ABSTRACT
Even in the first course in statistics, the slogan "Correlation is no proof of causation!" is imprinted firmly in the mind of the aspiring statistician or social scientist. It is possible that he leaves the course with no very clear ideas as to what is proved by correlation, but he never ceases to be on guard against "spurious" correlation, that master of imposture who is always representing himself as "true" correlation. The bridge from the identification problem to the problem of spurious correlation is built by constructing a precise and operationally meaningful definition of causality—or, more specifically, of causal ordering among variables in a model. In investigating spurious correlation, the chapter considers learning whether the relation between two variables persists or disappears when it introduce a third variable. The chapter suggests that the relations in question are linear and without loss of generality, that the variables are measured from their respective means.
