Blog Post

Toward a New Research Agenda to Improve Outcomes for Adolescent English Learners

In the wake of large-scale immigration since 1965, English learners (ELs) have become a significant proportion of the American public school population. These young people, who come from homes where a language other than English is spoken, now comprise 10 percent of all PK-12 students in the U.S. And in states with substantial immigrant populations this percentage is even higher—25 percent of California’s public school students and 15 percent of those in Arizona are English learners, for example.

Yet, as a group, ELs are “disproportionately underserved and underachieving.” English learners require academic supports to access the mainstream curriculum—they must often learn academic English while also learning content-based knowledge.

Given the scope of the challenges facing ELs, we recently brought together a group of scholars from diverse disciplines, a longtime practitioner, and staff from the Institute of Education Sciences (IES), to discuss the current state of knowledge about this particularly underserved population. The goal of this discussion was to begin developing a more comprehensive research agenda to identify better supports for adolescent English learners. This goal aligns with the Foundation’s focus on reducing inequality, through which we seek to support research that helps build, understand, test, and improve responses to inequality—be they programs, policies, or practices. In this focus area, we are especially interested in reducing inequality in outcomes for immigrant-origin youth, many of whom are ELs.

Participants identified three key issues that researchers should be aware of.

First, we need to build capacity in educational settings that serve English learners. This goal—long shared by educators, administrators, researchers, and policymakers—has been elusive. Both in districts that have been receiving immigrants for many decades (e.g., New York City or Los Angeles) and districts that have only recently begun to receive new immigrants (e.g., those in the Midwest and the South), ELs are entering under-resourced and often highly segregated schools.

Second, there is substantial diversity among the EL population.Practitioners, policymakers and researchers face the critical task of distinguishing among the diverse social-emotional, linguistic, and pedagogical needs of the distinct adolescent English learner populations in U.S. schools. Increasingly, there is a push among researchers to distinguish between more recently arrived “newcomer” students and those who have attended U.S. schools since early childhood and remain classified as ELs, or “long-term” ELs. While there is agreement that these distinctions need to be further refined, we draw on the meeting’s discussion to use the following as our baseline definitions: newcomers are those who have been in the U.S. schools up to 4 years, while long-term or U.S. ELs are those with at least seven years of U.S. schooling and a lack of English proficiency.

Third, practitioners and decision makers in districts seek the awareness, preparation, and resources to address the diverse needs of ELs in the midst of accountability mandates. Okhee Lee, professor of education at New York University, pointed out that Common Core State Standards (2010) and the National Science Education Standards (1996) have offered new opportunities and challenges with teaching all students, including ELs, and especially newcomers and long-term ELs (Lee et al. 2013). Lee suggests that a crucial ingredient in improving practice is research that shows what using language looks like in teaching students particular subjects, as well as how argumentation and evidence converge or diverge across different subject and language areas. She suggests that EL and subject area educators collaborate to address language practices and subject-based practices with ELs, especially newcomers and long-term ELs.

The meeting then shifted to a set of discussions about moving the research agenda forward in the policies, programs, and practices that can improve ELs’ academic, socio-emotional and behavioral outcomes, and what decision makers and practitioners need to know to better serve this population. These included a focus on the kinds of assessment, data, and methods that could motivate new research lines of inquiry, or deepen existing ones.

First, improve standardized assessment evidence and make assessments more linguistically and culturally valid. Different opinions emerged during our discussion of assessment, ranging from classroom testing, to standardized tests and high-stakes testing. Some participants believed that we need to be wary of the use of assessments that have not been normed on English learners. Others thought that the dichotomy between valid versus invalid assessment was too simplistic. Still, others spoke of the increasing consideration of formative (low-stakes quizzes that assess students’ understanding of lessons as it evolves) rather than summative assessment (higher-stakes, cumulative evaluations typically given at an end point) in classrooms.

There was agreement, however, that in designing assessments it is essential to involve culturally and linguistically savvy informants in test design and test modification as well as to carefully assess the linguistic complexity of items. Further, it is important to parse out the intended use of assessment—for instance, content knowledge versus language knowledge. A number of currently favored approaches to assessment, such as constructed-response items and scenario-based assessments, may be challenging for ELs. Constructed-response items involve open-ended essay questions developed to demonstrate the student’s cognitive knowledge and reasoning. Scenario-based assessments draw on a text-heavy vignette and then a series of questions related to the vignette. However, the settings for scenario-based assessments may be outside of ELs’ experiences. Further, while their passive/receptive English skills may be sufficiently developed so they understand the question they are being posed (and might be able to express the answer in their native language), their expressive English skills may not have developed enough for them to be able to articulate their response in their new language.

Other topics discussed included considering what is an adequate period of exposure to the new language prior to language assessment; multiple modalities of assessment beyond tests, such as standardized portfolios, which include varied types of student work that can be used to chart the individual’s progress over time; and the potential limits and untapped potential of computer-assisted strategies. This animated conversation provided glimpses into the rich terrain for future areas of research.

Second, do more to leverage available large-scale data sets. There are at least two challenges with the existing data that can be used to answer questions about EL outcomes. First, it is very difficult to accurately categorize newcomer and long-term ELs. Second, states vary widely in how they designate EL students, and the means of collecting data on this population differ from district to district.

In our meeting, participants discussed the kinds of data we want to have collected. These include data about the child’s previous education in the country of origin, as well as the language used at home, including with whom and the frequency of use. Although there was agreement that it would be difficult to get high-quality, valid, or comparable data on these points, one particularly promising suggestion that arose from this set of conversations was the recommendation to add to the child component section of state databases. This section typically includes demographic data, such as student name, identification number, gender, and date of birth. These data are not collected every year—rather, they are retained in the record that accompanies the student’s entire educational journey. David Francis, a professor at the University of Houston, saw the possibility of adding basic English language information to this record. For instance, Was the student an EL when she or he started school? Was the student designated as eligible for language services, and what was his or her level of English proficiency when first taking the assessment? Such information added to the demographic data would afford researchers a longitudinal approach without having to link records over time, and give researchers better precision in discussing how well schools and school systems are serving or not serving this population—information that educational leaders would want to know. Further efforts to leverage big data sets might allow us to track progress across EL groups in ways we are currently unable to do with accuracy.

Third, include broader voices and methodological approaches when working with diverse populations. Per the recommendations of the American Psychological Association, research with culturally and linguistically diverse populations requires a fundamental alteration of classic investigative frameworks. Using multiple methods can help address these complex methodological challenges.

While well-designed large-scale quantitative studies can make it possible to generalize findings to particular populations, it can be difficult to interpret those results without qualitative research. Qualitative methods provide an opportunity to better understand local terms and cultural norms. And using triangulated data collected from a variety of perspectives and including a variety of strategies provides greater confidence that data are accurately capturing the phenomenon under consideration.

At the same time, bicultural and bilingual researchers are essential in various phases of research, including protocol development, recruitment, data collection, analysis, and interpretation. These individuals provide a fresh interpretive perspective and may lend specific disciplinary expertise. Interpretive communities of “insiders” and “outsiders,” as well as individuals representing a range of disciplinary expertise can assume critical roles in research projects.

Because measures developed for mainstream English-speaking populations can be culturally and linguistically biased, the field would benefit from the development of new tools for research with ELs. In addition, existing tools should be validated and where needed, they can be adapted. The process of development could be a dynamic inductive one involving theoretically-based formulations and themes emerging from the field.

Despite their differences, newcomers and English learners who have lived in the U.S. for some time are often socially isolated, and many have pressing educational and social-emotional needs. Practitioners, policymakers, and researchers should work together to address the challenges facing these young people—both the newcomers who are adjusting to their homeland and dealing with acculturative stress, and the long term ELs who may not be hearing or engaging in enough high-level conversational discourse to develop their skills. Whether research identifies the extra supports needed to develop skills or opportunities to participate in rich language use, there is more we can do to improve outcomes for these English learners.

Inclusion is at the forefront of this research agenda, both in terms of the education that English learners have and their exposure to peer and teacher relationships and in developing research questions that practitioners and policymakers want to have answered. This will mean doing a better job of both identifying what works, in what contexts, and for which learners, and also developing ongoing and sustained in-service professional development to support teachers.

Systematically considering programs, practices, and policies that may move the needle in some of these important areas is the next frontier of research if we want to address inequality for this fast growing group of students.

Also See:

Bardack, S. (2010). Common ELL terms and definitions. Washington DC: English Language Learner Center, American Institutes for Research.

Short, D.J. & Boyson, B. (2012). Helping newcomer students succeed in secondary schools and beyond. Washington DC: Center for Applied Linguistics.

Solorozano, R.W. (2008). High stakes assessment: Issues, implications, and remedies for English language learners. Review of Educational Research 78 (2): 260-329.

Lesaux, N.K. (2012). Reading and reading instruction for children from low-income and non-English speaking households. Future of Children Report 22 (2): 73-88.

Suárez-Orozco, C., Gaytán, F. X., Bang, H. J., O’Connor, E., Pakes, J., & Rhodes, J. (2010). Academic trajectories of newcomer immigrant youth. Developmental Psychology, 46(3) 602-618.

Related content

Subscribe for Updates