|
Joseph Robinson
University of Illinois at Urbana-Champaign
Effects of native-language assessment accommodations in mathematics for kindergarten and first-grade English learners
FINAL REPORT:
Using a nationally representative sample of kindergarteners and first-graders, this study asks if native-language test translations help English language learners (ELLs) demonstrate their mathematics skills. ELLs frequently encounter testing hurdles because they are not proficient in English, the predominant language of assessments. Low scores on these assessments can limit ELLs’ academic opportunities through inappropriate track placements and decreased confidence. Intuitively, translations should help ELLs by making the content of the test accessible. Yet, actual research on the effects of mathematics translations is exceedingly scarce. Moreover, related research—e.g., research claiming that ELLs’ native-language skills are also underdeveloped—suggests translations may be ineffective.
This study uses a rigorous quasi-experimental design (regression discontinuity design) and finds that Spanish-speaking ELLs perform significantly better on mathematics assessments when tested in Spanish (instead of English) in kindergarten and first grade (Cohen’s ds > .85). In addition, and in contrast to prior research on older students, there was no evidence that the language of instruction moderates the effects of native-language test translations, suggesting that the home language may be more important than previously believed for testing policies in early grades.
Despite the implications for individual students and a variety of other stakeholders (schools, teachers, and education researchers), only nine states permit translations on annual state-mandated exams. Thus, policymakers should consider that test translations may provide ELLs opportunities to demonstrate their mathematics knowledge and consider adding translations to the list of available accommodations. Moreover, accommodations that improve test construct validity for the fast-growing ELL population are vitally important for teachers under value-added performance evaluations and for valid interpretations and uses of test results by education researchers and practitioners.
Back to Funded Research Grants Page
|