Dec. 9, 2013
During a recent visit to Chicago, I had an unsettling conversation with a kindergarten teacher. She explained that that recent changes to the district’s teacher evaluation system are shaping her pedagogical and instructional decisions in frustrating ways. Specifically, she worried that required third grade English literacy assessments make it difficult to justify research-based approaches to supporting her Spanish-speaking dual language learners (DLLs). While it’s easy to think of this as a unique problem, it illuminates broader implementation and alignment problems that serious education reformers should beware.
This teacher—let’s call her “Ms. Jones”—explained that Chicago Public Schools’ (CPS) new Recognizing Educators Advancing Chicago’s Students (REACH) system incorporates student academic growth in her evaluation each year. In other grades, student growth is measured on standardized assessments, but since kindergarten is an untested year, this element of her evaluation partially depends on schoolwide English literacy (standardized literacy assessment begins in third grade). (For more on the pros and cons of this accountability strategy, see Laura Bornfreund’s section on “Shared Attribution” in her recent policy paper, Oceans of Unknowns: Risks and Opportunities in Using Student Data to Evaluate PreK-3rd Grade Teachers.) Thus, though she teaches just one of her school’s many kindergarten classrooms, Jones is still incentivized to prioritize English acquisition above all other considerations for her DLL students.
Immersing students in a language other than their native tongue can undermine their broader linguistic development and negatively influence their long-term academic success.
Taken in isolation, we might not see anything wrong with emphasizing English acquisition for DLLs. English is the language of commerce and power in the United States—and globally. Every school system in the United States should strive to be sure that its graduates are proficient in English. Shouldn't we cheer efforts to immerse students in as much English as soon as possible?
Research suggests otherwise. In a 2005 “synthesis” of studies on DLL reading instruction, researchers Robert E. Slavin and Alan Cheung found that 12 out of 17 recent studies “found effects favoring bilingual education and 5 found no difference. None of the studies found results favoring English immersion.”
Fully immersing DLLs in English before they are developmentally ready can undermine their broader linguistic development and negatively influence their long-term academic success. In the early years, children learn both the content of their first language and how to use it. As they learn words’ definitions, the structure of verb tenses, etc, they are also learning how language (in the broadest sense) functions. They build an increasingly complex system of language concepts to fit their linguistic development. Programs that immerse young students in an unfamiliar language can interrupt this process and slow their linguistic and academic growth.
Thus, young language learners need support in continuing to develop their native language, even as they begin to learn English (the consensus on this, by the way, is the reason that many researchers prefer the term “dual language learner” to “English language learner”). This allows them to develop the complex language system they have been building in their native language even as they build a new, parallel system for English. In the long run, it’s not just better for DLLs’ language development—it’s better for their academic success as well.
The research is clear, the consensus is substantial—but the incentives aren’t so simple for DLLs’ teachers. Dual immersion programs that slowly phase in English over time usually take between five and seven years to bring students to full academic English proficiency. Jones doesn’t quite have that window. Her evaluations depend upon her students’ English proficiency in third grade. And while research shows that they will be better off in the long run, if she uses the dual immersion approach, many of the benefits of this approach would not be evident by third grade.
Linda Espinosa puts it clearly in a recent paper for the Foundation for Child Development, “It may take DLL children longer to respond to language tasks that require word retrieval, and they may not know as many words in each language.”
These are, of course, precisely the sorts of skills that an assessment measuring English literacy (in this case the NWEA MAP) would require.
Espinosa notes that dual immersion approaches offer other benefits, even in the short-term, such as “increased cognitive control and executive function skills.” But these are harder to measure, and the third grade test is focused on English literacy.
As a result, Ms. Jones teaches her DLL students in English. While this may not be best for them in the long run, she reasons, it may be the best she can offer them to prepare for the MAP in third grade.
Two other factors make her choice of instructional language (and approach) even more difficult.
First of all, the rest of the student growth portion of her evaluation depends on a performance task. At the beginning of the year, she chooses a task and has her students attempt it. She offers it again at the end of the year so that they can demonstrate progress.
Sound simple? Here’s the challenging part: Jones must also select the language in which she will administer the task. If she chooses Spanish (the native language of 100% of her DLLs), her students will have the benefit of completing the task in the language with which they are most familiar—even by the end of kindergarten. If she chooses English, they’ll have the benefit of completing the task in the language she will be teaching them that year.
Which choice would best measure the students’ abilities? Which would best highlight her “value-add” as a teacher?
The second factor is perhaps even more consequential. Jones is one of only a few bilingual teachers at her school. What’s more, she’s one of an even smaller group that has formal training in teaching a dual immersion program. In other words, whatever she does in her classroom this year, it’s unlikely that her students will continue to be supported in Spanish. If she offers a dual immersion program, she will be pedagogically misaligned with her students’ first (and second, and third, etc) grade teachers.
Unlike many teachers of DLLs, Jones hasn’t made this decision because of her own limits. She is well aware of the growing research consensus around how to best serve DLLs. She is capable of teaching a dual immersion program and knows that it would probably be best for the students, but her incentives around this choice aren't just misaligned—they’re chaotic.
Karen Nemeth, an expert on DLLs, notes that the problem isn't confined to Chicago. "It's a big problem and everybody knows it. Every researcher knows that assessment has to assess these kids in multiple languages. Otherwise you're not assessing literacy, you're just assessing English literacy."
The larger lesson here is that it’s much easier to design reform initiatives than to make them work. While there are good reasons to rework teacher evaluation systems and include DLLs in standardized assessments, it’s important to ensure that these reforms actually live up to their billing in practice. And nothing undercuts reform efforts like sloppy implementation of well-intended initiatives, no matter how necessary they were in the first place."