Cooking the Science Books

Report Finds Accounting Practices That Start at the Bottom Line.

The first thought that struck me when I read this headline was "That's how the Christianists approach science."

U.S. financial markets are supposed to be the most transparent and heavily regulated, with numbers you can trust. But as a 2,652-page report released yesterday shows, government-chartered Fannie Mae, one of the nation's largest and most scrutinized financial companies, was able to create the numbers it wanted.

The article is about Fannie Mae, and that's not what this post is about. It just draws a remarkable parallel with the way that "research" has been carried out by advocates for social repression -- the anti-evolutionists, the anti-gay rights activists, the anti-feminists, and the like.

It starts with cherrypicking data.

A great example of this is Paul Cameron, head of the Family Research Institute (which has been listed by the Southern Poverty Law Center as a "hate group," by the way). Cameron periodically comes out with "studies" that show gay men with a higher incidence of disease, shorter life spans, and a host of other presumably undesirable traits. Unfortunately, his results are all bogus because he picks and chooses the data he wants to use, which is not how science works. (Conspiracy theories aside, that is the reason he was censured and tossed out of the American Psychologicl Association -- bogus research, on a regular and consistent basis. It was not a conspiracy by gay activists, no matter what Donald Wildmon tells you.)

Study any creationist tome, or any tract on intelligent design, and you will find the same phenomenon. They start with the idea of creation then work back from there. (Check out Talk Origins to get a number of good examples of this approach at work.) That's if they can come up with anything original to begin with. Somehow, "God did it" doesn't really meet scientific standards of methodological rigor.

The "studies" so often cited by opponents of gay marriage and gay-parent adoption also play this game. The anti-gay apologists repeatedly start off with "Scientific studies show that children do better in homes with a mother and father." Nope. In most cases, they've taken studies that compare single-parent households with heterosexual, dual-parent households and present that as "evidence" against gay parenting. That's the only evidence they have, the stuff they make up. What studies actually show is that children raised by same-sex parents are exactly comparable to children raised by opposite-sex parents, except that they may actually be better at establishing and maintaining interpersonal relationships. I have lists of studies in this area, and I haven't found one that actually tests that question that supports the anti-gay arguments. Not one.

I remember a citation of a "study" conducted by a right-wing, anti-gay Christian group a number of years ago that "proved" that only 1% of the population was gay. (This was in the days when they were still trying to make the numbers argument: there are so few of them, the rest of us don't need to give them any rights.) It turns out the "study" was an exit poll at Baptist churches in the south after Sunday services. What's striking is that they still got 1% of respondents who would, under those circumstances, confess to being gay.

I've also seen the references to studies that abortion causes cancer. I haven't really seen anything definitive in that area, but just on first reading, I'm dubious. It's very easy to publish an impressive bibliography, and I suspect that the authors are relying on the fact that very few people will go back to the primary sources. And, as we've learned in the recent Korean stem cell research fiasco, if data is falsified, it's very hard to spot.

And do we really need to go into the Kansas school board's rewriting of science education standards that included a new definition of science as simply "a means of studying the universe." Astrology 101 is being offered in the fall. It's of a piece with the new "critical analysis" approach that is ID's new false moustache. I guess no one told them that's the way science operates anyway.

Here's how you prove a hypothesis if you're a Christianist: you start with the idea, for example, that God caused the earth to be covered in a flood. Then you go back and you find evidence that supports your conclusion, such as fossilized trees that extend through more than one geological layer. You ignore the other evidence, such as indications of rapid deposition of sediments, especially if the sediments were wind borne or riverine (oceanic sediments are OK, because they fit the conclusion of a world-wide flood). Voila! You have a "scientific study."

Cooking the books: it's not just for accountants any more.