Bilingual Classes Boost Performance, Study Finds

The most comprehensive national study of bilingual education ever conducted, tracing the long-term academic success of thousands of students, has found that those taught in their native language first fare better than those immersed quickly in English.

But the programs cited as best by two George Mason University professors are also the least common in the nation’s public schools: two-way immersion classes, where English-speaking and foreign language-speaking students sit side by side, learning each others’ languages.

Although bilingual advocates say the study should answer once and for all the nagging questions about how to teach the nation’s immigrant children, their English-only opponents point out that it clashes with other studies favoring more rapid immersion in English and proves only that determining which approach best teaches children depends largely on how researchers define success.

Some critics say the new research--the bulk of which is to be released in February--bears little resemblance to what happens inside California’s classrooms, where they contend that a heavy emphasis on native language instruction prevents many students from fully grasping English.

“If a student isn’t transferred out of bilingual ed by sixth or seventh grade, they’re lost . . . and that’s what’s happening here,” said Francine Hallcom, a linguistics professor in the Chicano studies program at Cal State Northridge. “The taxpayers are paying for these kids to read and write Spanish well.”

For California, which has been embroiled in an increasingly bruising battle over bilingual education, the national study is another reminder that before this state can settle on how to educate its 1.2 million students who do not speak English, it must decide exactly what it wants to accomplish.

Is it more important to move students rapidly into mainstream classes, which may save public money in the short run? Or is the state better served by a slower process aimed at producing future academic gains?

Focus of Studies Differs

“We need to come together and talk about what counts as literacy,” said Kris Gutierrez, an assistant professor of education at UCLA, who has done extensive fieldwork in local public schools. “The linguistic issues . . . [and] the social issues vary tremendously from community to community.”

At first glance, the George Mason findings seem to contradict an internal analysis completed last year of New York City’s schools--which educate children who speak more languages than any district outside of Los Angeles--showing that children immersed in English as early and completely as possible move into mainstream classes faster than those taught in their native language.

And the findings conflict with six years of research on El Paso public schools, which favored a compromise between the two extremes: a brief stint in one’s native language followed by a stepped-up diet of English.

But researchers say the findings of each study are valid. The key difference is what they set out to prove.

George Mason professors Virginia Collier and Wayne Thomas took the longest-term look, trying to uncover links between elementary level instruction and high school performance in five urban and suburban districts. The New York analysis took the shortest view, measuring the ability of its students to test out of bilingual education within three years. The El Paso review sought a middle ground.

The George Mason study also tried to tackle a chronic problem plaguing bilingual education research: high student transiency rates, which make it nearly impossible to track a student for more than a few years.

Rather than following individual students, Collier and Thomas spent four years sorting data from 42,000 students according to their backgrounds--including socioeconomic status and their parents’ education level--and then evaluated the progress of similar students.

By controlling for outside factors that might affect academic success, the researchers say they were able to isolate the influence of factors such as family background. They found that the students who did best in high school, regardless of the kind of bilingual instruction they received, were the ones who had had the most schooling before coming to the United States and those with the best-educated parents.

That, they said, may explain why the English immersion approach is deemed best by some studies. Because school administrators are most likely to place well-prepared students with better educated parents in those classes, their success may have more to do with the educational environment in their homes than with the effectiveness of their bilingual program instruction.

The earlier study of New York City schools found higher test scores among Korean, Chinese and Russian immigrants--who were likely to be enrolled in English immersion classes--than for students who spoke either Spanish or Haitian Creole, who were likely to be enrolled in native language classes and who come from poorer and less-educated families.

Children with those disadvantages need native language classes the most, the George Mason researchers contend, because they are unprepared to shoulder the double burden of learning new academic subjects in a new language.

“Most of the politicians and the people who don’t understand what the kids are going through say they have to learn English first,” Collier said.

“That would be true if they were adults, but a school-age child is going through three things simultaneously: language, academic work and they’re also developing cognitively,” she said. Teaching immigrant children only in English is like “asking an 8-year-old to think at the level of a 3-year-old--just stop thinking for three to four years until you can actually do work at your age level.”

Collier and Thomas drew their conclusions after examining records from students representing 150 language groups--though 70% of them were Spanish-speaking--gathered over the course of eight to 12 years of schooling.

They found that regardless of teaching approach, student showings on a variety of standardized tests remained relatively even through at least fourth grade.

But by seventh grade, the English immersion students’ scores had begun dropping, plummeting to below the 20th national percentile at the end of middle school. Meanwhile, those who had spent five to seven years in either native language or two-way immersion improved to score as well or better than their native English-speaking peers.

Measuring Later Success

From that performance gap, Collier and Thomas concluded that native language instruction gives students a solid academic foundation on which to build in any language, while English immersion’s greater focus on English acquisition causes those students to lose ground in other subjects.

Like each wave of bilingual education research that preceded it, the new study is likely to spark heated debate and spawn sweeping changes in how immigrant children are taught in school districts nationwide.

Just a year ago, bilingual opponents in California were using the New York City results to promote proposals for scrapping native language instruction made by school districts such as West Covina, Hacienda La Puente and Westminster in Orange County.

That pressure led California’s Board of Education to loosen its regulations in July and allow districts to use methods other than native language, as long as they could prove that student progress would not be slowed.

The New York City school board had undertaken its review because of concerns that the state might crack down on the length of time it would pay for a child to attend bilingual classes. New York state funds three years of study, but in the past had routinely granted extensions.

Some California politicians are proposing similar restrictions on bilingual funding here, though no limits have been set.

New York followed students who entered school in fall 1990 and fall 1991, tracking how long it took them to exit the programs and how they scored on reading and math tests after moving on.

Of more than 11,000 limited-English students who entered kindergarten in fall 1990, well over half left special programs within three years.

But the exit rates of those in English immersion versus those in native language classes were dramatically different: for kindergartners, 79% of immersion students moved on within three years, contrasted with 52% of native language students.

English immersion students also scored higher on standardized academic tests once they moved into mainstream classes. In the group that transferred out fastest, about half of those who had followed the English immersion curriculum were reading at or above grade level in English two to three years after they left the bilingual program, compared to 38% of those who followed the native language track. Math tests given in English brought similar results.

Since its release, some researchers have criticized the New York study as superficial. Students who had studied more English were bound to test better in English in grade school, but that says little about whether they are prepared for the mounting academic challenges of middle school and high school, critics contend.

Robert Tobias, director of New York’s Office of Educational Research, said the study was a snapshot of how long it took students to move into the mainstream and “not intended to be a definitive comparison of the relative success of [English immersion] versus bilingual.”

“I think it was taken out of context by politicians and other groups that wanted to further their own agenda,” Tobias added.

New York is evaluating data from subsequent years and Tobias said it appears that the native language students are “beginning to pick up steam” academically as they move through the system.

Before New York’s study came out, the research commonly cited by native language opponents came from a study conducted in El Paso, Texas, between 1985 and 1991.

There, University of Oregon professor Russell Gersten compared a native language program in five schools with a new, more rapid transition program in five others.

That new program began students with a half-Spanish, half-English school day, but moved most of them into all-English classes by fourth grade. The traditional program took at least two years longer to make the same transition.

Test scores for sixth- and seventh-graders were about even for the two groups of students, which Gersten said demonstrated that students could be moved more quickly out of bilingual programs without ill effects.

Surveys of teachers and students found more satisfaction with the quicker transition, and school administrators came to favor that approach because it helped them use bilingual teachers more efficiently.

A shortage of bilingual teachers has been cited as a major shortcoming in California’s schools, and helped influence the State Board of Education decision last summer to allow more flexibility.

Students in schools struggling to meet the state’s native language requirements have frequently found themselves in classes taught either by teachers who speak only English or bilingual instructors who lack formal teacher training and credentials.

El Paso officials reported that their program allowed them to concentrate fully bilingual teachers in lower grades and move them from room to room for the native language portion of the day.

“It seemed very appealing to me: the stress of learning a new language, but in a very supportive environment,” Gersten said. “It was kind of a training wheel program.”