Over at The Weekly Standard’s blog, Stuart Koehl disputes my point that the SAT Writing does a better job than the other two SAT tests at predicting college-freshman GPAs. To his credit, the post is a bit more reasonable than a similar letter he submitted to PBC, which featured a postscript beginning with “You had better print this.” (Which, human nature being what it is, led directly to me not printing it.)

Koehl is empirically wrong; the question of which test best predicts grades is a simple statistical one. As per the USA Today piece I linked to, Writing came out on top in two different studies. (One was by the College Board, which makes the SAT; the other was by the University of California.) This surprises everyone — as Koehl says, the writing test is the most subjective — but numbers are numbers. Putting “(!)” after the terms “USA Today” and “College Board” doesn’t change this.

Koehl also references a ridiculous Princeton Review exercise that fed various famous writers’ prose through the SAT Writing’s rubric — which is designed to test one’s expository-writing ability and strict adherence to modern grammar. Shakespeare, who wrote fictional plays in non-current English, oddly enough does not do well. The Unabomber, on the other hand, scores highly.

Koehl writes:

In other words, because of its mechanistic approach to writing, the writing test rewards people who follow arbitrary and generally false “rules,” such as short sentence structure, avoidance of the passive voice, and not beginning sentences with “and” or “but.”

I agree on “and” and “but,” but the other two rules dramatically improve readability. That’s something we should all strive for and tests should measure. (Or: But while I can possibly agree with Mr. Koehl that “and” and “but” might sometimes be considered appropriate starting words for sentences, he would be wise to recognize that passive voice should be avoided, and that word economy should be practiced, as these techniques make life easier for readers of one’s work and provide for clarity of thought; it should be deemed crucial that a measurement tool such as the SAT Writing sees to it that such practices are rewarded.)

He continues:

According to [an e-mail of mine], “Schools should not train students to turn in papers that sound like Shakespeare anyway,” which is debatable. One ought to ask, though, whether it should be encouraging people who write like the Unabomber.

Actually, yes, they should; he was nuts, but a good writer. You can see this by taking the “Gore vs. the Unabomber” test — oftentimes, the paragraphs that make sense are the Unabomber’s, not the former presidential candidate’s.

A slightly more convincing point he made in our e-mail exchange was that the SAT Writing is a young test. Future essay-question-writers and scorers might not do as good a job as the early ones clearly have. But this is a Food and Drug Administration kind of logic — sure, there’s solid evidence that this new medicine is better than what we have now, but let’s keep using the old, probably inferior method for another decade until we’re absolutely positive.

Also, given the degree to which the College Board is invested in the SAT’s strict standardization, I doubt this will be an issue. Two scorers independently grade each essay.

I’ll be keeping my eyes peeled for new studies, but two of them have come to the same conclusion — the SAT Writing is a valuable tool for predicting whether someone will do well in college — and we can only act on the information we have.

P.S. The “B” in my last name is capitalized.