So what do we know now? In the last couple of weeks, we’ve gotten a new set of numbers on school dropouts, an analysis of the effects of class-size reduction on student achievement and, most recently, this year’s statewide results on the SAT9, the Stanford 9 test in reading, math and other subjects, that’s given to all California public students in grades 2 to 11.

But the bottom line, to use one of state schools Superintendent Delaine Eastin’s favorite phrases, is mostly statistical noise. Eastin called the SAT9 results “good news,” — and indeed, the scores crept up in most subjects and most grades, more in math, a little more in the lower grades.

But this is the second time that the SAT9 has been given. And once you take away the effects of intensive test preparation and increased student familiarity with test forms, most of the real statewide gains are probably negligible.

But then, given the fact that California kids, one-fourth of whom are not native English speakers, were given a test that was standardized on a population of whom only 2 percent are not native English speakers, and that California has been shortchanging its schools for a generation, the results aren’t so awful either. Depending on grade, between 40 percent and 45 percent of our elementary-school kids score above average in reading; between 45 percent and 50 percent of our kids are above average in math.

The same uncertainty afflicts class-size reduction, which, at a price of $ 1.4 billion a year, is surely one of the costliest social experiments ever tried in any American state.

What can most safely be concluded from the recent analysis of the Class Size Reduction Research Consortium is that it’s been a wash. The small gains that it appears to have achieved for the majority of students were offset by overcrowding in other classes and by the shift of qualified teachers out of inner cities and their replacement by undertrained teachers in the schools serving the poor kids who can least afford them.

From there, things get even murkier. Hanging over from last year, and still unexplained, is the large drop in reading scores between the eighth and the ninth grades, a phenomenon that seems to have appeared in other states on the SAT9 as well.

It can’t be hormones, as some have suggested, since hormones seem to have had no effect on the math scores. That raises questions about the test itself. And since Harcourt Educational Measurement, which administers the test, managed to screw up the average scores for the state’s Limited English Proficient (LEP) students by mis-coding the tests of some 250,000 English-proficient students, there is reason for concern.

Harcourt promises corrected scores for the LEP kids by July 15, but after a year, there’s still no word on the ninth-grade dip — which, by the way, seems to persist at least through grade 11, which is the last time students take the SAT9.

Maybe the best way to see these numbers is as a sort of ideological Rorschach test, into which the various faiths in our pedagogical disputes read their own convictions. Ever since the Oceanside school district in suburban San Diego reported large gains for its LEP students, gains that were exaggerated by Harcourt’s coding goof, Ron Unz, author of Proposition 227, which seeks to end bilingual education, has been blitzing journalists with e-mails celebrating the now-proven blessings of his measure.

Meanwhile, surprisingly large leaps in SAT9 scores in Sacramento city schools have been attributed to an intensive set of reforms, among them a basic restructuring of the district and a major emphasis on phonics-based reading instruction. But surely if more intensive reading instruction made a difference in Sacramento, it could have made a difference in Oceanside as well. Success has many fathers.

Both theories may be right. But in the current frenzy of school reform, a sober dose of skepticism is in order. It’s a rare school reform that has any quick, dramatic — and lasting — effect. We are all for higher standards and more accountability, and California may be achieving them, but we should be aware that in our ongoing cycles of school reform, victory has often been proclaimed, only to be followed by yet another reform aimed at undoing the glorious victories of the past.

There’s no sign that the pattern is changing. In late June, a few days before the SAT9 scores were released, the Senate Education Committee approved a bill, already passed by the Assembly, that would exempt all LEP kids who’ve been in U.S. schools for less than two years from taking the SAT9.

The bill, AB 144 by Assemblywoman Carole Migden, is founded on the belief that students who have been in U.S. schools less than two or even three years can’t be expected to know enough English to be tested. This belief, too, is a tenet of one of our many churches of school reform. But if the state is trying to force the schools to teach kids in English, as 227 requires, it surely ought to be testing them in English as well.

The other day, Eastin stoutly insisted that despite the problems with the test, California should stay the course. Maybe, she said, the SAT9 is a Model T, but you can’t get to your modern sports utility vehicle without starting somewhere. It’s a funny illustration of progress. What made it particularly odd is that Eastin supports the Migden bill. This, it appears, is a part of the course she doesn’t want to stay either.

She’s right about staying the course — really staying it — because no set of numbers makes much sense without reliable bases of comparison with other places and with our own past. Keep changing the rules and you’ll never be sure of anything. And you’ll make the schools even more nuts. But do we really want to use just this test for the high-stakes consequences — judging kids and teachers and principals — that we’re about to attach to it?

Comments are closed.