ELL Assessment: One Size Does Not Fit All
¿Qué es lo que hace que una rama del gobierno no se vuelva demasiado poderosa?
If you know the answer to this question, congratulations! If you don’t, why not? Is it because you don’t understand the checks and balances built into the U.S. system of government, or because you don’t understand Spanish?
Every year, some 5 million public school students who are still mastering English take assessments to determine how much they know, how much progress they’ve made, and where they need support.
Unfortunately, the results of these tests are far from valid because many of these students are not sufficiently proficient in English to demonstrate their knowledge and abilities on assessments designed for native English-speakers. It is akin to asking someone to fill out a job application in a language he or she doesn’t understand even though that person can potentially do every aspect of the job.
Just as the prospective employer wants to know what the candidate can actually do, we need accurate information about English-language learners’ knowledge and skills. Unfortunately, the assessments that we administer in schools fail to fully uncover what we need to know.
This problem affects one of the fastest-growing groups of students in our nation’s public schools. English-language learners, or ELLs, are nearly 11 percent of the K-12 population, and about 80 percent of these students speak Spanish, with the rest speaking a wide variety of other languages. In California, more than half the children now entering public schools come from households where the first language is not English. Further, ELLs in the United States are not a monolithic group—they vary widely in proficiency in both their primary languages and in English.
The Partnership for Assessment of Readiness for College and Careers Consortium, or PARCC, and the SMARTER Balanced Assessment Consortium, or SBAC, are the two U.S. Department of Education-funded consortia charged with developing a new generation of state assessments aligned to the common-core state standards. Developing fair and accurate assessments for English-language learners must be a focus of their work from the very beginning as well as at every stage in the development of these new assessment systems. This must be an integral, rather than peripheral, part of any assessment system that’s expected to serve all students well.
The federal government has also organized a funding competition for consortia of states to develop assessments of ELL students’ English-language proficiency that better reflect the language demands of the common core. This is another critical problem, closely related to and yet different from the problem of fairly and accurately assessing ELLs’ academic performance. At a minimum, the work of the consortia should be closely integrated with the work of PARCC and SBAC to ensure that the assessments administered to English-language learners take accurate account of their developing fluency in English.
ELLs confront great challenges when they start school, working simultaneously to master their second language and to learn grade-level subject matter in reading, language arts, math, science, and other subjects. English-language proficiency is foundational to academic success and critical to measured performance. It affects a student’s ability to learn academic content taught in English and to demonstrate skills through assessments conducted in English.
Underperforming on tests because of a lack of language fluency can unfairly depress students’ scores. If they perform poorly, we must determine why. Is it due to a lack of content knowledge or a lack of English proficiency? Is the assessment responsive to differences in students’ levels of fluency? Today, we fail to disentangle these issues far too often, with grave consequences for students.
But the problems with assessment go further because of the way the language-learner subgroup is constituted for accountability purposes. The most linguistically and academically accomplished students exit the English-language-learner category over time, as they become fluent in English. Those not making sufficient progress remain in the category, where they are joined by newly entering ELLs who are by definition at lower levels of language proficiency. State-level assessment results typically ignore these revolving-door practices, which wrongly stigmatize the language-learner subgroup, demoralize students and teachers, and prevent accurate reporting of long-term outcomes, including graduation rates and college access and success data. What’s more, states have little incentive to track the long-term performance of English-learners who are reclassified as fluent in English. After two years of required monitoring, these students are no longer counted in ELL statistics. So there’s no reward if reclassified learners do well, and no accountability if they don’t.
Policymakers should take steps now to develop assessments and accountability practices that are more accurate and fair. The federal Education Department can require and ensure close collaboration among federally funded academic and English-language proficiency consortia.
A 2011 report from my organization, Policy Analysis for California Education, and the Rennie Center for Education Research and Policy makes clear that new technologies, including computer-adaptive assessment and accommodated virtual performance assessments, can improve access and support the implementation of fairer and more accurate assessments. PARCC and SBAC can also invest in formative-assessment processes and practices to support better instructional strategies for all teachers of English-learners. When Congress reauthorizes the Elementary and Secondary Education Act, lawmakers should set clear expectations for improving English-language proficiency and the performance of all students who have ever been categorized as English-language learners, with agreement to monitor these former language-learners for as long as they remain in the system.
The new assessment programs being developed by PARCC and SBAC must be more responsive to both the needs and strengths of English-learners. We owe it to students to ensure that assessments help them develop to their fullest potential, rather than leave them frustrated by low and inaccurate results. After all, if you can’t understand the test, how can you possibly pass it?
Vol. 31, Issue 02, Pages 20-21Published in Print: August 31, 2011, as Assessing English-Language Learners: One Size Does Not Fit All