June 26, 2008

The Role of Schools in the English Language Learner Achievement Gap

V. The Consequences of ELL School Concentration for Math Achievement

The concentration of ELL students in schools that report ELL test scores is positively associated with their lagging performance on mathematics achievement tests. Although white test-takers and ELL test-takers are largely not educated in the same public schools, there are public schools that educated both white and ELL students.1 To gauge how much of the ELL achievement gap is due to white students and ELL students attending different schools, this section of the report measures the difference in math proficiency, based on the state assessment tests, between ELL students and white students who attend the same schools. Figure 4 illustrates the difference in math proficiency rates between ELL and white students among the subset of public schools that educate both ELL and white students.2

When ELL students and white students attend the same schools, the measured difference in proficiency rates shrinks considerably (Figure 5). For example, statewide in Arizona 84% of white third-grade test-takers passed the state standard, compared with 49% of ELL third-grade test-takers, for an aggregate difference of 35 percentage points. However, if we examine the subset of Arizona elementary schools that educate both white and ELL students, ELL test-takers trailed their white classmates by only 27 percentage points. In each state, and in both elementary grades and middle school grades, a significant portion of the aggregate difference in math proficiency can be accounted for by the fact that ELL students and white students tend not to attend the same public schools.

Similarly, difference in math proficiency between black and ELL students may be due in part to their attending different schools. Figure 6 shows the aggregate statewide difference in ELL and black proficiency rates and the difference calculated on the basis of ELL test-takers and black test-takers who attend the same schools. Again, with the exception of Florida, the difference in math proficiency rates shrinks when examining students who attend the same public schools.

  1. numoffset=”8″ As mentioned in Footnote 6, ELL students can be of any racial/ethnic origin and some ELL students were also non-Hispanic white students. The vast preponderance of non-Hispanic white students were not English language learner students.
  2. The difference for each public school is obtained by subtracting the ELL proficiency rate from the white proficiency rate. However, the ELL proficiency rate, by definition, is available only for ELL reporting schools. And the white proficiency rate is available (among ELL reporting schools) only for ELL reporting schools educating enough white students to report the white proficiency rate. As a result, Figure 4 shows the average difference in proficiency among the subset of ELL reporting schools that educate at least a minimum threshold of white students. Appendix A reports on the overlaps of the different types of public schools in the NLSLSASD.