June 6, 2007

How Far Behind in Math and Reading are English Language Learners?

III. National ELL Achievement Gaps

Measuring the Gap

The NAEP, or the “Nation’s Report Card,” is the best-known assessment of student learning for the U.S. as a whole. NAEP assesses student learning in mathematics and reading at grades 4 and 8, providing national level results as well as results for some states.

For the nation as a whole, NAEP reveals that ELL students were far behind white students in their mathematics and reading skills in 2005. Performance on the main NAEP is reported in terms of four achievement levels: below basic, basic, proficient, and advanced. Since relatively few students from any NAEP student group perform at the advanced level—and ELL students nationally tend to be concentrated at the lower achievement levels—this report presents the NAEP achievement gap in terms of performance at or above the basic level of achievement. The National Center for Education Statistics also reports NAEP results in this fashion (NCES, 2005).The basic achievement level identifies “partial mastery of prerequisite knowledge and skills that are fundamental for proficient work.”

The 2005 assessment indicated that 46% of ELL students nation-wide achieved at the below basic level in math in grade 4 (Table 1). In reading 73% of ELL fourth grade test-takers were below basic. Among white fourth-graders nationally, 11% were at the below basic level in math and 25% were below basic in reading.

In the grade 4 math assessment 46% of ELL students performed at the below basic level and 54% performed at or above the basic achievement level. Among white fourth graders, 89% were at or above the basic achievement level in math.

This report assesses the gap in achievement as the difference in the percentages at or above the basic level for ELL students and a comparison group. For example, the gap in grade 4 math achievement between white and ELL student is 35 percentage points (89% for whites versus 54% for ELL students) (Figure 1). In conceptual terms, the 35 point gap is how far the ELL student group as a whole lags behind in demonstrating at least “partial mastery of prerequisite” skills.

Compliance with NCLB mandates will be determined not by performance on the NAEP but rather by testing programs developed and administered separately by each of the states. However, measuring the achievement gaps in the NAEP is a way of illustrating how much ground needs to be covered to accomplish the goal of having students of all groups meet the same standards of minimum proficiency.

In the 2005 NAEP, English language learner students significantly trailed black students in math and reading skills at the national level, although the national achievement gaps between ELL and black students were not as large as the gap between ELL and white students. For example, in grade 4 math, 60% of black students performed at or above the basic level. The ELL to black math achievement gap for grade 4 was 6 percentage points (Figure 2).

Yet, a majority of Hispanic students were not also English language learners.1 And some English language learner students were of Asian or Pacific Islander racial origin and not Hispanic.

In terms of learning, the 2005 NAEP indicates that ELL students trailed behind Hispanic students in their math and reading abilities (Figure 3). For example, in grade 4 math 67% of Hispanic students performed at or above the basic level, so ELL fourth graders trailed 13 percentage points behind their Hispanic peers.

National Assessment of Educational Progress

The National Assessment of Educational Progress (NAEP) is the nation’s only nationally representative assessment of educational achievement. Begun in 1969, NAEP is conducted by the National Center for Education Statistics under the direction of the National Assessment Governing Board.

In education circles, NAEP is often referred to as the gold standard of educational assessments. States can elect to participate in the state NAEP. The state assessment is identical to the national assessment in content. In 2005 all states and the District of Columbia participated in the grade 4 and 8 math and reading assessment, but state-level results for ELL students are not available for all states.

In 2005 a large sample of about 172,000 fourth-graders and 162,000 eighth graders participated in NAEP nationwide. NAEP does not provide scores for individual students or schools. Achievement is measured for students by grade and subgroups within those grades.

This report focuses on the reading and math abilities of students, but the NAEP has also assessed abilities in science, writing, U.S. history, civics, geography, and the arts. NAEP results are available at the NAEP Data Explorer: http://nces.ed.gov/nationsreportcard/nde/

Widening Gap, Changing Population Between 4th and 8th Grades

The ELL achievement gap widens at higher grades. For example, in fourth grade math, ELL students were 35 points behind white fourth graders. In grade 8, ELL students were 50 points behind white eighth graders (Figure 1). The widening of the ELL to white gap at higher grades is not unique to the 2005 NAEP assessment. It is also apparent in assessments the states are required to administer under No Child Left Behind (see the Appendix). In California, for example, student achievement results on the Stanford Achievement Test demonstrated large achievement gaps that increased at the higher grades (Gandara, et. al., 2003).

Is this widening gap from 4th to 8th grades evidence of failure on the part of the schools and the students, or are there other factors to consider? Indeed, change in the composition of the ELL population across these grades appears to explain some of the difference: Higher achieving students are removed from the ELL population while newly arrived immigrants just starting out in U.S. schools are added to it. These factors, explored below, help explain why ELL students fall further behind white students from grade 4 to grade 8. But these changes in composition do not diminish the challenges faced by students and schools in attempting to close the gap as mandated by federal policy.

ELL status is not permanent. Between 4th and 8th grade some students succeed in learning English. They are reclassified and no longer counted as English language learners. Meanwhile, because of immigration, new foreign born English language learners are added to the ELL population after 4th grade.

The U.S. Department of Education’s administrative data on English language learners (collected in the Common Core of Data) has little information on these students other than their school district. Using Census data, however, the characteristics of limited English speaking students can be examined. Limited English ability in the Census only refers to speaking abilities. ELL status depends on reading and writing abilities, in addition to speaking abilities, as well as other test scores, grades and teacher input (Jepsen and de Alth, 2005). The limited English population is frequently used as a proxy for the ELL population (Capps, et.al 2005).

Table 2 reports on public school enrollment in 2001 in grades 1 to 4. By 2005, most of these students had been promoted to grades 5 to 8. The number of limited English speakers enrolled in public schools clearly decreases from elementary school to middle school. There were 941,000 limited English speaking students in grades 1 to 4 in 2001. By 2005 there were 661,000 limited English speakers in grades 5 to 8. In addition to showing the decline in the number of limited English speaking students, Table 2 reveals the change in the composition of the limited English speaking students.

The number of native-born limited English speaking students declined from 656,000 in 2001 to 369,000 in 2005, or 44%. The share of limited English speaking students who were native-born fell from 70% in 2001 to 56% in 2005.

Although the number of foreign-born limited English speaking students remained roughly unchanged at 290,000 from elementary school to middle school, many foreign-born, limited-English speaking students in grades 1 to 4 did learn to speak English by 2005. Unlike native-born students, the total number of foreign-born students increased from 608,000 students in 2001 to 865,000 in 2005 due to immigration.

Many of the new arrivals were limited English speaking students. Of the 292,000 foreign-born, limited-English speakers in grades 5 to 8 in 2005, only 161,000 had been in the U.S. at least 4 years earlier and thus could have been in the U.S. grade 1 to 4 cohort in 2001. It appears that about 125,000 of the 286,000 foreign-born, limited-English speakers in grades 1 to 4 in 2001 learned to speak English by grades 5 to 8, a decline of 44%. Thus, foreign born limited English speakers in elementary schools appear to learn English at the same rate as native-born limited English speakers. However, those students were replaced in grades 5 to 8 by 131,000 foreign-born students who arrived less than four years ago and were not enrolled in grades 1 to 4 in the U.S.

Consequently, the middle school ELL population is composed of two student groups: newly-arrived, foreign-born students who were not in U.S. schools as well as ELL students from elementary school who have not mastered English. It is likely that the acquisition of English language skills and academic achievement are highly related. Those elementary school students who learned English rapidly also tended to score higher on their math and reading assessments. These students departed the ELL population by middle school and their higher achievement is no longer reflected in middle school achievement gap. The ELL to white achievement gap widens from elementary school to middle school possibly because the highest achieving ELL students in elementary school have departed by middle school.

The widening of the ELL achievement gap from grade 4 to grade 8 is distinctive in reading. In the national NAEP, the black-to-white achievement gap and the Hispanic-to-white achievement gap widens in math from grade 4 to grade 8. But those gaps tend to diminish in reading from grade 4 to grade 8.2 In the national NAEP, reading gaps narrow as the grades progress for black and Hispanic students. But not for ELL students. The English language learner population may be unique in featuring widening reading gaps between elementary school and middle school.

Nationally ELL students tend to trail further behind their peers in reading than in math. In grade 4, ELL students trailed 35 points behind white students in math, but the gap was 47 points in reading (Figure 1).

  1. The National Center for Education Statistics indicated that 3.8 million students received ELL services. It also reports that 8.9 million Hispanics were enrolled in school year 2003-04 (NCES, 2006). Thus, even if all English language learner students were Hispanic, less than a majority of Hispanic students could be ELL designated.
  2. The change in the black—white achievement gap as grades progress is the subject of considerable research. The assertion that the black to white and Hispanic to white reading gaps narrow is simply based on the NAEP and grade level comparisons within a NAEP assessment. It is not based on following the same cohort of children as grades progress. More sophisticated analysis also finds that the black-white math gap widens as children age but the reading gap remains relatively constant (Phillips, Crouse, and Ralph (1998)).