Peter Schrag: California schools' English language learner limbo

It's easy to see how CELDT, the California English Language Development Test, sinks to the bottom in the flood of latter-day school testing acronyms that Californians are routinely inundated with — CAHSEE, STAR, NAEP, API, AYP and countless others.

CELDT is taken by only one part of the state's students — those classified as EL, English-learners — who make up roughly 25 percent of California's school enrollment. It measures their progress in learning English.

And since new immigrants constantly replace those who move out of the EL category, statewide averages are determined as much by the population mix as they are by how the schools do. If a lot of immigrants enter the pool, scores would probably stay flat or go down.

Still the 2006-2007 scores released last week raise a host of questions, not only about how quickly these students become proficient, but about the rate at which districts, which make the decisions on reclassification, move individuals out of the category. (They then become RFEPs, Redesignated Fully English Proficient).

According to the new test numbers, 32 percent of English-learners are regarded as "early advanced" or "advanced" in English. But just over 9 percent were reclassified last year by their districts. That's not as strange as it sounds since reclassification is based not just on CELDT but on other tests, on teacher evaluations and on parental wishes. They're not to be reclassified until they can "effectively participate" in regular classes.

Still, when 70 percent of graduating EL seniors have passed the English part of CAHSEE, the state's high school exit exam, something seems out of whack.

Because the test was redesigned last year and the scoring made tougher to conform more closely to the state's academic standards, the 32 percent of students rated as "advanced" or "early advanced" is lower than the prior year's 43 percent. But when those scores are rescaled to make them comparable to last year's there's a small gain.

But the real significance of the data is in those questions about reclassification. Reclassified students score higher on other state tests than the average student, even in English. At the same time, according to a state-commissioned study by the Human Resources Research Organization released last fall, "more than half of the 10th-graders classified as English-learners have been registered in U.S. schools for 10 years or more." Many of them are poor and/or learning disabled.

Nonetheless, those numbers, like those for the high rate of EL students who pass the exit exam, are puzzling. Is the low percentage of students who are reclassified and the large number who remain in the EL category year after year due to the fact that they're learning too slowly or because the districts want to keep them there for reasons having nothing to do with their abilities?

Part of the answer, almost surely, is that districts get extra funding for EL students but none for RFEPs, even though state officials acknowledge that RFEPs may need some additional help after they're reclassified.

That creates an obvious incentive — maybe for the good of the students; maybe for the benefit of the district — to keep some students in EL longer than necessary. Nonetheless the state has no plans to provide any additional funds for reclassified students.

How to teach English learners is still something of a black box, despite the hundreds of studies purporting to prove that one or another method is superior to others and the great many true believers who cite those that support their views.

In 1998, California voters passed Proposition 227 to bar bilingual instruction except when parents specifically request it for their children. The big argument for the measure was that students were progressing too slowly into English proficiency.

But the ease with which people learn a new language depends a great deal on their age, on the degree of literacy in their native language, on socio-economic circumstances and on a host of other factors, including the competence of their teachers, regardless of method.

Nor does the state's approved method of training teachers for the task seem to make any great difference. A study released in May by EdSource, which analyzes and publishes data about California schools, found that the presence of more teachers with CLAD/BCLAD credentials, California's formal certificates for those teaching English-learners, made no difference in the achievement of those students.

What seemed to succeed was (1) coherence of program, (2) specific English language instruction in "pull out programs" -- extra classes in addition to regular instruction -- and (3) teaching math using English as a second language and/or immersion techniques.

But the most important thing the numbers tell is how much more California needs to know about how students are taught and reclassified, about the incentives that encourage or discourage districts in getting them qualified and about the fixes the system requires. It's not a trivial issue: These kids will be one fourth of the California labor force of the next generation.

About the writer:

• Peter Schrag can be reached at Box 15779, Sacramento, CA 95852-0779 or at pschrag@sacbee.com.