Are Too Many Students Being Identified as English Language Learners?

A new study suggests that the state of California may have massive problems in the way it identifies English Language Learner students in its public schools. In California, where one quarter or 1.6 million students are English Language Learners, this misidentification could be costly for schools and students alike.

The report, by Lisa Garcia Bedolla and Rosaisela Rodriguez of the University of California-Berkeley, suggests that the mechanisms used to identify children for English-learner status may not be flagging the right students. Once children are identified for screenings, the authors assert, the test they are given is too challenging for many, leading to over-identification for services.

No Child Left Behind requires that children be screened for English proficiency and given English language support if needed, but the misidentification of English Language Learners, over-identification of ELL’s for special education services, and consistent low-performance of English Language Learners on standardized tests all indicate that many states have yet to design well-functioning systems for children who are not proficient in English.

For elementary school teachers, particularly those who teach kindergarten and 1^st grade, the screening requirement presents a special challenge: Teachers must identify and test children to see if they speak English proficiently at an age when most children cannot read or write, and formal “testing” is often seen as inappropriate. At the same time, it’s especially important to correctly screen children at this young age, so they don’t waste years of school in classes that don’t support their language needs.

In order to identify children who will be screened for ELL services each year, states require teachers to send a Home Language Survey (HLS) to each student’s parents. California’s survey asks four questions, including, “Which language does your child most frequently speak at home?” and “Which language is most often spoken by adults in the home?” A “yes” answer to any one question flags a child for screening.

Bedolla and Rodriguez suggest that in California, many children are over-identified for screening because of the second question: “Which language is most often spoken by adults in the home?” If a child lives with both parents and grandparents, for example, it’s entirely possible that that child speaks mostly English with their parents, but that the parents and grandparents communicate with one another in Chinese. According to the report, only 12 percent of children who take the California screening test qualify as English proficient, indicating that the test may be too challenging for kindergarteners and 1^st graders to understand. If this is the case, there’s a good chance that this child might take the test and fail, even if he has a solid understanding of English for his age.

Assuming a district hasn’t experienced a sudden demographic change, the proportion of students who are flagged to take the EL screening test should be roughly correlate with the proportion of English learner students in the district. In over half the districts sampled for the study, however, rate of testing was 20% higher than the total percentage of English learners, indicating that too many children were being flagged for services in the first place.

Home Language Surveys appear straightforward, but conflicts surrounding the surveys are becoming more frequent. When Arizona streamlined its Home Language Survey in the summer of 2009 to ask only one question—“What is the primary language of the student?”—the number of ELL students in the state dropped by over 30,000 in a single year because the survey was flagging too few children for ELL screening. As Early Ed Watch reported at the time, the state settled an investigation with U.S. Department of Education’s Office for Civil Rights and the U.S. Department of Justice soon after and reverted to its original survey.

When I was conducting research for a forthcoming issue brief, a consultant who works with Illinois preschools and school districts told me she encourages providers to ask many questions on a Home Language Survey, but use only some of them to refer children for screening assessments. This is because providers frequently know too little about what mix of languages children hear at home, or whether parents are encouraging a child to learn only English, or are encouraging them to speak their home language as well. Gathering this information about kindergarteners and first-graders, who are less likely to explain their language backgrounds to teachers on their own, can be useful to both ELL and mainstream teachers.

The screening test taken by kindergarteners and first graders in California include reading and writing assessments and, according to the report, can take up to two hours to explain and implement. Additionally, those administering the test are often not specially trained or certified to do so—all factors that lead the authors to conclude, “it is unlikely that districts are targeting their scarce language resources as effectively as possible.” Most troubling, of the 37 districts surveyed, only fifteen percent had an appeal structure in place for teachers or parents to use if they felt children were misidentified. If too many children are being referred for testing and not enough of them can pass the screening test, it’s likely that there are children sitting in ELL classes who do not need to be there—and who don’t have a route to challenge their ELL designations.

It’s true that ELL instruction is costly and, in many areas, scarce. For true English Language Learners, the elementary years are key: Children who are well-supported are more likely to learn English fast and move into mainstream classrooms for the rest of school, while others struggle in years of special ELL instruction or, most inefficient of all, bounce between ELL and mainstream services throughout their school years. States have a long way to go when it comes to capitalizing on their existing ELL services and expanding them where needed. But first they need to be identifying the right students for special services.