I can't help noticing that the USA Today article doesn't mention how strong the correlation is between the SAT writing score and the GPAs, they just say that it's a "better" predictor. But the SAT is know to be a lousy predictor of college GPA--is the writing section merely less lousy? As I recall, SAT scores correlate most strongly with family wealth. I'm not going to interpret this study, done by the people with a deep conflict of interest, as meaning anything until I know what kind of correlation we're talking about, but I bet most news organizations won't be so careful. It makes me wonder if they're releasing this information now, well in advance of publication of the finalized study, precisely because the actual correlation is weak but they know the media will run with reports of "better prediction" in absense of the numbers.

The SAT does not measure one's ability to write well and produce good prose. It simply measures the quality of one's first draft.

Students _always_ write better when they write about themselves.

Actually, being successful on the SAT writing exam probably has a poor correlation with one thing: grades in their freshmen English courses, where us teachers need to deprogram them from the formalist crap they learn from taking standardized tests. Instead of having students think about the rhetorical issues that face anyone when engaging in a writing task (such as, for example, thinking about audience or genre), these tests encourage them to think in a rigid way about writing--they come to believe, for example, that all essays have an introduction that regurgitates the essay's main points. Then their college teachers (like myself) need to explain how introductions can be written in several different ways, depending on the purpose of the essay. And, just so you know, plenty of evidence exists that grammar instruction actually is completely ineffective tool for writing instruction (and that teaching students argument and dealing with competing evidence is actually the majority of the teacher's task, regardless of the level of the student).

Frankly, I think Robert Farley overestimates the value of many college classes. Students don't write well because at both the high-school and college levels, the teachers are too willing to overlook poor writing. If you don't demand greatness, you don't get it. If a student can't put together are good argument in a reasonably structured way, then that essay deserves no better than a D. The fact is, though, that those essays consistently earn B's or at worst C's. That's just not enough of a punishment to encourage the students to perform any better.

Now, if you happen to be a good writer already, this dynamic puts you at a significant advantage, but at the same time, there's nothing worse than a professor giving you an "A" on a paper that you would personally rate a "C." But in my experience, that is what college is: a never ending process to encourage students to under-perform because professors are too lazy to challenge them. Whether this is from pressure to keep the GPA's artificially high in order to improve a college's ranking, I don't know.

What I do know is that every student would benefit far more by reading than writing. Ultimately, reading is the best way to learn how to write. You can be reasonably sure that if a text is important enough for you to read, then the writer is excellent at structuring an essay. The more you read, the better you become at writing.

If colleges pointed their students in directions that encouraged them to read books that were interesting to them (but within the bounds of course objectives) rather than texts assigned with no consideration to the fact that students don't read books that bore them, then we'd probably see a lot more students actually reading the texts instead of glossing them.

I might argue that "glossing" over texts is the single biggest factor in contributing to poor writing among students. They don't write well, because they don't read well. They scan books, picking up the major points, but not the actual structure of the argument, then when they write their own essays, they aren't aware of how to adequately structure an argument. They work in broad strokes with little narrative flow and a nearly complete lack of nuance or genuine evidence.

One final point: topic sentences are for losers. They are a lazy writer's way of winning points. If a reader can't understand the point of a paragraph after reading it, then the author is a terrible writer and so-called "topic sentences" would merely serve to obscure that fact.

Not to mention those sorts of essays tend to be the driest, more boring essays imaginable.

The problem isn't with topic sentences, it's with papers lacking a thesis to begin with and then lacking the ability to structure whatever argument they're making and failing to back up arguments with sufficient or relevant evidence.

Or, to put it another way, I agree with much of what KR posted.

Another way of putting it is that the writing scores are more predictive because part of the test is "soft" grading, which is exactly how teachers in both high school and college grade.

However, colleges clearly consider the SAT to be predictive of actual ability (as opposed to the utterly useless GPA metric). That's why almost all of them allow students to skip out of courses or get out of remediation by achieving a certain test score.

"One final point: topic sentences are for losers."

Catchy topic sentence.

"If a reader can't understand the point of a paragraph after reading it, then the author is a terrible writer and so-called 'topic sentences' would merely serve to obscure that fact."

Wow, that's crazy. If topic sentences obscure the fact that a writer is terrible, isn't that an argument for using them?

Your whole paragraph is self-refuting. Your topic sentence attacks topic sentences. And your argument against topic sentences is a good reason to use them. Who said Jacques Derrida is dead?