Los Angeles Times

Monday, March 2, 1998

Amid debate over bilingual education comes supportive research from L.A. Unified and Tomas Rivera Policy Institute. But with critics calling research seriously flawed . . .
Opinions Vary on Studies That Back Bilingual Classes
By AMY PYLE, Times Education Writer

As the debate over bilingual education bounds toward a spring ballot initiative, two studies scheduled to be released today land on the side of teaching children in their native language first, then gradually switching to English.
     However, critics of bilingual education and even some supporters raised questions about the studies.
     One, by the Los Angeles Unified School District, focuses on students who remained at the same elementary school from first through fifth grade--a stability that is unusual in the state's largest school system. When the 4,200 students were given standardized English tests in fifth grade, those who had come through the native language bilingual program fared better than those who had been enrolled only in tailored English classes known as English Language Development.
     The other study, by the Claremont Colleges-based Tomas Rivera Policy Institute, registered a similar gap based on a survey of 11 previous studies of bilingual programs across the country.
     Political scientist Harry Pachon, the institute's director, said the findings shift the onus of proof to supporters of June's anti-bilingual initiative, English for the Children, which would replace bilingual education with just one year of English immersion.
     "It's up to them to prove that bilingual education isn't working," Pachon said. "If aspirin reduces a headache and we have data to prove it, why remove aspirin as an option?"
     But supporters of Proposition 227 described the research as heavily flawed--noting that of the 11 studies covered by the institute's survey, nine are more than 17 years old, and alleging that L.A. Unified could not see beyond its vested interest in the status quo.
     Ron Unz, the initiative's author, said he was particularly disturbed that even among this elite group of students--who stayed put for all of elementary school and received a consistent program--just 39% of the native-language-program students graduated to mainstream English by the end of fifth grade. The district's current goal is for all students to be weaned from bilingual education within five years.
     "Are they proud about the fact that over 60% of these students didn't redesignate?" Unz asked.
     The Unz campaign received an early boost from polls showing that a majority of the state's voters favored it, though the lead has decreased in more recent polling. The opposition campaign has kicked into gear in recent weeks, bolstered by such backing as the recent decision to oppose the initiative by the American Educational Research Assn., an international group of scholars and researchers.
     Even some of those who generally support bilingual education questioned the validity of the new studies. Of the L.A. Unified research, for instance, board member David Tokofsky remarked: "How can you have an evaluation done by your own people? Outsiders will just laugh at that."
     Indeed, a closer look at the L.A. Unified study raises some troubling questions that sent shudders through the district late last week as the release date approached. In particular, district staff acknowledged that 3,000 native-language-program students were not counted because they did not read English well enough in fifth grade to be tested on the English language Stanford 9 test. In the English Language Development comparison group, however, all students were tested regardless of their mastery of the language.
     Forrest Ross, the district's elementary bilingual administrator, said a hurried look at more comparable test scores--including only those students in the top English levels for both groups--suggested a smaller but still consistent gap favoring native language.
     The community activist who worked with parents at Ninth Street School to stage a student boycott two years ago to lobby for English classes said, "Their study makes our point; we can thank them for their study. They have described the problem in the LAUSD."
     Times' coverage of Alice Callaghan's protest caught Unz's eye, leading him to propose the initiative. It would essentially wipe out both types of L.A. Unified programs in favor of full English immersion, except where parents attain waivers from school boards.
     What the L.A. Unified study shows, Callaghan said, is that all limited-English-speaking children are failing, "which is exactly why we're proposing something different."
     Ross countered that Unz's proposal is completely untested.
     Data on what works best in acquiring language is plentiful, but also contradictory. Many studies indicate that students taught in their native language first need between five and seven years to transition to English, but perform better by the time they reach high school. Others have cited a negligible difference among the various programs, suggesting that public school districts might be smart to err on the side of quicker English immersion programs.
     Of the battling research, the Tomas Rivera report concluded: "Both sides have claimed that scholarly research supports their respective positions. Their reading of the literature, however, is often selective, exaggerated and distorted."
     The institute, on the other hand, claims that its assessment is "unsentimental," because it gathered a range of studies from many different camps and included only those that met stringent criteria.
     Yet even the institute reaffirms the most common finding of such reviews: "The vast majority of evaluations of bilingual programs are so methodologically flawed in their design that their results offer more noise than signal."
     That problem, the report concludes, makes it difficult to address one of the most pressing questions about bilingual education: How long should students be enrolled in such programs and what is the ideal ratio of native language versus English instruction?
     Government professor Jay Greene, who produced the Rivera study, said he was surprised by its conclusions because of his own mixed feelings about bilingual education.
     "I was skeptical," said Greene, an assistant professor at the University of Texas at Austin. He added that in the end the greater problem is the low quality of education in large U.S. cities.
     "I think it's probably true that quality of instruction matters much more than the language of instruction," he said. "But if we want to move toward higher quality instruction, it won't help to eliminate [the native language] option and in fact it's much more likely to hurt."
     * * *

     Performance Comparison
     The Los Angeles Unified School District compared test scores of students who had remained at the same elementary school for at least five years. The study found that Spanish-speaking students enrolled in bilingual programs outdid those who had taken the more English-intensive English Language Development Program. But both groups fell below the district's median percentile.

Results of Stanford 9 English Test, Grade 5

(English Only)
Reading 28 21 31
Language 35 29 37
Math 33 26 34