When two variables are correlated variation in one?

Can be cause of variation in the other only if the correction is positiv

may possibly cause variation in the other

Always suggests a causal relationship with the other

Can only be inferred with social validation