Correlation =/= Causation: Correlation Does Not Equal Causation

Mistaking correlation for causation is an everyday fallacy in which we are constantly bombarded with by the media, well meaning individuals, and our own brains. Causation is the action of one thing directly rising an action in one or more other variables. Correlation is a relationship between two or more variables. Although a causation between one variable and another requires a correlation, a correlation in no way implies a causation.

For instance, if Jim has lost some weight since he cut his diet to a normal level and began regular exercise, there is very likely some causation at play between Jim's new diet and exercise and his recent weight loss. If, however, one were to point out that Jim had lost some weight since his car broke down, we simply have a correlation. There no way of knowing whether the weigh loss is related to having to walk more or any number of other factors.

Although most people make mistakes assuming a causation where there is simply a correlation, the fallacy is most insidiously used by the media through the influence of public relations, advertising, and political interests. These forces intentionally manipulate statistics and other findings that correlate with their intended message and announce to the public that a direct causation has been made. Because there's often a chance that the correlation is actually a causation, these intentional fallacies can be difficult for even the most seasoned cynic to notice. The statements are usually dropped in such a matter of fact way that it is easy for the brain to just accept the intended but unsubstantiated implication (i.e. "Boy, Jim sure has lost some weight since his car broke down, huh?").

Examples of correlation being claimed as causation intentionally can readily be found in arguments and claims put forth in business and political arenas. For instance, taking a relatively weak correlation of a product being in used in correlation to desired results is a common ploy made by many marketers of snake oil and various pseudoscientific products. For example, if a company does a study with a herbal weight loss supplement and discovers that over half of the participants lost two or three pounds while taking the supplement in addition to proper diet and exercise, they will make the claim that the supplement works, never mind that it was a small amount of weight that was most likely shed through the changes in exercise and diet. Similarly, politicians and pundits love to take correlations and claim that they prove a "common sense" causation, such as a couple of hard winters or previous changes in the climate as arguments against global warming.

SOURCES:

http://en.wikipedia.org/wiki/Correlation_does_not_imply_causation