I have two contrary definitions of for the direct correlation between two variables $X$ and $Y$
- Their correlation coefficient is close to $1$.
- There is a direct causal relationship between the variables (ie. there is no $Z$ such that $Z$ causes $X$ and $Z$ causes $Y$)
Which one is correct?
I'm not a statistician (though I taught statistics about 35 years ago). But I'd say that the first definition is correct. One of the important points about correlation is that it should not be confused with causality. Remember the old joke: Smoking and cancer are correlated; does this mean that smoking causes cancer, or that cancer causes smoking. Neither one -- correlation by itself does not imply causality in either direction.