Achievement of ELLs Has Much to Do With the School
The Pew Hispanic Center today released a report by Richard Fry, a senior research associate at the center, showing that schools that report low achievement for ELLs also tend to have a set of characteristics associated with poor student performance on tests. Those characteristics include high student-teacher ratios, large student enrollments, and high levels of students eligible for free- and reduced-price lunches. The report says that when ELLs aren't isolated in such schools, they do considerably better on standardized tests. See my story published today at edweek.org, "Schools With Poor ELL Scores May Share Common Elements."
The report is consistent with the work of other researchers in showing that a high concentration of ELLs in schools is associated with low standardized test scores. But what I find interesting about Mr. Fry's report is that he uses a very small number of students as a threshold to examine the presence of ELLs, non-Hispanic blacks, and non-Hispanic whites in schools. He uses the thresholds that states set for schools to report test scores for student groups under the No Child Left Behind Act. In Arizona and Florida, for example, that number is 10 test-takers per grade for ELLs, whites, or other groups; in New York and Texas, the threshold is at least 5 test-takers per grade.
Essentially, Mr. Fry found out that ELLs that go to school with even a few white students do better on state math tests than students who have virtually no white students in their schools. And white and black students who go to school with even a few ELLs do worse on math tests than those who go to school with practically no ELLs.
Pew Hispanic Center researchers don't really speculate on what's behind their findings or make policy recommendations. Do any of you want to take a stab at making recommendations based on this information?