It’s SAT season. For the last week or so, all across this mighty land, millions of high-school kids who took the SATs in May have been getting their scores online. A young friend just e-mailed to tell me that she got 760 in critical reading, 640 in math, and 680 in writing.
What do these numbers mean? Well, arguably, a lot. Most college admissions offices still look at them and get very yippy and excited when they see a combined score in the top percentile or two. On the other hand, arguably, not much.
First, let’s be clear that the private nonprofits who administer the SATs, PSATs, and GREs, namely the College Board and its drooling hunchbacked assistant, the Educational Testing Service, were not appointed by God. They’re just people in offices who make this stuff up. Nor should we be fooled by the word “nonprofit” into thinking that they’re doing it for our own good: “nonprofit” is not a synonym for “volunteer.” Careers, paychecks, publications, and reputations are all on the line. Testing, even nonprofit testing, is a business, not a charity.
Second—let the joyous news be spread—it’s not a science, either. SATs don’t measure anything, any thing. Your “scholastic aptitude”—the original words behind the “SA” in SAT, which were officially dropped in 1994, leaving the acronym to float free—is not a fixed thing in your head, like an electrical charge or a bullet from an old war wound. It’s a construct, a fiction. An SAT score’s value, like that of a $20 bill, exists only because people have agreed to agree that it exists. And how are they persuaded to do that? Well, one little trick that is played to make the SATs look more objective, scientific, and reality-based is hidden in plain slight. All the scores end in zero.
Ever notice? You can’t get 761 or 653, only 760 or 650 or some other multiple of 10. That zero could be deleted right across the board and all SAT scores would still convey just as much information, or as little. Therefore, all SAT scores are inflated by a factor of at least 10. It’s worse, actually, because the section scale doesn’t go from 0 to 800 but from 200 to 800. That means there are really only 61 possible scores for each separate section of the SAT: i.e., 200, 210, 220, and so on, on up to 800. Dividing 800 (apparent resolution) by 61 (actual resolution), we find that all SAT scores are actually inflated by a factor of about 13.
In hard-science land, they would say that the SAT score displays spurious resolution. (Trust me, I’m an engineer. Heh heh.) If a hard-science person announced that they had measured some quantity on a scale of 800, they would be implying in no uncertain terms that they could measure differences or changes of one part in 800. If they could only measure differences or changes to one part in 61, they would use a coarser scale—one that doesn’t pretend to record many more than 61 distinctions. Tacking zeros onto one’s numbers to make them sound bigger, more important, and higher-resolution is pseudoscience. It’s bad.
But by all means, take the SATs or GREs or whatnot if you must. I did. We all have to jump through some hoops to earn our sardines from the trainers. Just remember that you are not your score on the ACT, SAT, GRE, IQ, LSAT, or any other obnoxious bite of alphabet soup. Nobody is.
I’ve always felt that a bad attitude—on test day, not during the prep phase—is a good ally. Try thinking rebellious thoughts, feeling terroristic feelings, as you pencil in the little ovals or click on the screen choices. When I sat down in the early 1990s to face the computerized GREs (Graduate Record Exams, like the SATs only meaner), I was positively pressurized with hatred and loathing. The verbal portion threw me question after question, even multiple-choice vocabulary questions for Chrissake (what do they prove?), getting steadily harder and harder, definitions and analogies and antonyms. Some little block of code was looking for my limit, trying to “adapt” (as the GRE brochure put it) to my supposed level of ability. Finally the test came up with the word scission. I’ve never seen scission in print before or since, but, having used scissors since kindergarten, I was able to guess its meaning correctly (“a cutting, dividing, or splitting; separation”). Choice (c). Take that, you stinking machine!
And then something, apparently, snapped. Not in me—in the stinking machine. The remaining questions were all absurdly easy, and I cruised through them bathed in a sense of cold, bitter triumph. John Henry beats steam hammer, and I didn’t have to die for it.
Take-home message: Tests are designed to classify and control us, not benefit us. And no matter how many numbers they spew, they are not science.