A Rant

CORRELATION DOES NOT EQUAL CAUSATION. 

Please memorize that phrase, and understand it as well. Just because A and B often happen together doesn't automatically mean A causes B, or for that matter, B causes A.

Mind you, it could mean either or it might mean some third thing C causes both. Or it could conceivably be a coincidence. Correlation suggests a possible relationship between two things which tend to occur together, nothing more.

Now, why the rant from your normally calm and serene blogger? Yet another example of reporters writing stories which demonstrate they do not understand what correlation does and does not mean.

The stinker is this article in the Atlanta Journal-Constitution. Hat tip to Lucianne.com for the link. For starters its title is misleading.

Where you live in Atlanta affects life expectancy.

This is a classic example of thing C causing both A and B. Where one lives in Atlanta is a function of race and income, probably mostly income as I'm certain there are good black neighborhoods in Atlanta as there are in D.C.

Have a decent income, you are likely to have good health insurance and healthier habits, less likely to use street drugs as recreation. A good income helps you live in a nice neighborhood where you are much less likely to be shot on your front porch. More likely to be busy at work 40-50 hours a week where the police are unlikely to hassle you.

The bottom line: it's true people in certain Atlanta neighborhoods don't live as long, but living there is a result of who they are and, more importantly, who they are not. Where one lives is an outcome, not a cause. The AJC headline should read:

Same factors influence both where you live in Atlanta and your life expectancy.