Three Types of Bilingual Education Effective, E.D. Study Concludes
Washington--Although they use different amounts of English and Spanish, each of the three most common bilingual-education methods effectively educates Spanish-speaking students with Limited English skills, a long-awaited Education Department study concludes.
The study, which was released last week, followed 2,000 Spanish-speaking elementary-school pupils through model bilingual programs of three types--some that immerse children in English, some that shift to mostly English over four years, and some that ease more slowly into English over six years.
Researchers found that children in all three programs kept pace academically with the general student population and outpaced other at-risk students.
Moreover, they found, children who received significant amounts of instruction in their native language were not impeded from learning English, and children who went through structured-immersion programs with little native-language instruction appeared to have learned mathematics and reading as well as many who had received significant amounts of native-language instruction in the short-term bilingual programs.
As relayed by the Education Department, the study results appear to affirm the agency's current policy, which holds that a variety of programs can be effective and that the methodology should be chosen at the local level.
"Based on this study," said Acting Secretary of Education Ted Sanders, "we can conclude that bilingual education benefits students, and school administrators can choose the method best suited to their students, confident that, if well implemented, it will reap positive results."
Rita Esquivel, director of the department's office of bilingual education and minority-languages affairs, said that, in light of the study, the Administration would endeavor to change the law that allows immersion programs to receive no more than 25 percent of the $121 million in federal bilingual-education funds.
The Reagan Administration earned the enmity of the bilingual-education community by fighting for years to allow more funding for immersion programs, which once were ineligible for any federal money.
Ms. Esquivel expressed hope that the new findings would put to rest the debate that has surrounded bilingual education for more than 20 years, so that, she said, educators "can forget the politics" and make decisions based on children's needs.
But advocates of both intensive English instruction and native-language maintenance were quick to assert that the study had serious shortcomings in its methodology and scope. They said it had examined too few sites and had failed to consider many areas that would have affirmed their respective views.
Advocates of native-language instruction asserted that the department was downplaying study findings that showed that long-term instruction in the native language increased achievement. Department officials said the data were insufficient to allow comparisons between such programs and those that use less, or no, native language.
Meanwhile, Ronald Saunders, executive director of U.S. English, a group favoring a constitutional amendment declaring English the official language of the United States, said the seven-year, $4.5-million study had failed to accomplish its primary objective: directly comparing the relative effectiveness of the three program models.
"My gut reaction is, 'It probably was a waste,"' he said.
The study was divided into two parts. The first analyzed the characteristics of each of the three instructional programs in comparison with their models and each other; the sec ond examined how well each approach helped limited-English-proficient children keep pace with the general student population while developing skills in mathematics, English-language speech and writing, and English reading.
Data were collected over four school years, from 1984-85 to 1987-4 88, in 46 schools in nine districts, located in California, Florida, New Jersey, New York, and Texas.
Children in immersion and early- exit programs were tracked from kindergarten to 3rd grade. Those in late-exit programs were tracked in two cohorts, one progressing from kindergarten through 3rd grade and the other from grade 3 to grade 6.
As defined for the study, the immersion programs offered instruction entirely in English and general ly used the primary language only to clarify English instruction, with the goal of preparing children to be mainstreamed by the end of the 1st or 2nd grade.
The early-exit programs, which sought to move children into the mainstream within the same time frame, provided some initial instruction in the primary language-- mostly in teaching reading skills-- and sought to phase into instruction entirely in English by 2nd grade.
Students in the late-exit pro grams generally spent a minimum of 40 percent of their instructional time working in Spanish to learn Spanish-language arts and the con tent areas. They typically remained in the program through 6th grade, regardless of whether they had been reclassified as proficient in English.
The study initially intended to compare the effectiveness of the three approaches. But the comparison was scrapped, Education Department officials said, after the researchers were unable to find all three approaches within the same school district and concluded that they could not control for variations among districts.
The study drew comparisons only between immersion and early-exit programs--which were found within the same districts--and among three varieties of the late-exit model.
One late-exit model started with substantial native-language instruction and slowly phased to 60 percent English; another consistently exposed students to about 60 percent English instruction; and the third sharply shifted toward English instruction in 2nd grade and used it almost exclusively in grades 5 and 6.$3 The study was based on student scores on standardized and non-standardized tests, as well as observations and interviews. Portions of the study, which was prepared by the California-based contractor Aguirre International, ere released to a National Academy of Sciences panel last fall. (See Education Week, Oct. 31, 1990.)
An N.A.S. review of the study is expected to be completed by December.
Among its other conclusions, the department's study found that:
- About 40 percent of immersion, 44 percent of early-exit, and 28 percent of late-exit students were reclassified as fluent-English-proficient by the end of three years.
- About 67 percent of immersion, 72 percent of early-exit, and 51 percent of late-exit stu dents were reclassified by the end of four years.
- About 79 percent of late- exit students were reclassified by the sixth year in the program. Most students remained in the immersion and early-exit programs much longer than the models called for and longer than the researchers had expected, with fewer than 26 percent in either program being mainstreamed by the end of 3rd grade.
It was clear from interviews that teachers believed students were better off remaining in the programs even after many had been reclassified. Among the late-exit students, the group that received high levels of native-language instruction and that had English instruction phased in over time appeared by the end of the 6th grade to be gaining in math, English-language, and English-reading skills faster than the general student population. The group that was rapidly moved into receiving mostly English instruction appeared to be losing ground in these areas, compared with students in general.
Proportionately more families of late-exit children were from lower-income levels than those of pupils in other programs. Parents of late-exit students were far more likely to help them with homework than were parents of immersion or early-exit students, and late-exit teachers assigned homework more often.
Teachers in all three programs offered a passive language-learning environment, with students responding orally to only half of their interactions, and then typically with short answers to closed-ended questions.
"Teachers are creating a corps of passive students who don't speak and who rely on rote memory to answer questions," said Ms. Esquivel, who asserted that bilingual teacher-training programs need to be improved to encourage more active instruction.
The early- and late-exit-program teachers in the study were required to have bilingual teaching credentials, while the immersion teachers were required to be skilled in the students' primary language and to have credentials in bilingual education or English as a second language.
The report cautions that the programs examined represented the optimal implementation of each model, and that the results are applicable only to programs with the same characteristics as those in the study.
The results also apply only to programs for Spanish-speaking children, as research suggests that children from other language minorities acquire English skills differently. Ms. Esquivel said that about 60 percent of the students in federally funded bilingual programs are Spanish-speakers.
Mr. Saunders of U.S. English said the study was seriously flawed in that it failed to directly compare the effectiveness of all three bilingual programs. He noted, for example, that it did not follow immersion and early-exit students into the 5th and 6th grades to see how they compared with their late-exit counterparts.
Noting that all three program types were found not to effectively teach higher-order thinking skills, Mr. Saunders asked, "Who cares if all of the programs are equal if they are equally poor?"
James J. Lyons, executive director of the National Association for Bilingual Education, meanwhile, said the data showed that "all of these program treatments are better than doing nothing."
He also cited the superior gains of late-exit students, and asserted that the study would have more conclusively favored programs that heavily use the native language if it had tracked students into high school and measured dropout rates.
Vol. 10, Issue 22, Page 1, 23Published in Print: February 20, 1991, as Three Types of Bilingual Education Effective, E.D. Study Concludes