Thursday, February 02, 2006

Coffee Causes False Cause; Correlation Error To Be Acceptable

I teach a course on alcohol and drugs at Walladumpdung University (I use the fake name Walladumpdung so that no-one knows that I actually teach at Griffith University). I don't drink alcohol, I don't smoke, and I don't use illegal drugs or abuse pharmaceuticals; so I don't experience a crisis of conscience when I expound on drug and alcohol issues.

However I do drink coffee. I LOVE coffee. So naturally when I look at drug and alcohol research on coffee I make sure I only pay attention to literature which shows that coffee is either harmless or good for you. If for example, a study finds that people who drink lots of coffee are less prone to dementia, I will assume, as do the researchers, that coffee consumption is protective against the onset of dementia. Of course, both I and the coffee-loving researchers know that the causal relationship could be the other way around - that is, people who are not naturally prone to dementia are more likely to like coffee.

But hey, if you're looking for an excuse to drink coffee, the occasional False Cause; Correlation Error can be safely ignored.

Um... on a personal note... if anyone knows of a study which shows that weight gain causes excessive food intake I would be interested in following it up.