enter the fray: our reader discussion forum
Re: Reply to: Why are Jamaicans so good at sprinting?
by Vashti

"Correlation DOES say something about causation. Correlation is a necessary condition for causation, although not a sufficient one. Thus establishing a strong correlation is suggestive of causation, but does not actually prove causation"

It depends. Correlation doesn't ALWAYS say something about causation. Because two variables can be correlated just by chance alone, and in fact be completely unrelated.

Correlation speaks more to association rather than causation, and even that association can be completely spurious or driven by something else. Also saying that strong correlation is suggestive of causation is extremely sloppy science.

I do agree that correlation is necessary but not sufficient criterion. Correlation is a very small slice of the pie and we should not let it carry the burden of proof.

View complete thread