Exam scores spark debate

New results: Students speaking limited English show improvement.

In the first year under a new law banning most California bilingual education programs, students who are not fluent in English improved their performance on the statewide achievement test, according to results released Thursday.

But while critics of bilingual education claim victory, state officials say it’s too soon for a verdict on the new law, Proposition 227, and warn that other factors such as small classes may have helped boost scores.

It’s too early for anybody to celebrate or throw stones, said Delaine Eastin, state superintendent of public instruction. We need a little more research to see the effects of 227. Even though one in four students who speak little or no English showed gains on the basic-skills test, Eastin noted they still lagged behind their English-speaking counterparts.

A series of technical snafus surrounding scores for limited-English students had delayed the state’s public release of STAR test scores for 4.3 million public school students past the June 30 deadline. The Mercury News published 1999 scores for area schools July 7, using a file it downloaded from the California Department of Education Web site and spot-checked for accuracy. But the paper did not break out scores for limited-English students, since there were questions about the results’ accuracy.

Although state officials said most problems were resolved Thursday, about 420,000 students still were not classified by their language fluency and their scores were not included in either the English-proficient or limited-English-proficient categories. State officials said they will not know for several weeks what effect the omission has on the scores.

In Santa Clara County, limited-English students generally scored higher than the state average. The exception was 11th-grade reading scores, which dropped one percentage point below the average. Mirroring a statewide trend, primary grade students fared better than high school students in reading.

Measure credited

The strides encouraged Proposition 227 supporters, who said state-mandated classes — where nearly all instruction is in English — made the difference.

This absolutely confirms everything we’ve been saying about 227, said Ron Unz, the Silicon Valley entrepreneur who spearheaded the initiative. I’m very encouraged with these numbers. Unz singled out San Diego’s Oceanside Unified School District as the poster child for English-only instruction. Known across the state for aggressively following Proposition 227, the district posted significant gains for limited-English students in the primary grades.

Educators and testing experts, though, suggest the improved performance statewide may have been more a product of a growing familiarity with the two-year-old exam and statewide efforts to reduce class sizes in primary grades. A study released last month showed the number of third-graders from small classes who scored above the national average in math and reading on the state’s achievement test last year was slightly higher than for those taught in larger classes.

State education officials plan to spend the next several months analyzing the test scores to gain a better understanding of 227’s effects on student achievement.

It’s hard to pull out what specific impact 227 has, said Gerry Shelton, a top official with the state’s testing program. There’s been a lot of other things going on, too. Clouding the issue is the fact that not all school districts carried out Proposition 227 in the same manner. Many schools maintained bilingual classes if enough parents obtained waivers to exempt their children from English-only instruction.

San Jose’s Alum Rock Union School District continued to teach about 70 percent of its 7,000 limited-English students in bilingual classes. Still, the district’s English learners improved their scores.

Teacher Kathleen Leclair, who taught a first-grade bilingual class at Alum Rock’s Cesar Chavez School last year, said the improvement is largely due to the district’s heightened emphasis on test preparation — not Proposition 227.

Students drilled

One of the state’s poorest-performing school districts, Alum Rock hired a testing consultant and required teachers to spend three months drilling students on sample standardized test questions.

Districts are really afraid of not doing well on the exam, said Leclair, who has left the district.

Students also posted gains in San Jose Unified School District, the only district in the state to win court approval to continue teaching Spanish-speaking students in their native language. At Washington Elementary School, scores improved by as much as 17 percentage points in math and seven percentage points in reading. The downtown campus, where 78 percent of students are learning English, taught limited-English students almost exclusively in bilingual classes.

If we develop a strong primary language first, the second language will be strong as well, Principal Albert Moreno said.

Others do well

At the same time, schools that taught students predominantly in English also boosted their scores. At Santee Elementary School, in San Jose’s Franklin-McKinley School District, limited-English students in all but third grade improved their scores in every subject. In math, second-graders improved their scores by 19 percentage points. Principal Suzanne LaBare said she was reluctant to attribute the gains to Proposition 227.

We’ll have to wait two or three years to see the full impact of 227, she said.

Much the same as English-fluent students, children still learning English did not do well on the section of the STAR exam tied to the state’s new academic standards. They answered fewer than half of the standards-based questions correctly, results that educators say partly reflect that teaching was not yet tailored to the new criteria.

For the second year in a row, the release of the Standardized Testing and Reporting (STAR) results was riddled with controversy.

Problems with the score reports led the state to release only statewide numbers last month. Harcourt Educational Measurement, the test’s publisher, had mistakenly combined the scores of limited-English students with those who had been reclassified as fluent English speakers.

Even with the release of the remaining STAR scores Thursday, state officials cautioned that data glitches could still exist. Scores for three large Southern California districts were not released by the department because of continuing problems.

Earlier results

The errors reported by the state do not appear to affect the test scores published earlier by the Mercury News, although Eastin said the glitches minutely affected some statewide scores for students in grades 2, 4, 6 and 10. In most cases, those scores are a percentage point below what was originally reported.

Next month, the California Board of Education will determine whether to reduce the state’s payment to Harcourt because of the publisher’s flawed reports. Under an agreement with Harcourt, the board can withhold up to $1.1 million — 5 percent of the $22 million contract.

Board members also have the option of revoking the $2.3 million performance bond posted by Harcourt.

Under the state’s new accountability program, schools will be expected to show annual progress on the STAR exam.

This year’s scores will be used to determine which schools will participate in intervention programs designed to boost achievement at low-performing campuses.

In light of the accountability efforts, Eastin cautioned against scrapping Harcourt in favor of another testing program.

“We have started down this path, and for the time being, California must have these year-to-year scores,” she said.



Comments are closed.