how to improve writing in the schools...

Like everything else: do it a lot and get good feedback from those with more experience and skill!

From Michael Prince in the Chronicle Review (hat tip: Jay White)...

Even the manufacturer of the SAT admits that the new test, which includes writing, is no better than the old test, which didn't...

Prince then cites a Boston Globe story which reported that the new SAT "rated 0.53 on a measure of predictive ability" of "first-year college grades". (Was this adjusted R-squared??) This compared with 0.52 for the old SAT. Whether statistically significant or not (there's no mention of an F-test or "statistical significance"), the additional impact is certainly not substantial.

A few thoughts:

-This is not to say that the SAT has no predictive power-- but that the new SAT adds little predictive power to what the old SAT provided. So, this should not be seen as a call to eliminate the SAT-- as Prince seems to claim later in the article. (Ironically, his writing is not clear on whether he thinks the SAT or only its writing portion should be sacked.)

-Does this result hold for all schools-- or does the new SAT provide more info at certain types of schools (Ivy League vs. large state vs. regional university vs. community college)?

-The study implicitly assumes that schools are using the SAT properly as a predictor of student performance and adjusting accordingly (turning aside unqualified students, categorizing students as probationary, etc.). If the SAT is ignored-- for example, because of poor statistical modeling by the university, an open-door desire for tuition dollars, or affirmative action policies which trump the data-- then one cannot say that the SAT-writing is "not predictive".

How could this happen? College professors frequently ask their students to write. Shouldn't a test that includes actual writing tell us more about scholastic aptitude than a test that doesn't?

Well, it's easy to imagine that writing is important and predictive in theory-- but that it may be difficult to test in practice, for a number of reasons.

Here's another (valid) angle explored by Prince:

...[perhaps] the test asks students to do something categorically different from what college professors generally ask their students to do. Is that the problem with the SAT? You be the judge.

Prince provides an obtuse sample question used to prompt writing on the test and then analyzes it:

Most college professors—especially those outside the humanities—would view the SAT essay prompt as significantly unlike their own writing assignments. First and foremost, we ask students to read....The SAT writing test fails for the simple reason that it ignores reading comprehension, overrates argument, and plays down grammar and prose mechanics. My advice: Toss the test; upgrade the skills it neglects...

His second argument is a long discussion of what is meant by (and taught with respect to) critical thinking.

That brings me to the third inexpensive change that faculty and administrators can make to foster the success of their high-school English and college composition programs...imitation-based pedagogies that view students less as budding cultural critics and more as apprentices to a craft.

This is an important point. If writing is a craft, then budding young writers need to be crafted-- through their own efforts and the wise counsel of others. In a word, students should write early and often-- and their teachers should provide candid and competent feedback to their apprentices.

Prince concludes with a debate within the university on how to improve writing: should it be done in writing courses or throughout the curriculum?

What, then, should writing courses be about? Enlightened instructors and administrators will respond that they should be about what all other college courses are about—not writing itself, but a learnable body of information: literature, art history, biology, political science, or any other substantial topic that furthers a students' real education. Yes, there are rhetorical strategies that good writers know and weak writers lack, but those are best taught in every class, by faculty members who themselves have mastered not only a body of knowledge but also the skills for writing publishable work and sharing those skills with apprentices to their craft.

His inference should not be surprising-- and is, I think, quite correct. Writing should be suffused throughout the curriculum-- in order to be practiced often and to be shaped/informed by the perspectives of a variety of experienced writers.