A change to how a widely used English-proficiency test is scored has led to thousands of students being retained in English-language-learner classes and created budgeting and staffing challenges for some school districts.
The change has rattled educators in states with an established English-learner student body, such as Nevada; those with fast-growing populations, such as Tennessee; and even states, such as Maine, that have a relatively small percentage of students who don鈥檛 yet communicate fluently in English.
Those three states and more than 30 others belong to the WIDA, or the World-Class Instructional Design and Assessment consortium, and use a test it designed to assess students鈥� English proficiency.
The scoring adjustment to the test known as ACCESS 2.0 raises the bar for English-language proficiency and took effect in the 2016-17 school year. With the change, English-learners must demonstrate more sophisticated language skills in four domains鈥攍istening, speaking, reading, and writing鈥攖o achieve the same proficiency-level scores.
Student scores in many states dipped under the more demanding standards this past spring, extending the time that many will need to remain in support services, such as English-as-a-second-language classes.
In Nevada, state administrators estimate that only 2 percent, or 1,500, of the state鈥檚 75,000-plus English-learners will move out of all language-development courses this year. In past years, between 12 percent and 14 percent of students moved out.
鈥淲e anticipated that there would be a drop,鈥� said Karl Wilson, Nevada鈥檚 program supervisor for ELLs. 鈥淲e were surprised that the drop was so significant.鈥�
Figuring out whether students need that support and when they no longer require it is crucial to the success of English-learners. When moving students out of language-support services鈥攃ommonly called reclassification鈥攐ccurs too early, English-learners can find themselves struggling.
If the process happens too late, students may be restricted from taking higher-level courses that would prepare them for college.
The scoring change鈥攔olled out in the spring to reflect the new language demands in the more-rigorous academic standards adopted by states in recent years鈥攊s the first for WIDA in more than a decade.
鈥淲hat鈥檚 being asked of students, not just English-learners, is categorically different than it was in the past,鈥� said H. Gary Cook, the director of research for the WIDA consortium, a group of education agencies that share English-language-proficiency standards and assessment for ELLs that are aligned with the Common Core State Standards. 鈥淚f the education system doesn鈥檛 change accordingly, I don鈥檛 foresee progress.鈥�
ELL Performance Is Critical
The consternation over the test scores comes as schools brace for another monumental change.
This coming school year, the performance of English-language learners will have a much greater impact on how schools are judged under the federal Every Student Succeeds Act.
It is another signal that the significance of English-learners鈥攚ho now make up nearly 10 percent of K-12 enrollment鈥攊s on the rise as this population grows and the changes in federal law mandate a more precise measurement of how those students are performing in school.
鈥淪tates are paying a tremendous amount of attention to this issue and really working hard to figure out how to get this 鈥� right,鈥� said Scott Norton, the director of strategic initiatives in standards, accountability, and assessment for the Council of Chief State School Officers.
For WIDA-member states, getting it right includes setting the cutoff scores that students earn to transition out of English-learner services. States have the discretion to set the cut scores that students earn to transition out of English-learner services.
Tennessee plans to temporarily adjust its English-learner exit criteria to avoid 鈥渦nnecessarily鈥� retaining students in ESL classes.
鈥淎lmost all of our state鈥檚 English-learners would be retained in ESL programming because they didn鈥檛 meet our state鈥檚 exit criteria,鈥� said Sara Gast, a spokeswoman for the Tennessee education agency.
To set its permanent criteria, the state plans to study the relationship between student scores on the ACCESS 2.0 exams and state tests in other subjects. Maine is among the states that have already lowered their exit criteria in response to the ACCESS cutoff-score changes.
The shock that states are experiencing is similar to the adjustments made when adopting the common-core-aligned assessments known as the Partnership for Assessment of Readiness for College and Careers and Smarter Balanced.
鈥淢oving from one test to another and raising standards 鈥� is not really a new thing,鈥� Norton said. 鈥淚t takes a while for everybody to adjust, both on the instructional side and on the student-expectation side. And it usually causes some concern out in the districts until people adjust.鈥�
Count Ignacio Ruiz among the concerned.
Ruiz, the assistant superintendent for ELLs in the Clark County, Nev., schools, expects the change to be significant for his district鈥檚 61,000 English-learners.
After the 2015-16 school year, roughly 16 percent tested high enough to exit English-support services.
That percentage dropped to less than 8 percent in the latest round of testing. That means about 5,400 fewer students tested out of support services.
A More Accurate Test?
Similar scenarios have played out in districts around the country.
From Baltimore to Boise, Idaho, fewer students have transitioned out of English-support classes. With fewer students moving out of English-language-development classes, educators face the prospect of educating hundreds, if not thousands, more ELLs than district leaders budgeted and planned for.
That鈥檚 most likely a result of more states, including WIDA-member districts, leaning almost exclusively on English-language-proficiency exams to determine when students no longer need extra English support.
鈥淲hile this test is more rigorous, we think in the long run it鈥檚 a more accurate test,鈥� said Stacey Roth, Boise鈥檚 administrator of student programs. 鈥淲e want to ensure that our kids truly have reached proficiency.鈥�
But sometimes that鈥檚 not even enough to determine when students are ready to move on. Research from the federal Regional Education Laboratory at WestEd has found that middle- and high-school-age English-learners who are deemed proficient still often struggle to grasp academic concepts.
In acknowledgment of that research and other data, ESSA requires states to monitor former English-learners for four years, two years longer than what was previously required.
鈥淚t鈥檚 not just about the proficiency piece, but that we have the right information to get students to proficiency,鈥� Ruiz said. 鈥淲e want to make sure that when we say a student is exiting, that will translate to them being able to have success in those high-stakes tests.鈥�
While Ruiz supports the tougher standards, he鈥檚 worried about students, including long-term ELLs鈥攖hose not considered proficient after years in U.S. schools鈥攚ho may become frustrated at the prospect of another year in support services.
In Boise, educators saw fewer elementary students reclassified as proficient, even those who are clearly bilingual, because their writing skills, like those of almost all children in that age group, are still developing.
Educators and advocates remain worried about other potential pitfalls of relying on a single assessment.
A report written for CCSSO concludes that states should use at least two measures, including observations of how students use language in classroom settings, to determine whether a student no longer needs English-language instruction.
Prior to ESSA, most states had a mishmash of ELL exit procedures, with districts often deciding when the students no longer needed support. With the new law鈥檚 aims to make policies consistent, fewer states may rely on teacher input or evaluation鈥攎aking tests such as ACCESS 2.0 the only way out of English-support classes.
鈥淪tates really need to pay attention to these students. Many are trying,鈥� said Cook, the WIDA research director. 鈥淯ltimately, we rely on our member states to support the decisionmaking with the data.鈥�
