Teaching English-Language Learners: What Does the Research Tell Us?

Leer en español (read in Spanish)

Nearly 3 in 4 American classrooms now includes at least one English-language learner, and these students make up roughly 1 in 10 public school students.

While their numbers continue to rise quickly, the evidence on what works best to help non-native speakers become proficient in English—particularly the more formal academic language needed for school success—has been harder to come by.

What does the federal law say about how schools should approach ELL instruction?

The federal requirement stems from the landmark 1974 case Lau v. Nichols, in which the U.S. Supreme Court found that Chinese-American English-learners in California who were not given educational accommodations to help them learn English did not receive equal access to education.

In essence, this was discrimination due to their language and national origin, a violation of the 1964 Civil Rights Act. Lau‘s mandate has been preserved in subsequent versions of the main federal K-12 law, including the version approved by Congress late last year which states that school districts must take “affirmative steps” to counter students’ language barriers and ensure ELLs can “participate meaningfully in schools’ educational programs.”

In 2015, the Education Department’s office for civil rights issued a letter updating how districts should approach ELLs. Districts must use instructional practices and programs that are backed by scientific evidence and effective in helping students speak, listen, read, and write English and meet challenging state content standards.

What are the most common types of instruction for students learning English as a second language?

Most U.S. schools use variations on one or all of the following:

Pullout/push-in tutoring: English-learners attend core academic classes in English, while being provided separate instructional support in the language either by an ELL specialist during the class or in a separate session outside of class. This method is most often used for English-learners with at least some proficiency in the language.

Sheltered English instruction: English-learners, particularly those with low English proficiency, are taught in a stand-alone classroom. The teacher may focus several hours of the day on direct language instruction as well as academic content. Within a classroom, students often are grouped by their English proficiency so that lessons can be tailored for different levels. Most of these programs are designed to be short—as little as a single year—but some critics have argued that such programs can delay ELLs’ access to regular content. Among the most common versions of this is the Sheltered Instruction Observation Protocol, or SIOP. Three states—Arizona, California, and Massachusetts—have laws requiring sheltered English instruction and limiting the use of bilingual instruction. (California voters will have an opportunity to overturn the restrictions on bilingual education later this year.)

Bilingual instruction: Students receive ongoing language and subject matter instruction in both their native language and English. These programs may serve ELLs only, in a multiyear “developmental” program or a short-term “transitional” program. By contrast, dual-language immersion programs include both native and non-native speakers. These often begin with most of the content taught in the target, or non-English language. Gradually, the time spent teaching in both languages is evenly split, with the goal of making all students exit the program proficient in both languages. This is most commonly used for programs with a high percentage of ELL students of a single native language, such as Spanish or Chinese.

What does research say about the effectiveness of different ELL instructional methods?

While all three main types of ELL instruction have been in use for decades, there is relatively little rigorous research on the general effectiveness of each method, and evidence is particularly scarce on the most effective methods for specific ELL populations, such as young versus older ELLs, or those of different language groups. This is particularly concerning since federal civil rights law requires districts to take into account an ELL’s English-proficiency level, grade, educational background, and in some cases, native-language background to determine appropriate services.

A series of Stanford University studies, including a 2015 study in the journal Educational Evaluation and Policy Analysis, found that English-learners in bilingual programs had language arts and math scores that grew as fast or faster than those of ELLs in sheltered English immersion, but students in developmental bilingual programs showed slower growth in math than those in other types of bilingual and sheltered-English instruction.

Moreover, in 2015, a four-year randomized controlled trial evaluation of the Portland, Ore., dual-language immersion programs found that students who participated in the programs outperformed their other English-learner peers in English-reading skills by a full school year’s worth of learning by the end of middle school.

A rigorous federal research review in 2013 found that no evaluations of sheltered English immersion met its quality standards. There have been a few studies since then, including a cluster-randomized study of Project GLAD, a version of sheltered immersion, which found mixed results for the approach, in part because teachers implemented it very differently from school to school.

“It would be hard right now to do a good [randomized controlled trial] of SIOP because of its broad spread in schools,” said Theresa Deussen, a co-author of the Project GLAD study. “Most teachers don’t use [structured immersion] as a coordinated package of integrated strategies. ... Instead, they think of it as [individual] ‘tools in the toolbox.’”

What instructional practices help ELLs learn academic content?

Regardless of the overall program structure, the Institute of Education Sciences, the Education Department’s research agency, has identified rigorous evidence that the following teaching practices are effective in teaching academic content to ELLs:

• Teach a set of academic vocabulary words intensively, over several days and a variety of activities.
• Integrate instruction in spoken and written English into content-area teaching, such as using science laboratory reports to teach writing in English.
• Provide ongoing, structured chances to develop writing skills.
• Provide small-group interventions for students struggling with specific problems in literacy or language development.

How long does it typically take for English-language learners to become proficient in English?

A landmark study of California ELLs in 2000 found students in both bilingual and sheltered English programs typically took three to five years to become proficient in oral English and five to seven years to become proficient in academic English. This timeline is still generally considered standard by ELL educators, but the new version of the federal K-12 law gives districts three years to bring students to full proficiency and allows them to include former English-learners in the ELL accountability subgroup for up to four years.

A 2015 study by Education Northwest of ELLs entering kindergarten in Washington state found that half reached proficiency in 3.8 years, but 18 percent of the students were not proficient within eight years. The timelines varied significantly by the English level students had upon entering kindergarten, and also by their home language.

For example, Korean-speaking students reached proficiency on average in less than three years, while Spanish-speaking students took on average more than four years. However, the study did not have enough data to suggest why ELLs of different language groups had different rates of learning English.

“It seems like it would be more difficult for a Chinese speaker to learn English than a Spanish speaker, but it doesn’t always hold true,” said Jason Greenberg Motamedi, an Education Northwest senior researcher and the author of the study.

“It may be less the fact that they speak a particular language than other characteristics we can’t see here. ‘Spanish’ may be just standing in for a whole host of other things [such as low income or immigrant status]. Half of the Spanish speakers are second or third generation in Washington. They’ve grown up there, but clearly there are structural barriers that are preventing them from [reaching English proficiency].”

How long it takes students to reach proficiency has a huge bearing on longer-term outcomes.

A 2013 study found English-learners who reached proficiency by the end of kindergarten showed no academic gap with native English speakers, while students who did not reach proficiency by the end of 1st grade showed significant gaps in reading and math compared to native English speakers. While these gaps narrowed in reading over time, they grew in math.

Is effective ELL instruction the same for immigrant and native-born students?

While many English-learners do arrive as immigrants, the vast majority—some 80 percent—are born in the United States and enter U.S. schools at the beginning of their academic careers.

For ELLs who enter the United States before the start of their school years, the instructional approach is generally the same, though Motameti and other researchers’ studies have found that students who enter kindergarten with very low English proficiency take longer to catch up. There have not been significant studies looking at whether particular instructional approaches are more effective for immigrant versus native-born ELLs who start in kindergarten or preschool.

Research suggests older ELLs, particularly “newcomers” who enter in middle and high school, have needs, particularly in content-area language and instruction, that are quite separate from those of ELLs who were born in the United States or who came in early grades.

A 2015 case study of so-called “newcomer schools” in Ohio and New York City suggested that they can be more supportive environments for older ELLs, but may be associated with lower academic achievement. An earlier three-year study by the Center for Applied Linguistics found that the most effective “newcomer schools” provided: flexible course scheduling; teachers skilled and regularly trained in ELL supports; basic adolescent literacy interventions coupled with ELL interventions; content instruction designed to fill gaps in academic learning; and ongoing monitoring of student progress.

The most effective programs also provided significant extended-learning time, including before and after school, on Saturdays, and in summer. They connected immigrant students with family and social services, and provided supports to help students transition to college, careers, and practical life after high school.

Do federal civil rights laws related to ELL instruction apply to charter schools, too?

Yes. The Education Department’s office for civil rights issued guidance in 2014 confirming that charter schools, like any public schools, must take steps to support students learning English and ensure their admissions, disciplinary, and other policies do not disproportionately affect ELLs or their parents. For example, OCR entered a resolution agreement with the BASIS DC Public Charter School after finding that students who did not speak English at home were not appropriately screened for their English-language skills, and teachers incorrectly believed that only the school’s reading lab teacher was responsible for providing ELL services.

Is there a bilingual advantage?

Students who become fully fluent in multiple languages generally perform better academically than either fluent monolingual students or students who are not fully proficient in more than one language. However, researchers are still not sure how much of an advantage there is or what accounts for it.

In the past decade, cognitive and neuroscience studies have suggested that fully bilingual students can switch between cognitive tasks faster than monolingual students. However, a 2014 analysis in the journal Psychological Science found that studies between 1999 and 2012 that found a link between bilingualism and executive control were more likely to be published than those that found either no effect or a negative result. This suggests that journals may be more willing to publish studies that find bilingual benefits.

What can we expect from ELL research in the near future?

More than 45 states now use one of two English-language proficiency assessments: the World-Class Instructional Design and Assessment, or WIDA, or the new ELPA 21 test. Because these two tests have become so common, researchers are in the process of developing crosswalk studies to compare proficiency and achievement across different states. This would enable better comparisons of different state and district approaches to identifying, supporting, and eventually reclassifying English-learners.

“For the first time ever we can get an image of what proficiency development looks like across the nation,” Motameti said. “A year ago or two we couldn’t do that.”