Concrete Academic

Numbers don’t lie, we’re told, but they lie all the time. Ok, in and of themselves they may be innocent, but numbers are open to such variegated interpretation that they shouldn’t be completely trusted. I’m somewhat suspicious of the widely reported article in Reproductive Health, a peer-reviewed medical journal, last month. What the press reported, with utmost fidelity, is that the more religious the state the higher the rate of birth. But whether or not it is spelled out, this implies quite a lot more.

At stake are questions of religious hypocrisy, religio-economic and -educational stereotypes, and even the very morality of religion, Christianity in particular. Science does a far better job at insulating itself from politics than religion does. Religion wallows in it. Nevertheless, even when someone is speaking from the hallowed halls of science, we still have to think carefully about what we read and watch for agendas just as closely. The “democratization” of news in the age of the Internet, and more recently “Web 2.0,” has increased our awareness of effect of mediation: the news is most certainly filtered, it’s only the degree that’s in question. With the plurality of sources available we can even filter it ourselves, pumping only the sources that agree with our point of view.

But in this case there is further complicating factor beyond the media. Those involved in higher levels of study in the sciences or the humanities know that there generally is a significant disparity between the state-of-the-art in the academy and the state of public understanding. This is a fitting subject for Concrete Academic, I should think. Scientific findings have to be interpreted, and when they are not contextualized responsibly—forgive me this cliché—a little knowledge can become a dangerous thing.

So much for the preface. The study, “Religiosity and teen birth rate in the United States,” is only fourteen pages long, so it is worth reading. The authors begin by noting the influence of religiosity in the debate over sex education in school, and the importance of such education along with contraception, as opposed to teaching “abstinence only,” for lowering the birth rate. Given this, the relationship between religiosity and birth rate should be examined. After reviewing a few previous studies, which did not speak univocally, they explained their own research.

So far so good, but now a couple of things give me pause. First, they did not do their own field work but rather mined existing data. Sometimes this is necessary, and it’s certainly easier logistically, but it’s harder to get fix on what you’re looking for. The data they worked with came from two separate banks: birth rates from the Center for Disease Control, and a poll on religious feelings by the Pew Forum on Religion and Public Life (plus another, negligible, to control for income). Both were done at the state level. Using one study they compiled a ‘religiosity score’ for each state, and then using the other they looked at birth rates for teens. The hard statistical work came in correlating the numbers. They no doubt did good work with what they had. But this is not a very close look at the issue. Certainly it is not close enough to account for a plurality of sociological factors, such as the propensity for the religious to marry younger than the irreligious.

The authors admirably warn against over-interpretation of their statistics at the end of the paper, so much so that one wonders about the usefulness of the work. I think it is suggestive, but far from impressive. The media’s coverage of it wasn’t either. It’s hard to fault the AP and others for reporting the ‘facts,’ which they did, but they lacked sufficiant context. The professionals reading the journal know how to weigh the article, and they speak the language well enough to understand what the authors can and can’t do with it; the general public does not. This requires us as readers to be aware of our own ignorance and to be modest in drawing conclusions.