent wants to oppose that view, then simply saying 'correlation is not causation' is bulldust. It is a slogan, not an argument. I agree that the correlation does not 'prove' anything. But nothing in science does that. But it does provide evidence, and contrary slogans are not acceptable as contrary arguments.
    Reply With Quote  
     

  5. #4  
    Genius Duck Moderator Dywyddyr's Avatar
    Join Date
    Jan 2013
    Location
    Scunthorpe, UK
    Posts
    10,827
    Science works by not assuming that correlation means there is causation: it has to be shown by evidence.
    "[Dywyddyr] makes a grumpy bastard like me seem like a happy go lucky scamp" - PhDemon
    Reply With Quote  
     

  6. #5  
    Brassica oleracea Strange's Avatar
    Join Date
    Oct 2011
    Location
    喫茶店
    Posts
    16,602
    Quote Originally Posted by skeptic View Post
    If I pose an argument based on a correlation, that is evidence.
    At best, it is very, very poor evidence. It might just suggest that is worth spending a bit of effort to look for some real evidence. But, really, unless you can show that it isn't due to 2, 3 or 4 then it is no more than a mildly curious random factoid.

    For instance, the linguistic diversity of a country is correlated with the number of fatalities due to traffic accidents in that country, even controlling for country nominal GDP, per-capita GDP, population size, population density, length of road network, levels of migration, whether the country is inside or outside of Africa (a strong predictor of road fatalities), distance from the equator and absolute longitude.
    Linguistic Diversity and Traffic Accidents: Lessons from Statistical Studies of Cultural Traits
    Seán Roberts, James Winters
    PLOS ONE: Linguistic Diversity and Traffic Accidents: Lessons from Statistical Studies of Cultural Traits
    MrMojo1 likes this.
    Without wishing to overstate my case, everything in the observable universe definitely has its origins in Northamptonshire -- Alan Moore
    Reply With Quote  
     

  7. #6  
    Forum Radioactive Isotope skeptic's Avatar
    Join Date
    Nov 2008
    Location
    New Zealand
    Posts
    4,843
    Quote Originally Posted by Dywyddyr View Post
    Science works by not assuming that correlation means there is causation: it has to be shown by evidence.
    Correlation IS evidence. It is one of the most common sources of evidence used in science.

    To Strange.

    Correlation is not very, very poor evidence. Often it is very potent evidence. It varies. At the very least, it is a beginning for scientific investigation. The correlation seen between smoking and lung cancer began a process that led to conclusions that are as near to cast iron as science will allow.

    None of which alters the fact that when someone says "correlation is not causation", that person is not presenting an argument. He or she is just quoting a slogan. And slogans are valueless as rational points in a debate.

    If I tell you that a rooster crows every morning at dawn, and suggest that it does so in response to the rising sun, I have presented a very potent argument. If you want to dispute the fact that it responds to the rising sun, you have to present another argument, not just quote a stupid slogan.
    Reply With Quote  
     

  8. #7  
    Forum Cosmic Wizard
    Join Date
    Dec 2013
    Posts
    2,408
    There is indeed a fifth possibility.
    The correlation might just be a random coincidence and entirely unrelated to any other factors.
    Reply With Quote  
     

  9. #8  
    Forum Radioactive Isotope skeptic's Avatar
    Join Date
    Nov 2008
    Location
    New Zealand
    Posts
    4,843
    Quote Originally Posted by dan hunter View Post
    There is indeed a fifth possibility.
    The correlation might just be a random coincidence and entirely unrelated to any other factors.
    That is the same as option 4.
    Reply With Quote  
     

  10. #9  
    Northern Horse Whisperer Moderator scheherazade's Avatar
    Join Date
    Sep 2012
    Location
    Yukon, Canada
    Posts
    4,066
    Reply With Quote  
     

  11. #10  
    Forum Radioactive Isotope skeptic's Avatar
    Join Date
    Nov 2008
    Location
    New Zealand
    Posts
    4,843
    Scheherazade

    That is probably a valid correlation. It probably follows option 3. That is : a third factor causes both.

    In this case, we see murder rates falling along with technological advances. This has been the case for not just years, but decades and centuries. See Prof. Steven Pinker's book "The Better Angels of our Nature" for details.
    Reply With Quote  
     

  12. #11  
    Brassica oleracea Strange's Avatar
    Join Date
    Oct 2011
    Location
    喫茶店
    Posts
    16,602
    Quote Originally Posted by skeptic View Post
    At the very least, it is a beginning for scientific investigation.
    That is what I said. But you cannot draw any conclusions from it because .... well, you know what I'm going to say.

    It is evidence that there might be something worth investigating. It is not evidence for any particular conclusion.

    If I tell you that a rooster crows every morning at dawn, and suggest that it does so in response to the rising sun, I have presented a very potent argument.
    That the sun rises because the cock crows?

    There is a very obvious mechanism there. And so it easy to jump from the correlation to the apparent causation. Although, even in that case, the leap would not be justified without some further evidence. ("Obvious" is the enemy of good science - another slogan for you.)

    It is not a slogan; it is an essential warning against the sort of bias that the scientific method is supposed to avoid.
    Without wishing to overstate my case, everything in the observable universe definitely has its origins in Northamptonshire -- Alan Moore
    Reply With Quote  
     

  13. #12  
    Forum Junior TridentBlue's Avatar
    Join Date
    Jan 2013
    Posts
    207
    There is no fifth possibility. If you want to argue that a correlation is not causation, you need to point out that one of those alternate possibilities is, or is likely to be correct.

    The thing is you often don't know what alternate possibility is true. For instance, consider the statement "eating rice makes your hair black, as their is a strong correlation between people who eat a lot of rice and black hair". Now its clearly false, the correlation between black hair and rice comes from people in Asia eating lots of rice, and tending to have black hair. Its clearly absurd that having black hair makes you eat rice. And variable C is this huge unknown. Why did rice become part of Asian culture, or why did a gene for black hair pop up in the same place? What's the common factor? Totally unknown, I can't tell it to you at all. But correlation here is clearly not causation.

    But in other contexts, I understand your frustration. Correlation is generally significant in the information landscape, and shouldn't be ignored.
    Reply With Quote  
     

  14. #13  
    Brassica oleracea Strange's Avatar
    Join Date
    Oct 2011
    Location
    喫茶店
    Posts
    16,602
    And I don't think anyone is saying it should be ignored. Just that we can't say which of 1 to 4 it is, without further evidence, preferably of a mechanism.
    Without wishing to overstate my case, everything in the observable universe definitely has its origins in Northamptonshire -- Alan Moore
    Reply With Quote  
     

  15. #14  
    Forum Junior TridentBlue's Avatar
    Join Date
    Jan 2013
    Posts
    207
    Yeah, you are correct. I was thinking about it a minute ago, and its really damn hard to say what defines causation in a probability landscape. I can define it in terms of the statement "A causes B" where if I observe A, I will observe B with probability 1, but I will observe B with a non-one probability otherwise. But what about cases where A often causes B? I looked it up. Wikipedia says:
    Probabilistic causation - Wikipedia, the free encyclopedia
    The central idea behind these theories is that causes raise the probabilities of their effects, all else being equal
    So we're talking about conditional probability? Then all else being equal, being born in Asia raises the probability you will have black hair, therefore being born in Asia is a cause of black hair? So blond couples can have babies in Asia to get kids with black hair? The wikipedia article has more. Apparently, something called "do operators" have to be introduced to have causation even make sense in terms of probability theory: (same link as above)

    Indeed, observing the barometer falling increases the probability of a storm coming, but does not "cause" the storm; were the act of manipulating the barometer to change the probability of storms, the falling barometer would qualify as a cause of storms. In general, formulating the notion of "probability raising" within the calculus of do-operators[4] resolves the difficulties that probabilistic causation has encountered in the past half-century.
    NOW it makes sense to me. But I did not know about that, and I bet 99% of people don't either.
    dan hunter likes this.
    Reply With Quote  
     

  16. #15  
    Forum Cosmic Wizard
    Join Date
    Dec 2013
    Posts
    2,408
    Not quite the same as 4.
    A correlation could be quite valid but still not related in any way to any kind of common cause. Gathering more data might not help at all.
    Reply With Quote  
     

  17. #16  
    Suspended
    Join Date
    Apr 2007
    Location
    Pennsylvania
    Posts
    8,822
    Quote Originally Posted by skeptic View Post

    If I tell you that a rooster crows every morning at dawn, and suggest that it does so in response to the rising sun, I have presented a very potent argument. If you want to dispute the fact that it responds to the rising sun, you have to present another argument, not just quote a stupid slogan.
    No you really haven't, because you could have just as easily said that the sun rises in response to the crowing. You are relying on other information for your potent argument, not the correlation alone.
    Reply With Quote  
     

  18. #17  
    Northern Horse Whisperer Moderator scheherazade's Avatar
    Join Date
    Sep 2012
    Location
    Yukon, Canada
    Posts
    4,066
    Hmmm...
    There are several of us who have started taking a Vitamin D supplement in response to European studies that conclude that people living at northern latitudes are at risk of deficiency, especially during the winter months. We have recently compared experiences when we found out that we had all embarked upon the same course without consulting each other.

    All of us have noticed that we feel significantly better, having more energy, less aches and pains, less given to stress over small details (even larger ones, lol) and observable improvement in hair and fingernail texture.

    Although a very small 'study group', there would appear to be a correlation between supplementing with Vitamin D and physical indicators of well-being, at least in women, because this data was only gathered from female participants. All parties were supplementing with 1000 mcg once a day in addition to whatever they might be getting from their diet.
    Reply With Quote  
     

  19. #18  
    Forum Masters Degree mat5592's Avatar
    Join Date
    Mar 2013
    Location
    West Virginia
    Posts
    601
    Quote Originally Posted by scheherazade View Post
    Reply With Quote  
     

  20. #19  
    Northern Horse Whisperer Moderator scheherazade's Avatar
    Join Date
    Sep 2012
    Location
    Yukon, Canada
    Posts
    4,066
    LOL, yes. The food industry will really be trying to prove that correlation.
    mat5592 and dan hunter like this.
    Reply With Quote  
     

  21. #20  
    Your Mama! GiantEvil's Avatar
    Join Date
    Apr 2010
    Location
    Vancouver, Wa
    Posts
    1,919
    There is a basic error in both graphs presented. Note that each individual graphs have have two different data axis plotted to time.
    In the IE/Murder graph, market share is a percentage, and murder rate is a whole number that is not a percentage of population.
    And in the Organic/Autism graph... Do I even need to explain this? People can see what's wrong with these graphs, right?

    Although being a MicroSuck hater, I find the IE/Murder graph to be emotionally satisfying.
    I was some of the mud that got to sit up and look around.
    Lucky me. Lucky mud.
    -Kurt Vonnegut Jr.-
    Cat's Cradle.
    Reply With Quote  
     

  22. #21  
    Northern Horse Whisperer Moderator scheherazade's Avatar
    Join Date
    Sep 2012
    Location
    Yukon, Canada
    Posts
    4,066
    Quote Originally Posted by GiantEvil View Post
    There is a basic error in both graphs presented. Note that each individual graphs have have two different data axis plotted to time.
    In the IE/Murder graph, market share is a percentage, and murder rate is a whole number that is not a percentage of population.
    And in the Organic/Autism graph... Do I even need to explain this? People can see what's wrong with these graphs, right?

    Although being a MicroSuck hater, I find the IE/Murder graph to be emotionally satisfying.
    I expect we could graph a correlation between internet failure and anxiety attacks, lol.

    Man, do people loose their biscuit when the grid goes down.
    I just stoke the wood stove, light a candle and grab a book while my stew is warming on the heating appliance. There is something to be said for living in the boondocks.
    Reply With Quote  
     

  23. #22  
    Forum Radioactive Isotope skeptic's Avatar
    Join Date
    Nov 2008
    Location
    New Zealand
    Posts
    4,843
    Much of what is being said here is valid. However, I stick to my statement that quoting a slogan is not an argument. If I state a correlation happens, and the debate opponent says :"correlation is not causation", that person has not made an argument. My correlation represents evidence. Certainly it is not proof, but there is no proof in science, anyway. The correlation/causation slogan is not contrary evidence, either. It is just a slogan. Certainly that slogan is often correct, but it is still a slogan, and stated by itself is a very poor argument.

    My four possibilities still stand. I may have worded possibility 4 badly. Possibility 4 really just means we got it wrong. We quoted a correlation which is not correct. There is no correlation, for whatever reason. Every genuine correlation will have validity as possibilities 1, 2 or 3. But 4 means no real correlation.

    It was pointed out to me about the cock crow/sun rising example that we could claim the cock crowing caused the sun to rise. Certainly we could, but that is obviously absurd. The sun rising causing the cock to crow is pretty obviously the correct mechanism. Possibility 1 is correct. To respond to my statement that the cock crows in response to the sun, as evidenced by the correlation, with the correlation/causation slogan, is pitiful.

    There are many examples of possibility 3 ( a third factor). For example, heavy beer drinking correlates with lung cancer. This is explainable by the fact that heavy beer drinking is associated with cigarette smoking. So factor C (smoking) is correlated with both A (beer drinking) and B (lung cancer).

    But a debate response of :"correlation is not causation" is not an argument, and is not useful. Further argument is needed.
    Reply With Quote  
     

  24. #23  
    Brassica oleracea Strange's Avatar
    Join Date
    Oct 2011
    Location
    喫茶店
    Posts
    16,602
    Quote Originally Posted by skeptic View Post
    Much of what is being said here is valid. However, I stick to my statement that quoting a slogan is not an argument.
    It depends. If you have said, "there is a correlation between A and B therefore A causes B" (and there is no other evidence) then it seems absolutely right to point out that you are making an unjustified assumption. If they choose to do that by means of a succinct cliche ... <shrug>

    However, if you are saying, "here's an interesting correlation; do you think there could be cause and effect here?" then the response is pointless (and cliched - so you win).

    Further argument is needed.
    Maybe. But sometimes the only further argument is, "we just don't (and can't) know."
    Without wishing to overstate my case, everything in the observable universe definitely has its origins in Northamptonshire -- Alan Moore
    Reply With Quote  
     

  25. #24  
    Forum Masters Degree mat5592's Avatar
    Join Date
    Mar 2013
    Location
    West Virginia
    Posts
    601
    Quote Originally Posted by GiantEvil View Post
    There is a basic error in both graphs presented. Note that each individual graphs have have two different data axis plotted to time.In the IE/Murder graph, market share is a percentage, and murder rate is a whole number that is not a percentage of population.And in the Organic/Autism graph... Do I even need to explain this? People can see what's wrong with these graphs, right?Although being a MicroSuck hater, I find the IE/Murder graph to be emotionally satisfying.
    The graphs were obviously not intended to be taken seriously.
    Reply With Quote  
     

  26. #25  
    Northern Horse Whisperer Moderator scheherazade's Avatar
    Join Date
    Sep 2012
    Location
    Yukon, Canada
    Posts
    4,066
    Quote Originally Posted by skeptic View Post

    But a debate response of :"correlation is not causation" is not an argument, and is not useful. Further argument is needed.
    Perhaps skeptic is bored this day and seeking out worthy debating partners, lol...

    (Not me...I am just stirring the pot.)
    Reply With Quote  
     

  27. #26  
    Northern Horse Whisperer Moderator scheherazade's Avatar
    Join Date
    Sep 2012
    Location
    Yukon, Canada
    Posts
    4,066
    Last year, I went nearly gluten-free for a year. I did not buy expensive gluten free products but rather just avoided those that contained gluten and reworked a few of my recipes to do gluten free baking at home. I did indentify that I experience a number of symptoms when I ate certain test foods on occasion and that these symptoms were absent when I avoided gluten.

    I am presently reading an article on the topic of gluten free diets and whether they are beneficial or harmful when something else came to mind, prompted by the article.

    For years, I have had a very pale rash on my chest that the doctors were not sure of it's nature. It caused me no discomfort. It was just there and I became accustomed to living with it.

    The thing is...it is now gone and I'm not quite sure exactly when it went away.

    Is there a correlation between my dietary changes which continue to exclude most gluten from my life and this rash going away?
    Reply With Quote  
     

  28. #27  
    Life-Size Nanoputian Flick Montana's Avatar
    Join Date
    Jul 2012
    Location
    Flatland
    Posts
    5,438
    I see what skeptic is saying. This seems absurd:



    But we do have evidence which suggests that increasing temperatures lead to more erratic behavior which can cause a rise in violent crimes. It just isn't quite the correlation you might expect.
    "Sometimes I think the surest sign that intelligent life exists elsewhere in the universe is that none of it has tried to contact us." -Calvin
    Reply With Quote  
     

  29. #28  
    Forum Radioactive Isotope skeptic's Avatar
    Join Date
    Nov 2008
    Location
    New Zealand
    Posts
    4,843
    I do not mind slogans if the person using the slogan is smart enough to offer an argument to justify the slogan.

    If someone says :"correlation is not causation" and leaves it at that, then the person is not being smart, since that is not an argument. But if there is a follow up argument, in which a reason is given to doubt the conclusion from the correlation, then that becomes a smarter argument.
    Reply With Quote  
     

  30. #29  
    Forum Junior TridentBlue's Avatar
    Join Date
    Jan 2013
    Posts
    207
    Quote Originally Posted by dan hunter View Post
    Not quite the same as 4.
    A correlation could be quite valid but still not related in any way to any kind of common cause. Gathering more data might not help at all.
    Hey Dan, Its hard to tell on this forum who's responding to who, I am assuming you are responding to me.

    Yeah, I'm not a physicist, but I remember hearing something related to the ... was it EPR experiment? Does "Bell" ring a bell in regards to that? Anyway, let me look... Ah, here is the article:
    The Experiment That Forever Changed How We Think About Reality - Wired Science
    I remember when I read it, I was struck by the lack of a hidden variable, that could be considered the "cause" of what was observed. (Expressed in this metaphor as the "hidden stamps theory") I am drinking a beer and can't formulate a clear example of it, but intuitively this seems to back up what you are saying at a very deep level.
    Reply With Quote  
     

  31. #30  
    Suspended
    Join Date
    Apr 2007
    Location
    Pennsylvania
    Posts
    8,822
    Quote Originally Posted by skeptic View Post
    I do not mind slogans if the person using the slogan is smart enough to offer an argument to justify the slogan.

    If someone says :"correlation is not causation" and leaves it at that, then the person is not being smart, since that is not an argument. But if there is a follow up argument, in which a reason is given to doubt the conclusion from the correlation, then that becomes a smarter argument.
    The trouble is that the slogan is always justified. Even if you don't have a specific reason to doubt the significance of a correlation, there can still be hidden reasons which nobody thought of yet.
    Reply With Quote  
     

  32. #31  
    Universal Mind John Galt's Avatar
    Join Date
    Jul 2005
    Posts
    14,169
    The slogan is a shorthand way of presenting the entire argument. In the absence of further evidence to support the apparent causation it is unscientific to accept it as probable. Iand see what happens. t is up to those who suspect that correlation is through cause to seek to demonstrate this, not up to others to disprove it. You assert the reverse is true. That is nonsense.

    Why does the cock crow upon the sun rising? Perhaps it is related to temperature: conduct experiments in which the temperature is varied, but the cock remains in the dark. Is it due to an inbuilt timer? Repeat experiment in the dark. Look at each other possibility and do the same. With the elimination of each, the probability that it is the appearance of the sun that causes the crowing, increases.
    Reply With Quote  
     

  33. #32  
    Forum Cosmic Wizard
    Join Date
    Dec 2013
    Posts
    2,408
    Quote Originally Posted by dan hunter View Post
    There is indeed a fifth possibility.
    The correlation might just be a random coincidence and entirely unrelated to any other factors.
    Quote Originally Posted by skeptic View Post
    Quote Originally Posted by dan hunter View Post
    There is indeed a fifth possibility.
    The correlation might just be a random coincidence and entirely unrelated to any other factors.
    That is the same as option 4.
    Quote Originally Posted by TridentBlue View Post
    Yeah, you are correct. I was thinking about it a minute ago, and its really damn hard to say what defines causation in a probability landscape. I can define it in terms of the statement "A causes B" where if I observe A, I will observe B with probability 1, but I will observe B with a non-one probability otherwise. But what about cases where A often causes B? I looked it up. Wikipedia says:
    Probabilistic causation - Wikipedia, the free encyclopedia
    The central idea behind these theories is that causes raise the probabilities of their effects, all else being equal
    So we're talking about conditional probability? Then all else being equal, being born in Asia raises the probability you will have black hair, therefore being born in Asia is a cause of black hair? So blond couples can have babies in Asia to get kids with black hair? The wikipedia article has more. Apparently, something called "do operators" have to be introduced to have causation even make sense in terms of probability theory: (same link as above)

    Indeed, observing the barometer falling increases the probability of a storm coming, but does not "cause" the storm; were the act of manipulating the barometer to change the probability of storms, the falling barometer would qualify as a cause of storms. In general, formulating the notion of "probability raising" within the calculus of do-operators[4] resolves the difficulties that probabilistic causation has encountered in the past half-century.
    NOW it makes sense to me. But I did not know about that, and I bet 99% of people don't either.
    Quote Originally Posted by dan hunter View Post
    Not quite the same as 4.
    A correlation could be quite valid but still not related in any way to any kind of common cause. Gathering more data might not help at all.
    Quote Originally Posted by TridentBlue View Post
    Quote Originally Posted by dan hunter View Post
    Not quite the same as 4.
    A correlation could be quite valid but still not related in any way to any kind of common cause. Gathering more data might not help at all.
    Hey Dan, Its hard to tell on this forum who's responding to who, I am assuming you are responding to me.

    Yeah, I'm not a physicist, but I remember hearing something related to the ... was it EPR experiment? Does "Bell" ring a bell in regards to that? Anyway, let me look... Ah, here is the article:
    The Experiment That Forever Changed How We Think About Reality - Wired Science
    I remember when I read it, I was struck by the lack of a hidden variable, that could be considered the "cause" of what was observed. (Expressed in this metaphor as the "hidden stamps theory") I am drinking a beer and can't formulate a clear example of it, but intuitively this seems to back up what you are saying at a very deep level.
    I was responding to the original poster, but yes to your understanding of what I was trying to get at.
    Reply With Quote  
     

  34. #33  
    Forum Radioactive Isotope skeptic's Avatar
    Join Date
    Nov 2008
    Location
    New Zealand
    Posts
    4,843
    Quote Originally Posted by Harold14370 View Post
    The trouble is that the slogan is always justified. Even if you don't have a specific reason to doubt the significance of a correlation, there can still be hidden reasons which nobody thought of yet.
    Harold

    That applies to everything, in science or out of it. For example, you could give the best scientific explanation ever of why global climate change is correct, and the denier says : "there can still be hidden reasons which nobody thought of yet."

    I doubt that would impress you much. Nor am I impressed by your comment above.
    Reply With Quote  
     

  35. #34  
    Forum Radioactive Isotope skeptic's Avatar
    Join Date
    Nov 2008
    Location
    New Zealand
    Posts
    4,843
    Quote Originally Posted by John Galt View Post
    The slogan is a shorthand way of presenting the entire argument.
    No. A slogan is a slogan. A lazy person's way of avoiding having to present an argument.
    Reply With Quote  
     

  36. #35  
    Universal Mind John Galt's Avatar
    Join Date
    Jul 2005
    Posts
    14,169
    Darwin seemed happy enough to adopt the slogan "survival of the fittest" as a summary of his "one long argument" of "On the Origin of Species". So I believe you to be (seriously) mistaken.
    Reply With Quote  
     

  37. #36  
    Forum Radioactive Isotope skeptic's Avatar
    Join Date
    Nov 2008
    Location
    New Zealand
    Posts
    4,843
    Darwin did not originate or use the phrase "survival of the fittest." That came from Herbert Spencer.

    Either way, it is a slogan, and a slogan is pretty much worthless. Intelligent thinking out trumps slogans.
    Reply With Quote  
     

  38. #37  
    Forum Professor river_rat's Avatar
    Join Date
    Jun 2006
    Location
    South Africa
    Posts
    1,497
    The biggest problem with pointing at a correlation between two time series is that the common cause could just be the passage of time - for example global warming and number of iPhones owned by parliamentarians would be nicely correlated but declaring that warming climate induces an irrational love of apple products seems a bit of a stretch. The correct response to someone just pointing out a correlation without some causative explanation should just be: "So what?"
    As is often the case with technical subjects we are presented with an unfortunate choice: an explanation that is accurate but incomprehensible, or comprehensible but wrong.
    Reply With Quote  
     

  39. #38  
    Forum Cosmic Wizard
    Join Date
    Dec 2013
    Posts
    2,408
    I wonder if River Rat's idea of using a dimension like time as a cause would be valid. I have a hard time imagining the passage of time as the cause of anything, except possibly aging.
    Reply With Quote  
     

  40. #39  
    Forum Masters Degree
    Join Date
    Feb 2013
    Location
    INDIA
    Posts
    548
    Quote Originally Posted by skeptic View Post
    One of the things that annoys the shit out of me, is when I present an argument based on a correlation, and the reply is "correlation does not equal causation". While that reply may be true, it is not an argument. Instead, it is a slogan, and argument by slogan is an invalid argument. To argue against a correlation, and make a credible claim that it does not equal causation, you need to go a bit further, and say why it does not equal causation.

    If we see a correlation, there are four possibilities. Imagine that we observe that A is correlated with B. What possibilities exist?

    1. A may cause B.
    2. B may cause A.
    3. Both A and B may be jointly caused by a third factor, C.
    4. The correlation is not correct, and we need better data.

    There is no fifth possibility.
    5. There may be other factor or factors causing A,B or other

    There is no sixth possibility
    "No law of Physics is surprising & can not beat commonsense until it does not give enough explanation logically or I did not understand it rightly or simply it is wrong "
    Reply With Quote  
     

  41. #40  
    Forum Radioactive Isotope skeptic's Avatar
    Join Date
    Nov 2008
    Location
    New Zealand
    Posts
    4,843
    The passage of time may, indeed, be the third factor, making it a case of possibility 3. The thing is, though, that correlation is evidence. It is not proof, but it is evidence. Evidence deserves to be considered - not shrugged off by quoting a brainless slogan.
    Reply With Quote  
     

  42. #41  
    Universal Mind John Galt's Avatar
    Join Date
    Jul 2005
    Posts
    14,169
    Quote Originally Posted by skeptic View Post
    Darwin did not originate or use the phrase "survival of the fittest." That came from Herbert Spencer.
    That is exactly why I chose the example and why I said he adopted the phrase. Adopted means to take ownership or responsibility for something not of ones own creation. If I thought Darwin had originated the phrase I would have written "Darwin seemd happy enough to coin and use the phrase......"

    Intelligent thinking recognises the depth and breadth of the implication of a slogan and saves time that would be otherwise spent stating the bloody obvious.
    Reply With Quote  
     

  43. #42  
    Forum Radioactive Isotope skeptic's Avatar
    Join Date
    Nov 2008
    Location
    New Zealand
    Posts
    4,843
    John

    Slogans in general are means of obscuring and misleading. I have been in debates on gun control and had to face the slogan :"Guns do not kill people. People kill people."

    Like other slogans, that one is correct up to a point, but used to mislead rather than enlighten. The slogan "correlation is not causation" is used in exactly the same way. Not to enlighten, but to confuse. It should not be used on its own in a science debate. If used, you need further argument to explain why, in that case, correlation is not causation.
    Reply With Quote  
     

  44. #43  
    Northern Horse Whisperer Moderator scheherazade's Avatar
    Join Date
    Sep 2012
    Location
    Yukon, Canada
    Posts
    4,066
    A recent study would seem to demonstrate a correlation between an excess of added sugar in the diet and triple the risk of heart related disease.

    Many Americans consume too much added sugar, a habit that not only increases the risk of obesity, but may also increase the risk of dying from heart disease, a new study suggests.
    Between 2005 and 2010, about 71 percent of Americans consumed 10 percent or more of their daily calories from added sugar, according to the study. The World Health Organization recommends limiting calories from added sugar to less than 10 percent of your daily total.
    What's more, people who consumed between 17 and 21 percent of their daily calories from added sugar were nearly 40 percent more likely to die from cardiovascular disease over a 14-year period than those who consumed about 8 percent of their daily calories from added sugar, the study found. [9 Snack Foods: Healthy or Not?]

    It's important to note the study found only an association, and cannot prove a cause-and-effect relationship
    between consuming added sugar and dying from heart disease, the researchers said. It's possible that other factors not accounted for in the study, were responsible for the link, they said.

    The researchers also assessed consumption of added sugar only at the study start, and it's possible that people changed their consumption over time.
    Still, the study findings contribute "to the growing body of research on sugar as an independent risk factor in chronic disease," Laura Schmidt, a professor of health policy at the University of California, San Francisco, wrote in an editorial accompanying the study. "It underscores the likelihood that, at the levels of consumption common among Americans, added sugar is a significant risk factor for [cardiovascular disease death] above and beyond its role as empty calories leading to weight gain and obesity," Schmidt said.
    Added Sugar May Boost Risk of Heart Disease, Death | LiveScience
    Reply With Quote  
     

  45. #44  
    ▼▼ dn ʎɐʍ sıɥʇ ▼▼ RedPanda's Avatar
    Join Date
    Aug 2012
    Location
    UK
    Posts
    2,737
    Quote Originally Posted by skeptic View Post
    Slogans in general are means of obscuring and misleading.
    No they aren't.
    SayBigWords.com/say/3FC

    "And, behold, I come quickly;" Revelation 22:12

    "Religions are like sausages. When you know how they are made, you no longer want them."
    Reply With Quote  
     

  46. #45  
    HTM fan
    Join Date
    Dec 2012
    Location
    U.S.
    Posts
    227
    Quote Originally Posted by dan hunter View Post
    I wonder if River Rat's idea of using a dimension like time as a cause would be valid. I have a hard time imagining the passage of time as the cause of anything, except possibly aging.
    The way things happen matters. Maybe he meant the "shape" of all 4 dimensions. The passage of time causes everything. Still, observations based on time are just associations. I think sort of simulating what happens is good evidence. If you can say A-->B-->C and fundamentally how those transitions occur (e.g. 2 proteins bind because of blah forces), and if you can explain everything in the system studied (be it test tube or dirt), then I think you have enough evidence. Then people keep testing it or perhaps recreate it to do something useful.
    "It is the ability to make predictions about the future that is the crux of intelligence."
    -Jeff Hawkins.
    For example, you can predict that 3+5=8. You can predict what sequence of muscle commands you should generate during a conversation, or whether an object is a desk or a chair. The brain is very complicated, but that is essentially how intelligence works. Instinct, emotions, and behavior are somewhat seperate.
    Reply With Quote  
     

  47. #46  
    Forum Cosmic Wizard
    Join Date
    Dec 2013
    Posts
    2,408
    Quote Originally Posted by NNet View Post
    Quote Originally Posted by dan hunter View Post
    I wonder if River Rat's idea of using a dimension like time as a cause would be valid. I have a hard time imagining the passage of time as the cause of anything, except possibly aging.
    The way things happen matters. Maybe he meant the "shape" of all 4 dimensions. The passage of time causes everything. Still, observations based on time are just associations. I think sort of simulating what happens is good evidence. If you can say A-->B-->C and fundamentally how those transitions occur (e.g. 2 proteins bind because of blah forces), and if you can explain everything in the system studied (be it test tube or dirt), then I think you have enough evidence. Then people keep testing it or perhaps recreate it to do something useful.
    No, after rereading his comment I think he was referring to time being the cause only of the correlation. I was taking his use of "cause" very literally.
    Reply With Quote  
     

Similar Threads

  1. Replies: 471
    Last Post: February 18th, 2014, 01:15 AM
  2. Correlation Causation Chocolate Nobel Prize
    By adelady in forum In the News
    Replies: 25
    Last Post: October 18th, 2013, 02:08 PM
  3. intelligence correlation
    By free radical in forum Scientific Study of Religion
    Replies: 53
    Last Post: December 18th, 2010, 03:35 PM
  4. Correlation and regression question
    By gamelogic in forum Mathematics
    Replies: 4
    Last Post: December 1st, 2008, 09:28 AM
  5. Maths project on correlation
    By Maxime Maison in forum Mathematics
    Replies: 1
    Last Post: December 3rd, 2005, 07:37 PM
Bookmarks
Bookmarks
Posting Permissions
  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •