Two studies find mixed results on grouping English learners ^{CEFR B1}

Two recent studies asked whether putting many English learners (ELs) together in the same classes helps their academic outcomes. The high school study, published in the Peabody Journal of Education, analyzed New York City Public School records for 31,303 ninth graders in 2013, 2014 and 2015. Researchers divided students by years learning English before ninth grade: newcomer EL (1–3 years), developing EL (4–6 years) and long-term EL (7 or more years). They measured the percent of ELs in students' courses and then compared graduation and college enrollment across schools with EL shares of 50% or more and those with less than 50%.

The team found that higher concentrations of ELs in courses were associated with lower chances of graduating and lower college enrollment, regardless of prior years learning English or the overall school share. Reported effects included 10–15% lower four-year graduation, 6–11% lower six-year graduation, 12–13% lower immediate college enrollment and 9–13% lower enrollment within 2.5 years.

The elementary randomized trial assigned 84 fourth- and fifth-grade ELs to small groups that were either all ELs or mixed with non-EL peers. All groups had the same 12-week language and literacy curriculum from the same teachers. On average, researchers found no statistically significant differences in language, reading comprehension or argumentative writing, though heterogeneous groups helped higher-skill students and homogeneous groups helped earlier-stage learners. Observations showed different teacher practices in each group type. Both studies conclude grouping ELs together should not be a default practice.