Talk:Correlation

Misunderstanding[change source]

The following examples show what correlations do not mean:

Positive correlation: The more the cows get fed, the fatter they will be. (The correlation is only: Many cows are fat, many cows are fed well. That the second is the cause for the first in not content of the correlation. But it is the meaning of the sentence "The more the cows get fed, the fatter they will be.")

Negative correlation: If more umbrellas are sold, less sunscreen will be. (dito)

Not correlated: It might well be that in a year where there are fewer storks are in a certain region, fewer babies are born there. Storks usually have nothing to do wirh childbirth. The two variables might however be linked by some other variable. There may be fewer storks because breeding spaces are harder to find (because of urbanization)

The last one is a Positive correlation and the only good exampel for correlation, because is does not say anything about the reason why the correlation is there. --Cethegus 15:55, 18 July 2007 (UTC)[reply]

Nice save! You're absolutely correct. We'll have to replace them with examples that actually make sense. · Tygartl1^·talk· 16:14, 18 July 2007 (UTC)[reply]

Oops, I didn't realize this was a new article. I figured it was an older article with that information added by an IP a while ago. Looking into the situation a little further, I realize my response may have been a little inappropriate. At any rate, I think examples are an excellent idea. Correlation can be very confusing to many people so I think examples will be helpful to understanding this difficult concept. We just have to make sure they are true positive and negative correlations. (I may have to brush up on my college statistics!) :-) · Tygartl1^·talk· 16:36, 18 July 2007 (UTC)[reply]