March 2, 2004 | Volume 2 | Number 5
The Effects of Bilingual Education Programs on English Language Learners
The Effects of Bilingual Education Programs on English Language Learners
How effective are bilingual education programs for improving the English reading skills and general content knowledge of English language learners (ELL)?
Bilingual education has been a political hot-button issue in many states. Initiatives in Arizona, California, Colorado, and Massachusetts (to name only a few states) have sought to curb the use of bilingual programs for ELL instruction and to replace these programs with English immersion. In a recent meta-analysis of research on bilingual education, Jay Greene points out that both supporters and critics of bilingual education claim that scholarly research supports their positions. According to Greene's meta-analysis, however, bilingual education is significantly more effective at improving ELL achievement. Another recent systematic review of research on the effectiveness of bilingual programs for improving student reading, conducted by Robert Slavin and Alan Cheung, found that students who were taught to read both in their native language and in English outperformed their peers in English-only programs on tests of reading achievement. Despite these general findings, bilingual education programs are not monolithic. A wide variety of programs emphasize primary and secondary language instruction to differing degrees and with different levels of effectiveness.
Wayne Thomas and Virginia Collier conducted the study highlighted in this issue of ResearchBrief (see below for full citation). The study examined the type of instruction English language learners (primarily Spanish-speaking students) received in five school districts over five years (1996–2001), as well as the achievement of these students on English and Spanish tests of academic problem solving across the curriculum (math, science, social studies, and literature). The five districts encompassed more than 210,000 students and included an inner-city district, a large and a medium urban district, and two rural districts. Students were tracked as they progressed through the programs from kindergarten or 1st grade through 4th or 5th grade. The study had both qualitative and quantitative components; however, this ResearchBrief focuses on only the quantitative aspects of the research.
Districts were selected for the study on the basis of recommendations from state agencies, as well as their willingness and ability to initiate, conduct, and maintain instructional reform and collaborative research processes. Readily available student-level data and the technology to compile such data were also important. Districts in 26 states were examined, and 16 sites in 11 states were chosen for the project. This selection ensured a viable sample for analysis at the end of the data collection (researchers thought that some of the school districts might undergo changes during the five years that would compromise the programs under examination or the data collection process). Researchers identified eight types of ELL programs:
- 90-10 two-way bilingual immersion: Two language groups receive integrated instruction in English and a second language, in this case, Spanish. In a 90-10 program, 90 percent of instruction is initially delivered in the minority language (Spanish), and 10 percent of instruction is in English, gradually evolving to 50-50 instruction over five years.
- 50-50 two-way bilingual immersion: The two language groups receive half their instruction in English and half in Spanish.
- 90-10 one-way developmental bilingual education: In one-way bilingual programs, one language group is taught using two languages. As explained in the previously mentioned 90-10 program, 90 percent of instruction is initially delivered in the native language, 10 percent in English, evolving to a 50-50 mixture.
- 50-50 one-way developmental bilingual education: One language group receives half the instruction in the native language and half in English.
- 90-10 transitional bilingual education: ELL students receive 90 percent of their instruction in their native language and 10 percent in English until grade 5, followed by immersion in the English mainstream.
- 50-50 transitional bilingual education: ELL students receive 50 percent of their instruction in English and 50 percent in their native language over three or four years, followed by immersion in the English mainstream.
- English as a second language (ESL): ELL students receive bilingual and ESL instruction for two or three years, followed by immersion in the English mainstream.
- English mainstream: All bilingual and ESL services are refused, and the student is initially placed in the English mainstream.
Student achievement within these programs was measured by looking at the achievement gap between ELLs and non-ELLs and the degree to which each intervention narrowed that gap.
The complete study (available at the link below) includes both qualitative and quantitative findings, as well as additional analysis stemming from more complex regression models and blocking of data. Stated here are the general findings regarding the effect on long-term student academic achievement in bilingual and English immersion programs. The findings are based on students' performance on relevant district tests in their most recent grade level.
- 90-10 two-way bilingual immersion: These students completed 5th grade with scores at the 51st percentile on standardized reading tests.
- 50-50 two-way bilingual immersion: By the end of grade 5, 58 percent of these students exceeded Oregon state standards for English reading.
- 90-10 one-way developmental bilingual education: These students reached the 34th percentile by grade 5.
- 50-50 one-way developmental bilingual education: These students scored at the 61st percentile by grade 7.
- 90-10 transitional bilingual education: These students scored at the 32nd percentile on standardized reading tests at the end of grade 5.
- 50-50 transitional bilingual education: By grade 11, these students scored at the 45th percentile on standardized reading tests.
- English as a second language (ESL): ESL students scored at the 23rd percentile by graduation.
- English mainstream: ELL students mainstreamed without bilingual support showed large decreases in achievement (three-fourths of a standard deviation by grade) when compared to their peers in bilingual programs. The immersion group accounted for the largest number of dropouts and scored at the 12th percentile on standardized reading tests.
Students in bilingual programs continued to perform well on tests of Spanish achievement and scored higher on other core academic subjects. Dual language (two-way) programs generally explained a greater percentage of achievement scores than did socioeconomic status.
The Bottom Line
Well-designed and carefully implemented bilingual education programs can have a significant positive effect on student achievement both in English literacy and in other academic core courses when compared to English immersion.
English language learners and schools instituting ELL programs.
A number of important caveats relate to this study. While students who bypassed bilingual services were outperformed by their peers in such programs, intentional immersion programs—which may differ from immersion as defined in this study—frequently include some native language instructional experience and other supports for ELL students. This study also focused on young learners entering ELL programs in kindergarten or grade 1 and may not be generalizable to older students entering the system. (In fact, one of the strongest predictors of education success for ELL students is the extent to which they received educational services prior to entering the United States.)
Program implementation and appropriate assignment of students is also important, and this study focused on well-designed and fully implemented programs. The population in this study was largely limited to Spanish speakers, so the effects of bilingual programs may be different for students with other language backgrounds. The purposive selection of districts may limit the degree to which these findings can be generalized to other districts.
Finally, the program findings may be explainable by other district characteristics, since every type of program was not necessarily available in each district.
Thomas, W., & Collier, V. (2002). A national study of school effectiveness for language minority students' long-term academic achievement. Santa Cruz, CA and Washington, DC: Center for Research on Education, Diversity & Excellence.
A Meta-Analysis of the Effectiveness of Bilingual Education
Effective Reading Programs for English Language Learners: A Best-Evidence Synthesis
All comments regarding ReseachBrief should be sent to RBfeedback@ascd.org. To speak directly with an ASCD staff member, please Contact Us.
Dan Laitsch serves as ASCD's consultant editor for ResearchBrief. Laitsch is an assistant professor in the Faculty of Education at Simon Fraser University in British Columbia, Canada, and is coeditor of the International Journal for Education Policy and Leadership.