American schools are “failing” students whose second language is English, according to two researchers tracking the progress of those students in 16 school districts around the country.
Since 1996, George Mason University researchers Wayne P. Thomas and Virginia P. Collier, working in collaboration with districts, have tried to get a handle on what happens to English-language learners as they move through the education system and enter mainstream classes. Their federally financed project, among the largest to take a long-term look at such students, is based on student records from 1982 to 2000.
The researchers presented some of their findings during the annual meeting of the American Educational Research Association, held here April 11-14.
By the end of high school, they found, students who started school knowing little or no English trail far behind native speakers of the language on achievement tests. They typically score at about the 10th to 12th percentile on English-language versions of national standardized tests, such as the Iowa Tests of Basic Skills, the Stanford Achievement Test-9th Edition, and the TerraNova edition of the Comprehensive Test of Basic Skills.
Measured in terms of “normal curve equivalents,” which are essentially more evenly spaced percentiles, the gap between the two groups amounts to about 25 normal curve equivalents. Five normal curve equivalents translates to about 15 months of schooling, the researchers said.
“This is a colossal gap,” said Mr. Thomas, a professor of research and evaluation methodology at George Mason in Fairfax, Va. “In each and every school system we’ve ever been in, we’ve never failed to see this type of thing going on.”
The researchers said their findings were essentially a “best-case scenario,” since the study included only foreign-language speakers who had been in the districts at least five years or who had already been formally schooled in their home countries.
What Works?
The researchers concluded that some approaches to teaching students new to English are more effective than others in bridging the achievement gap. Most promising, they said, were one-way and two-way developmental bilingual education programs.
Used in about 300 schools nationwide, two-way, or dual-immersion, programs involve teaching groups of students with limited English in classes with native English- speakers. Half the instruction is given in the newcomers’ first language; half is in English. The one-way classrooms use the same approach but do not include native English-speakers.
When students who were not native English-speakers stay in such classes for six years, they score on a par with other students by the end of high school, the researchers found.
The method that the researchers judged least effective at narrowing the test-score gap—English-as-a- second-language pullout programs—is the approach most commonly used in schools. In those programs, students leave their regular classrooms for instruction in basic English and get no academic lessons in their native languages. In those settings, the achievement gap barely narrows at all by the end of secondary school, the research team said.
The districts involved in the study include a mix of urban, rural, and suburban systems from every region of the country. But only one district—the Houston Independent School District in Texas—has officially agreed to allow the researchers to reveal its role.
The study does, however, include data from California, where a 1998 ballot initiative curtailed bilingual education statewide in favor of all-English classes for students with limited English proficiency.
“It’s not looking pretty at all in California,” said Ms. Collier, a professor of bilingual-multicultural-ESL instruction. The achievement gap there has widened in the past two years despite rising scores among English-language learners, she said, because scores for native speakers have increased faster.
The findings are not likely to end debates over such programs any time soon, said Kenji Hakuta, who headed a 1997 National Research Council panel focusing on instruction of LEP students. He said researchers need more statistical and methodological data to assess the George Mason researchers’ findings.
Mr. Thomas and Ms. Collier promised that data would come in August, when they turn in their final report to the U.S. Department of Education’s office of educational research and improvement.
