Wednesday, April 14, 2004

Virtually Meaningless Statistics

Here's an article that is full of statistics on heights of populations that are about as meaningful as the old joke that the average adult has one breast and one testicle.

While the historical height averages are interesting, the modern ones (and especially the comparison of historical to modern) are virtually meaningless. For example, the average height of Americans is compared with the average height of the Dutch, and the results are supposed to be caused by prenatal clinics. What a crock! There are so many confounding factors that it would drive you nuts just thinking of them all. One of them is the fact that America, unlike the Netherlands, is a country that welcomes the "huddled masses yearning to breathe free", and to which many previously malnourished people like my mother- and father-in-law emigrate. (My husband's family alone will drive the average height down at least two inches.) The American population is also a hotbed of interracial childbearing, which tends to dilute the effects of any "tall genes" that might be floating around out there.

Because these mixing conditions did not hold historically for many populations (e.g. the 18th century Londoners), the historical height averages are meaningful. But the modern data is simply not comparable to the historical data.

Look, everybody knows that malnutrition and lack of prenatal care (on your mother's part) can make you short. But it's just not true that shortness is always due to malnutrition or lack of prenatal care-- just ask Tiny Princess, who is fed three squares a day and received the world's finest prenatal care, and is still a full foot shorter than her classmates. And trying to pin shortness on unemployment is just ridiculous. This guy is further than out on a limb with his conclusions-- he's out past the last bud on a twig.