Classifying English proficiency varies by district, with mixed outcomes for students

The findings, published in a special centennial issue of the American Educational Research Journal, inform conversations about statewide policies for English learner reclassification, which are now mandated under the 2015 Every Student Succeeds Act.

"Our study examined the effect of state-level decisions as implemented at a local level. With the Every Student Succeeds Act stipulating that reclassification policies be standardized within a state, our work gives a glimpse into what we might see more broadly once these policies are standardized," said Joseph Robinson Cimpian, associate professor of economics and education policy at NYU Steinhardt and the study's lead author.

"Notably, our findings revealed a wide array of effects of reclassification on achievement and graduation, ranging from large negative effects in some districts to large positive effects in others, even when considering students subject to the same state-level policies."

By 2050, over a third of school-age children in the U.S. are projected to be immigrants or the children of immigrants, and many of these children will be classified as English learners. Effectively educating the large English learner population requires appropriate instructional services and settings throughout the time students are learning English, as well as during their transition to fluent English status. Because reclassification often entails a change in services and settings, it is important that the switch occurs at the appropriate time: when the student no longer receives added benefit from the English learner setting.

Reclassification criteria vary across states, districts within a state, and even within districts. However, they all use some determination of whether an English learner is achieving at a pre-specified level on an assessment determined by local policymakers.

A recent wave of education research uses a technique known as regression discontinuity designs, which use large amounts of data to compare the outcomes of students who just barely attained the reclassification criteria with those who just barely failed to attain it, helping researchers to understand the effects of reclassification on students with very similar profiles.

While previous studies have examined effects within a single school district, the current study is the first to examine reclassification effects across multiple districts and measure variability between them. The researchers used longitudinal data from 107,549 students in two states -- one in the Southeast and one in the Northwest -- and applied regression discontinuity designs to assess the effects of reclassification on later achievement and graduation. All students studied were at some point considered English learners and were enrolled in elementary, middle, or high school when they were reclassified.

The researchers found remarkable variability between districts in how reclassification affected student achievement and graduation rates. In some districts, reclassification had negative effects on graduation, where students who were reclassified but just barely made the cut-off were as much as 80 percent less likely to graduate. Meanwhile, reclassification in other districts had positive effects on graduation. Just-barely reclassified students were 38 percentage points more likely to graduate than peers who stayed in an English learner setting.

"This variability tells us that we should not default to a belief that reclassification is universally beneficial or detrimental," said Cimpian.

The researchers conclude that this type of research can help policymakers see misalignment between their thresholds for reclassification and the instruction and services in their state and districts. Fixing this misalignment might include lowering or raising the reclassification threshold as well as modifying instruction and services for students near the threshold. For example, if reclassification was found to have a negative effect on student outcomes, a district might consider providing additional language development support for students right after reclassification.

Of note, the passing of the Every Student Succeeds Act requires states to establish uniform reclassification criteria for all districts, meaning districts may not have the freedom to adjust their thresholds or add in additional considerations for reclassification.

"On one hand, establishing the same criteria has the benefit of facilitating comparisons across districts and providing a common metric by which to assess English learner status for students who move across district boundaries. On the other hand, requiring a common threshold across the state restricts the ability of a district to adjust the threshold to meet the needs of its own students given the services that the district provides," Cimpian said.

Commenting on the importance of this research in the American Educational Research Journal, Kenji Hakuta of Stanford University said it is a "wonderful display of the power of combining current analytic methods, present-day educational datasets, practitioner-researcher collaborative arrangements, and key policy questions pertaining to the educational outcomes of English learners.

"This paper's important message is that policies and implementation matter at the state and local levels in ways that demonstrably affect the probability of graduation for individual students," Hakuta continued.