On Statistical Assumptions

I work for a market research firm, so I deal with statistics on a daily basis. Listening to radio programs where statistics are thrown left and right to support absolutely opposing views, I recognize that anyone can pull a study that will show that their viewpoint is upheld by some report or another. People have this bizzare faith in statistical analysis that is only matched by their assertion that statistical analysis can be used to say whatever you want it to say. It reminds me, in fact, of the way people have historically used religious texts to validate their politics. When data is left open to interpretation, whether it's a biblical parable or a statistical report, people need to take whatever rationalization that comes out if it with a grain of salt (or in some cases, a block).

The reason I bring this up is that I was listening to a radio program that had a guest on, talking about a research study that supposedly shows that calorie intake, as well as other dietary factors, correlates with having one gender or another. This person was talking about how calorie intake makes evolutionary sense, since when there is a plenty of food to be distributed, the human race can afford more males (the study showed that the more calories, the more likely to have a boy). So, female babys can survive with less, while it takes more, statistically, to have a boy. They finally asserted, as a result of these findings, that a woman can influence her likelihood to have a boy or girl based on her diet.

The major, though not the only, problem with this is that it overlooks the obvious fact that while boys are genetically more dependent on calorie intake in the womb, the child's gender is determined relatively from conception. It's only a survivability factor that this study shows to exist, not a gender-influence factor. A woman who has conceived what will be a boy cannot decide to eat less and eventually have a girl. It will only increase the chance that her baby will die from a lack of caloric intake. Obviously, these conclusions are in totally different ballparks. It surprised me that the host did not make this glaring contradiction to the guest's assertion more apparent.