An article in the Smoky Mountain News comparing the health of North Carolina counties may serve well for a case study in statistics. First of all, a newspaper-length article cannot do justice to the number of variables that must necessarily be in a study billing itself as indicator of “health.” The reporter did a good job of trying, however.

Typically, the scientific method only tracks two variables at a time. That usually works for hard sciences, but can only make broad-brush conclusions for the chaos-based equations governing living organisms combined with their free agency. The study ranked each county’s health on the basis of “health behaviors, clinical care, social and economic factors, physical environment, and percentage of smokers.” To make matters worse, an expert quoted in the article says Americans just look less healthy than people in Thailand. Another spokesperson noted variables were selectively applied to whatever algorithms were used in computing health; and some were double-counted.

The study integrated obesity into its calculations. Hauling just a few extra pounds of flab is considered obese by today’s standards, but was healthy in Marilyn Monroe’s heyday. Some people can get obese without contracting heart problems and type 2 diabetes. In and of itself, obesity is not a problem. Education is another variable considered. The rationale being that women who can read tend to get their children vaccinated more than those who don’t.

Smoking is another, and we all know about George Burns. Swain County fared miserably in the study. One reason was its high teen pregnancy rate. Years ago, the UN considered the availability of birth control an indication of the degree to which a country’s healthcare system had advanced. The Wisconsin study used “the number of diabetic Medicare patients getting annual blood sugar control tests” as a standard. Some people maintain health by avoiding doctors unless something is broken.

And, of course, studies like this are needed for planning.