Correlations between orange juice and automobile accidents?
November 11, 2013 4:03 PM
I would like to find a cite-able reference for an anecdote that I heard — drinking orange juice is correlated to a lower frequency of automobile accidents. The point of this example is that correlation is not causation by using a true correlation with a ridiculous causation. But, where did I hear of this example? Is the correlation true?
Clearly, drinking orange juice does not make safer drivers, but there are (supposedly, according to my memory) correlations between drinking orange juice and having breakfast, as well as an inverse relationship between having breakfast and being a motorcycle driver, which then have higher automobile accident rates. I swear that I did not come up with this idea, but I cannot track down where I heard it. A friend wants to use it in an upcoming speech, but he needs a reference for it (or, at least proof that the correlation is true).
Clearly, drinking orange juice does not make safer drivers, but there are (supposedly, according to my memory) correlations between drinking orange juice and having breakfast, as well as an inverse relationship between having breakfast and being a motorcycle driver, which then have higher automobile accident rates. I swear that I did not come up with this idea, but I cannot track down where I heard it. A friend wants to use it in an upcoming speech, but he needs a reference for it (or, at least proof that the correlation is true).
I'm not familiar with the orange juice/car accident anecdote, but would it help your situation if we provided other examples of spurious correlations?
Global average temperature vs number of pirates
posted by obscure simpsons reference at 4:25 PM on November 11, 2013
Global average temperature vs number of pirates
posted by obscure simpsons reference at 4:25 PM on November 11, 2013
I'm a big fan of Internet Explorer Market Share vs US Murder Rate.
posted by Jairus at 4:28 PM on November 11, 2013
posted by Jairus at 4:28 PM on November 11, 2013
This is what you're looking for.
It references this which I believe was the origin of the meme.
posted by Pogo_Fuzzybutt at 4:34 PM on November 11, 2013
It references this which I believe was the origin of the meme.
posted by Pogo_Fuzzybutt at 4:34 PM on November 11, 2013
One of my favorite random correlations of a similar nature, from the classic How to Lie with Statistics (quoted here):
"Perhaps the trickiest of them all is the very common instance in which neither of the variables has any effect at all on the other, yet there is a real correlation. A good deal of dirty work has been done with this one. The poor grades among cigarette smokers is in this category, as are all too many medical statistics that are quoted without the qualification that although the relationship has been shown to be real, the cause-and-effect nature of it is only a matter of speculation. As an instance of the nonsense or spurious correlation that is a real statistical fact, someone has gleefully pointed to this: There is a close relationship between the salaries of Presbyterian ministers in Massachusetts and the price of rum in Havana.posted by zachlipton at 4:54 PM on November 11, 2013
Which is the cause and which the effect? In other words, are the ministers benefiting from the rum trade or supporting it? All right. That's so farfetched that it is ridiculous at a glance. But watch out for other applications of post hoc logic that differ from this one only in being more subtle. In the case of the ministers and the rum it is easy to see that both figures are growing because of the influence of a third factor: the historic and world-wide rise in the price level of practically everything." (p. 90)
Another example is the correlation of ice cream sales and drowning.
The point of this example is that correlation is not causation by using a true correlation with a ridiculous causation
It is important to bear in mind that while these examples like orange juice/driving are fun, correlation is actually very good evidence of causation if there is a plausible causal link between the proposed cause and effect. For example, there is no plausible causal link between ice cream sales and drowning, but both share the plausible causal correlation of warm weather. To use an uncontroversial example, there is a positive correlation between smoking and lung disease. Most people are comfortable deeming this to be a causal link.
Too often, "correlation is not causation" is used by the speaker to dismiss causal links that they don't like. It is important to make sure that one does not let the pendulum swing too far the other way by thinking that all correlations can be safely dismissed.
posted by Tanizaki at 6:15 PM on November 11, 2013
The point of this example is that correlation is not causation by using a true correlation with a ridiculous causation
It is important to bear in mind that while these examples like orange juice/driving are fun, correlation is actually very good evidence of causation if there is a plausible causal link between the proposed cause and effect. For example, there is no plausible causal link between ice cream sales and drowning, but both share the plausible causal correlation of warm weather. To use an uncontroversial example, there is a positive correlation between smoking and lung disease. Most people are comfortable deeming this to be a causal link.
Too often, "correlation is not causation" is used by the speaker to dismiss causal links that they don't like. It is important to make sure that one does not let the pendulum swing too far the other way by thinking that all correlations can be safely dismissed.
posted by Tanizaki at 6:15 PM on November 11, 2013
Similiarly, could orange juice consumption and fewer car accidents both be caused by an increased interest in personal well-being? That is, a balanced breakfast is an example of self-care, and people with higher levels of self-care also drive more safely?
posted by charlemangy at 1:54 AM on November 12, 2013
posted by charlemangy at 1:54 AM on November 12, 2013
The Google book result is dead on, but there isn't any data in that example. Perhaps an old stats teacher had just read that book, and used the example in class? But, without cite-able data, I'll look into using a different example.
posted by Peter Petridish at 11:21 PM on November 12, 2013
posted by Peter Petridish at 11:21 PM on November 12, 2013
This thread is closed to new comments.
posted by koroshiya at 4:16 PM on November 11, 2013