can correlation

ever equal causation or can it only imply it?

What?

First of all, correlation does NOT imply causation, that's the whole point of the saying.

Second, while we can measure correlation with a statistic, I don't think we can measure causation. So how would they be equal?

I thought they were inversely proportional ! ?

Quote: dwheatley

What?

First of all, correlation does NOT imply causation, that's the whole point of the saying.

Second, while we can measure correlation with a statistic, I don't think we can measure causation. So how would they be equal?

I'm not talking about the "saying"....the question is, can the correlation of 2 variables = causation?

as to your statement of "correlation does NOT imply causation" I think a better statement would be correlation does NOT AUTOMATICALLY imply causation

Yes, correlation can mean causation, but the fact that two variables are collarated does not necessarily mean that one causes the other.
In certain cases the causation is obvious and there is consensus about that.
In other cases, causation is often the subject of debate especially in areas like econimics and social sciences.
But even in other areas like medicine and health, correlation does not necesarily mean causation.
I think many scientists make the mistake of attributing causation when they find correlation.
This is especially wrong when they take existing data and they look into them to find correlations between many different variables and when they find correlation they try to attribute causation according to what makes sense to them. It is inevatable that in any set of data of a population there will be some correlation in some of the variables, some times this will be just pure chance.
It is a completely different scenario when a scientist has a hypothesis based on a theory and then he does the test and checks for correlation on the variables based on his theory, to confirm it or not.

Correlation is a number. Causation is a concept. No, one cannot be equal to each other, that just does not make any sense.
Also, correlation does not imply causation period. Automatically or otherwise. The implication goes the opposite way.

It would be impossible for causation to exist without 2 variables being correlated, i.e. correlation is necessary for causation. So... of course the correlation of 2 variables CAN '=' causation.

In this context, the word imply means 'to involve by logical necessity'. The necessity is obviously not there, so saying 'does not automatically imply' is painfully redundant.

To get back to the original question, "yes" (and how could it be anything but yes?)

You are certainly entitled to your opinion. But emperical evidence can be sited to prove you are wrong.

the point of this thread was just to bring up a debate that I was having with someone.....where they made a statement of correlation=/=causation is always true....

Not arguing the validity of that statement.

Correlation between X and Y does not mean X causes Y. However, X and Y can be correlated and X can cause Y.

Smoking and Lung Cancer are correlated, and Smoking causes Lung cancer. Lung Cancer does not cause Smoking, but there is a correlation between lung cancer and smoking.

A correlation is a reason to investigate an underlying mechanic or phenomena, and is a start, not an end.

Quote: vert1276

ever equal causation or can it only imply it?

The usual example given is the number of storks observed in an Austrian village's bell tower and the number of babies born in the village: no matter how many hundreds of years there has been a direct 1:1 correlation this is not to be considered proof of causation!