Five years ago, researchers from the University of California, Davis, took a look at reading-exam scores across the K-12 grade span for students growing up in non-English-speaking homes.
Predictably, the researchers found that the students who struggled the most with learning English lagged well behind their English-speaking peers at all levels of schooling, never really catching up at any point along the spectrum.
A more curious pattern emerged among students with stronger English skills. During the first few years of school, this group鈥檚 achievement levels were almost on par with those of English-speaking students. But the more skilled English-learners began to drop back after 4th grade. By middle and high school, the gap separating them from the higher-achieving English-speaking students stretched into a chasm.
鈥淲e don鈥檛 know what that鈥檚 due to,鈥� notes Kenji Hakuta, an education professor at Stanford University who has long studied this population of students. 鈥淵et, these are the most successful of our English-language learners in schools.鈥�
To Hakuta and others, the pattern suggests much about what scholars know鈥攁nd what they have yet to learn鈥攁bout teaching the estimated 5.1 million English-language learners in the nation鈥檚 schools.
Researchers have learned a lot, for instance, about how to teach basic reading skills in the early grades to English-language learners. What they have yet to nail down is how to help this vulnerable and challenging population of students over the learning hump that comes later in elementary school; how to teach higher-order reading skills, such as comprehension; how to teach adolescents who are new to English; and how to boost achievement in academic subjects other than English.
鈥淭he bad news is that we鈥檙e not where we鈥檙e supposed to be. There鈥檚 a lot we don鈥檛 know,鈥� says Claude Goldenberg, who is also a professor of education at Stanford. 鈥淭he good news is that the research is growing.鈥�
More Study Needed
That answers are urgently needed now, as the nation鈥檚 ranks of non-English speakers grow to historic levels, goes without saying. Goldenberg has noted, for instance, that on the 2007 National Assessment of Educational Progress tests in reading, 4th grade English-language learners scored an average of 36 points behind non-ELLs in reading and trailed by 25 points in math on a scale of 500.
Yet the pool of credible studies on how best to teach these students is far shallower than it is for other much-debated areas of education, such as general reading. Goldenberg says, for instance, that when the National 69传媒 Panel met to synthesize the research in that subject, it found 400 studies that met its methodological criteria. In comparison, the National Literacy Panel on Language-Minority Children and Youth, focusing on English-language learners, turned up just 17 in its own review of the research.
Making matters worse, the existing research on the topic has been dominated by a single, politically explosive question: Should English-language learners be taught, either initially or for an extended period of time, in their native languages?
Goldenberg and other experts say that five independent research reviews addressing that question over the last 25 years conclude that teaching students in bilingual settings is more effective鈥攁t least modestly so鈥攖han teaching them only in English.
鈥淚 think the evidence is there,鈥� says Diane August, a senior research scientist at the Center for Applied Linguistics, a private research center in Washington. 鈥淭here鈥檚 a lot of transfer that occurs from the first language to the second language.鈥�
Still, some experts remain skeptical of that research. One is Russell Gersten, a professor emeritus and co-founder of the Instructional Research Group, a private research firm based in Los Alamitos, Calif., who describes studies on that question as 鈥渋nconclusive.鈥�
鈥淓ven those studies that are considered the 鈥榞old standard鈥� have major problems, such as having just one teacher per [experimental] condition, so you don鈥檛 know if it鈥檚 the teacher or the method that鈥檚 making the difference,鈥� he says.
As part of federal reporting requirements, states document the types of language-instruction programs funded under Title III of the No Child Left Behind Act. All but two of the 48 reporting states supported English-only instructional programs. The majority of states (36) also provide programs taught dually in English and another language. The specific Title III programs most commonly reported are content-based English-as-a-second-language (ESL) and ESL pull-out instruction, used in 43 and 42 states respectively. Only 15 states offer developmental bilingual programs.
NOTE: Data not available for California, Michigan, and Virginia.
SOURCE: EPE Research Center, 2009. Analysis of data from Consolidated State Performance Reports, 2006-07.
Scholarly views diverge even more over how long it should take for students to master English, with estimates ranging from three to eight years.
One hope is that the small boom in studies under way now, much of it fueled by the U.S. Department of Education鈥檚 Institute of Education Sciences, will yield more definitive outcomes.
For instance, Johns Hopkins University researcher Robert E. Slavin and his research partners are tracking 661 Spanish-speaking students from across the nation who, upon entering kindergarten, were randomly assigned to either English-only classes or to one form or another of a transitional bilingual education classroom. In the latter settings, students are taught in Spanish for some part of the school day in kindergarten, 1st, and 2nd grades.
The final results from that experiment are not due until next year, after the students complete 3rd grade. However, Slavin, like a growing number of other experts, contends that the language-of-instruction issue may turn out to be of limited value to policymakers grappling with whether to implement鈥攐r steer clear of鈥攂ilingual education.
鈥淢y guess is we鈥檙e not going to find much difference between kids taught in Spanish and kids taught in English all along,鈥� Slavin says.
Doing Something Well
Even if experts agreed that teachers should incorporate students鈥� native languages in lessons for English-learners, many districts don鈥檛 provide such instruction.
Urban districts with large and growing immigrant populations may have a hard time finding teachers conversant in Haitian Creole, Urdu, or the dozens of other languages their students speak. And seven states鈥� Arizona, Arkansas, California, Connecticut, Massachusetts, New Hampshire, and Wisconsin鈥攅ither ban or restrict the use of native-language instruction with English-learners, according to a survey by the Editorial Projects in Education Research Center.
鈥淵ou can have bilingual instruction and do it poorly, and you can have English-only instruction and do it poorly,鈥� says David J. Francis, a psychology professor at the University of Houston and the director of the federally funded National Center for Research on the Educational Achievement and Teaching of English Language Learners, which is based at the university. 鈥淢y argument has been to figure out what it is you can do best, do it well, and worry less about the language of instruction.鈥�
On the other hand, researchers widely agree on the idea that teaching basic English-reading skills to young English-language learners is not very different from teaching young English-speakers how to read.
When teachers teach students how to decode words, how to spell, and how to recognize phonemes, which are the basic sounds that make up words, ELLs nearly catch up to English-speaking classmates in the early grades, according to the research. That鈥檚 the point illustrated in the statistical patterns that turned up in the University of California, Davis, study, which was conducted by researchers Patricia G谩ndara, Russell Rumberger, Julie Maxwell-Jolly, and Rebecca Callahan.
Beyond Decoding
Teaching students to comprehend what they read, particularly when it comes to more academically oriented text, is another matter.
鈥淲e鈥檝e done a good job of building up reading skills to the point where students can decode words and read them, but they don鈥檛 necessarily have the language abilities that would allow them to construct a representation of the text at a very high level,鈥� says Maria S. Carlo, an associate professor of teaching and learning at the University of Miami in Coral Gables, Fla.
In an effort to build a bridge to higher literacy levels, Carlo, August of the Center for Applied Linguistics, and their colleague Isabel Beck developed a curriculum for teaching 4th and 5th graders in mainstream classes vocabulary words that are commonly found in print, but rarely used in conversational English or on the playground鈥攚ords such as 鈥減redict,鈥� or 鈥渢herefore,鈥� or 鈥渋ncremental.鈥�
The target words are embedded in weekly English-language reading assignments that the minority-language students 鈥減review鈥� on Mondays in their native language. On Tuesdays, the teacher defines the words for the whole class and leads a discussion on the reading, which the ELL newcomers then read in English. Throughout the rest of the week, working in small groups with English-speaking classmates, students learn synonyms and antonyms for the targeted words, pick out their root structures, and discuss any multiple meanings for the words, among other activities.
In a 2004 experiment on the program, researchers documented reading and English-language-development gains for both the English-learners and their classmates after 15 weeks of lessons. The gains paralleled one another, so the gap in achievement between the two groups didn鈥檛 budge initially, according to Carlo. A subsequent study showed, however, that the learning differences narrowed when students stayed in the program for two years.
Such studies come amid growing interest in finding ways to help English-language learners cultivate 鈥渁cademic English,鈥� the language they need to succeed in subjects beyond English literacy.
鈥淢astery of academic language is arguably the single most important determinant of academic success for individual students,鈥� Francis and his colleagues write in a research-based guide for educators of ELLs, adding that its importance 鈥渃annot be overstated.鈥�
Research has yet to identify a proven strategy for imparting those skills to ELLs, although a flurry of studies are under way in Francis鈥� lab and elsewhere.
Such methods will need to do more than teach vocabulary words if they are going to make a difference in students鈥� overall learning, according to August. They have to teach content, too.
鈥淭ake 鈥榩hotosynthesis,鈥� 鈥� she says. 鈥淚t鈥檚 not like you can give kids the definition and they get it if they don鈥檛 understand the science behind the word.鈥�
Talking the Talk
Researchers are also zeroing in on oral-language skills in helping English-learners overcome academic roadblocks.
鈥淟ots of ELLs are very shy,鈥� says Slavin of Johns Hopkins. 鈥淭hey don鈥檛 want to use English in class because they鈥檙e afraid they鈥檒l be laughed at or they don鈥檛 feel confident.鈥�
To prod students to talk more, especially in the academic arena, many experts recommend setting up structured cooperative-learning groups so that students can practice speaking under less-threatening circumstances. In fact, a research-based practice guide published last year by the Institute of Education Sciences calls for English-learners to spend at least 90 minutes a week working one-on-one on carefully designed activities with students of different ability and English-proficiency levels.
To Goldenberg and others, however, the bottom line is that the research suggests that English-learners need some sort of classroom support if they are ever going to succeed in American classrooms. 鈥淵ou cannot do sink or swim,鈥� he says.
Yet he estimates that 10 percent to 50 percent of ELLs are in classrooms where few, if any, modifications are made to help them overcome their language difficulties. And their numbers are growing, he adds, even as pressure builds in some states to enact policies that block teachers from using students鈥� primary language in classes or limit instructional modifications for English-learners.
Such practices, he concludes in an essay published in the Summer 2008 issue of the American Federation of Teachers鈥� American Educator magazine, 鈥渁re simply not based on the best scientific evidence available.鈥�
