Bad Science

Ok, stay with me here. The following is the beginning of an article that was linked from the Yahoo! main page:

"Do Diet Sodas Make You Fat?

The short answer(s) to this question is no and, maybe, yes. One recent study has shown that people who drink diet soda still have a 41 percent chance of being overweight."


There are more problems with this than I could point out. But they range from scientific mispractice, to the state of "press release" journalism, to unfortunate cultural underpinnings.

Suffice it to say, the first day of any science class one hears a famous mantra that should be in every functioning human being's lexicon: CORRELATION IS NOT CAUSATION.

Example. Students who take Latin do better on the SAT. This is actually true and lead to a spat of "have your kids take Latin!" newspaper articles. The problem is that with correlation A can cause B, B can cause A, or an unknown factor C can be the cause of both A and B. That's the case here. Schools that can afford Latin classes are not inner city ghettos, and they provide a better educational experience overall, not simply in the Latin class. The same could likely be said for kids who play Lacrosse.