Study Challenges 'Model' E.S.L. Program's Effectiveness
FAIRFAX COUNTY, VA--The English-as-a-second-language program here, often hailed as a model alternative to bilingual education, may be less effective in developing the language skills needed for school success, according to a long-term study released this month.
Researchers found that it took Fairfax ESL students four to nine years to reach grade level on standardized tests in reading and other subjects. They concluded that certain groups of children "might acquire English for academic purposes more rapidly'' if they received at least two years of instruction in their native tongue.
The study, by Virginia P. Collier and Wayne P. Thomas of George Mason University, is the first in the United States to compare the rates at which children of different ages achieve "cognitive-academic proficiency'' in a second language. And it is one of the few attempts to chart the progress of limited-English-proficient students in an all-ESL program.
Educational researchers and advocates for language-minority children are calling the study significant for both theoretical and political reasons.
Its findings appear to confirm the hypotheses of Jim Cummins, a Canadian researcher who has influenced the design of bilingual programs in this country, regarding the "transfer'' of literacy and subject-area mastery between languages.
The results are also raising new questions about the decision by the Congress, in reauthorizing the Bilingual Education Act through 1993, to permit expanded funding for programs featuring no use of the native language.
Proponents of the change, including Secretary of Education William J. Bennett, had cited Fairfax County, where at least 75 languages are spoken in the schools, as an example of the need for "flexibility'' in instructional approaches. Opponents had argued that evidence for the effectiveness of "English only'' approaches was insufficient to support the diversion of scarce bilingual-education funds.
Fairfax County was not identified in the study, as it was presented April 7 at the annual meeting of the American Educational Research Association in New Orleans. But district officials confirmed that Fairfax students had been the subjects; they stressed, however, that the study was not a full-fledged evaluation of the Fairfax ESL program.
The researchers focused on a subgroup of 2,014 LEP students--those who had entered school with grade-level skills but no exposure to English, and who received varying amounts of ESL instruction over a one-to-three-year period.
Using a "cross-sectional'' approach rather than the longitudinal method of following individual children, Ms. Collier and Mr. Thomas examined student achievement as measured by Science Research Associates tests in reading, language arts, social science, mathematics, and science. They examined scores for the 4th, 6th, 8th, and 11th grades over a six-year period beginning in 1981.
Some 20 percent of the students in the study were Hispanic and 65 percent were Asian, including speakers of Korean, Vietnamese, Chinese, Khmer, Lao, and Urdu. Coming largely from middle-to-upper-class backgrounds in their home countries, they were deemed "advantaged,'' even if their families had low incomes in the United States, in part, because of parents' "strong middle-class aspirations.''
The study found clear differences between age levels in the ease with which pupils learned English. Children who arrived at ages 5-7 scored poorly in all subjects, as compared with those who started the ESL program at ages 8-11. By the 6th grade, older arrivals were outperforming younger arrivals who had been in the United States two to three years longer. A third group, those who arrived at ages 12 to 15, did worst of all.
"The only known variable that differentiated'' the two youngest groups, the study said, was the limited mother-tongue schooling of the 5-to-7-year-olds as compared with the older children.
According to the researchers, these age-group findings are consistent with Mr. Cummins's hypothesis that at least two years of literacy development in the first language is the necessary "threshold'' for the transfer of literacy and other cognitive skills to the second language.
Mr. Cummins, a researcher at the Ontario Institute for Studies in Education, has conducted similar research with Canadian children. He distinguished between two types of second-language proficiency: "basic interpersonal communications skills,'' or "playground English,'' and "cognitive-academic language proficiency,'' or the more abstract linguistic tools required in the classroom.
While playground English takes two to three years to achieve in the second language, he concluded, the latter takes five to seven years. And it takes longer and is more difficult for LEP children who begin school in all-English programs rather than in a language they already know, he found..
Programs that have applied Mr. Cummins's theories in the United States, such as the Case Studies project of the California State Department of Education, have stressed intensive native-language development--as a foundation for long-term achievement--and a later transition to English mainstream classrooms (See Education Week, April 1, 1987.)
Older Arrivals Struggled
In Fairfax County, where there is no bilingual instruction, the researchers concluded that the 8-to-11-year-old arrivals were quicker to achieve cognitive proficiency because they had gotten a head start in their native tongues. "The cognitive base in first language seems to help them significantly with their second-language schooling,'' the study said.
On the other hand, the 12-to-15-year-olds were most affected by the language barrier.
The study attributed this phenomenon to the quicker pace and higher cognitive demands of the upper grades.
In mathematics, newly arrived 8th graders scored well above the 50th percentile, reflecting the transfer of skills learned back home; particularly in Asia, math curricula tend to be more demanding than in this country. But over the next three to four years--the time it took to become proficient in English--these students slipped steadily behind their classmates.
The Collier-Thomas study suggested that in Fairfax both the youngest and the oldest groups would benefit from native-language instruction, but for different reasons. Bilingual methods would speed up the acquisition of cognitive language skills for the youngest children, and help the oldest students keep up in class, the researchers said.
In an interview, Ms. Collier noted that Fairfax County's mean achievement-test scores are well above national norms. LEP high-school students, despite high aspirations, cannot hope to compete for college admission if they fall behind while learning English, she said.
"I find this study very provocative,'' said Esther Eisenhower, director of the Fairfax ESL program. But she cautioned against drawing firm conclusions until the district replicates the findings. She said a thorough evaluation of the program, assisted by Ms. Collier, is now under way and should be completed by May 1989.
The poor performance of the 5-to-7-year-olds "threw me for a loop,'' Ms. Eisenhower said. But the question, she added, is "whether this is a transitory or permanent negative effect, whether they are handicapped for their [entire] academic career.''
At the same time, she questioned some of the researchers' premises--such as whether the LEP students were really at grade level on arrival and whether they were more "advantaged'' than other immigrants.
She also said the study failed to consider the length of time children spent in ESL programs, a point on which the district has no reliable data.
Asked whether she would consider the study's recommendations for native-language instruction, Ms. Eisenhower said no alternative would be ruled out. But she added that she had not changed her assessment of bilingual education since 1980, when Fairfax officials won a protracted battle with the federal office for civil rights over their alleged neglect of LEP children.
In 1980, Fairfax County became the first district in the nation to be allowed to remedy such civil-rights violations without providing bilingual instruction.
"Given scattered and multilingual populations like ours, in an English-only treatment--if it is provided properly--the children can achieve just as well,'' Ms. Eisenhower said. "We would be hard put to find the number of teachers to deliver the Fairfax County curriculum'' in 75 languages.
Although Spanish-speaking teachers could be found, she said, "I cannot in all fairness say that the 20 percent of Hispanics we have here are entitled to a treatment that the Asians are not.''
U.S. Education Department officials, who have often lauded "the Fairfax model,'' had no official comment on the Collier-Thomas study.
'A Long, Long Time'
"It's difficult for me to talk about the significance of the results without seeing more numbers in each category,'' said Edward Fuentes, director of research for the office of bilingual education and minority languages affairs. "It's an ongoing study. You have to look at it when it's complete, take it in toto.''
A different assessment was voiced by Richard Tucker, president of the Center for Applied Linguistics. "The evidence is pretty strong,'' he said, on the validity of Mr. Cummins's theories about the transfer of cognitive skills. Even though the Collier-Thomas study did not control for all possible variables, he said, the numbers of children involved and the consistency of the age-group patterns "lend confidence'' to the results.
Mr. Tucker also endorsed the study's most general finding--as expressed by Ms. Collier--that acquiring academic proficiency in a second language "is a process that takes a long, long time.''