New Research: DLLs’ Language Acquisition Patterns are Anything But Simple

Blog Post
Oct. 12, 2015

Education policy is fertile turf for jargon. And that poses a real problem for efforts to improve policies for underserved kids—because it can be hard to cut through all the terminological jumbles to explain to non-educators what those policies actually do.

Take “reclassification policy,” for instance. “Reclassification” sounds like the result of a particularly good haircut and wardrobe update. Re-class-ify must be what happens when enough wealthy people move into a neighborhood and start getting snooty about the not-wealthy people in the area. Or maybe it’s what happens when the Army decides you’re not quite used up yet and can still serve in the trenches. On still another hand, it sounds a little like CIA retrieval of covert documents from a snooping newspaper. Shoot, it sounds like something that cricket teams do between “overs.”[1]

Ok, ok. I’m done.

As it turns out, “reclassification” refers to a deeply important moment in DLLs’ educations. Reclassification is the process that school districts, states, and the federal government[2] use to determine which multilingual students are still learning English — and which have reached full academic proficiency.[3] As I wrote in Chaos for DLLs about a year ago, reclassification policies are: 

extraordinarily important and notoriously challenging to set. Since they determine which students remain formally designated as DLLs, they also determine who will receive targeted language supports, accommodations on assessments, content instruction alongside native English-speaking peers, and much more.

When these students are deemed fully proficient in English, language services generally evaporate. So knowing how long it takes DLLs to be reclassified is a matter of considerable pedagogical and budgetary importance. And while this might seem like a relatively straightforward area of policy, research suggests otherwise.

This is why a new study from Oregon State University Professor Karen D. Thompson is so important for considering how to think productively about fixing reclassification policies in U.S. schools. In it, Thompson studies DLL reclassification patterns from the Los Angeles Unified School District (LAUSD) over nine years (between the 2001–02 and 2009–10 school years).

LAUSD DLLs were most likely to be reclassified out of DLL status in late elementary school—usually after approximately six years in the district’s schools. This isn’t so bad: it approximately tracks the research, which suggests that DLLs take 4–7 years to reach full academic English language proficiency. But some results were much more discouraging: one in four LAUSD DLLs remained in language services beyond nine years.

These findings need some context. Because of a 1998 state referendum, almost all DLLs in California receive their language support services in “Structured English Immersion” (SEI) programs. Recent research suggests that these English-only services are unlikely to support DLLs’ academic success in the long run. While it might benefit DLL students to remain in bilingual settings for longer periods of time, it may not be as beneficial to continue giving DLLs SEI services if they are not showing progress. Students who aren't succeeding in these SEI settings after nine years are unlikely to suddenly meet reclassification standards in their tenth year of language services.

And yet, this focus on topline numbers is probably a mistake. As tempting as it is to focus on the big question of how long it takes DLLs to reach reclassification, Thompson provides a useful caution. She found that the answer varies with a host of factors:

Boys, native Spanish speakers, students with lower levels of initial academic English proficiency, students with lower levels of initial academic [native language] proficiency, students in special education, and students whose parents have lower levels of education all have lower probabilities of reclassification than their peers, controlling for other factors…For example, estimates suggest that students who enter the district with beginning levels of initial academic [native language] proficiency and beginning levels of academic English proficiency are 24% less likely than their peers who enter with high levels of both.

Thompson’s research suggests that California—and other states—should be wary of requirements that appear to make reclassification particularly difficult for DLLs who may have met all other metrics for leaving DLL status. If a student scores proficient on the state’s English language proficiency assessment, but fail to demonstrate proficiency on the state math assessment, should they remain a DLL for another year?

This is particularly important in California, where districts have considerable latitude in setting their reclassification policies. This flexibility could be helpful for DLLs, if it meant giving local educators flexibility (and guidance) about how to think through reclassification for individual students. But that is not how California's decentralization works in this case. Instead, districts are required to consider a variety of factors and set a variety of benchmarks for success within each. Districts that set particularly stringent criteria will have different reclassification patterns than districts that set less stringent criteria. In particular, DLLs who miss what Thompson terms the late-elementary school “reclassification window” may be at particular risk of stagnating academically. That is, their ongoing language services may limit their opportunities to fully participate in the academically rigorous coursework available to students who are not receiving these services. Given doubts about the efficacy of California’s English-Only language services for DLLs, variations in reclassification patterns can be deeply consequential for these students.

So while reclassification policy isn’t as prominent or provocative as, say, fights over inequity in neighborhood schools, it’s extremely important. And, as Thompson notes, current “accountability targets for how long it should take [DLLs] to attain proficiency in English have too often been based on wishful thinking rather than empirical data.” This is a useful reminder in a moment when policymakers are reconsidering these very accountability provisions—and even whether to have any accountability at all.


This post is part of New America’s Dual Language Learners National Work Group. Click here for more information on this team’s work. To subscribe to the biweekly newsletter, click here, enter your contact information, and select “Education Policy.”

[1] (Or test matches. Or wickets. Or whatever it is they call it when they’re swapping out which all-rounder will be in the popping crease out of his ground. God, cricket is confusing). As in, “Right, Neville, let’s get another round of cucumber sandwiches after the forthcoming reclassification! Also some Pimm’s!”

[2] No Child Left Behind, the main federal law governing PreK–12 education, requires states to develop language assessments and appropriate language proficiency levels to flesh out that process.

[3] Note: I’m aware that there are enormous problems with this sentence. All students are still learning English at all points in their educational career. Even 30-something education policy writers are still learning English. And the definition of “full proficiency” could be — and frankly, is — interpreted to mean any number of things from basic oral conversation to full academic literacy and so on and so forth. This is precisely the point I’m after."