Page 1 of 3

English Learners in North Carolina Dual Language Programs

Year 3 of this Study: School Year 2009-2010

A Report by Wayne P. Thomas, Ph.D. and Virginia P. Collier, Ph.D.

Executive Summary Prepared for the North Carolina Department of Public Instruction

Background Like other states, North Carolina has experienced a tremendous increase in students whose

first language is not English in recent years. Graduates of English-as-a-Second-Language

(ESL) programs still exhibit substantial achievement gaps compared to students whose home

language is English. As a result, NCDPI has contracted with Wayne P. Thomas and Virginia

P. Collier of George Mason University, nationally recognized for their studies of English

language learners, to research the effectiveness of dual language programs in addressing

these gaps. This report presents the third year cross-sectional results of a longitudinal study

to evaluate education programs for English learners in North Carolina Public Schools.

Dual

Language

Education

Defined

The programs studied in this report are two-way dual language programs: programs in which

English language learners and native English speakers are educated in the classroom together

in both English and another language (the home language of the English learners). The

curriculum (North Carolina Standard Course of Study) is presented in English part of the

time and a second language at least half the time (Spanish, Chinese, French, German, or

Japanese). All state-level tests are administered in English.

Data Sample The data sample analyzed consists of all available 2009-2010 (School Year 2010)

information on students in the school districts operating at least one dual language program

for at least four years:

(1) Chapel Hill-Carrboro City: Carrboro Elementary (PK-5), Glenwood Elementary (PK-5),

and McDougle Middle School (6-8)

(2) Charlotte-Mecklenburg: Collinswood Language Academy (K-8), Oaklawn Language

Academy (K-8), and Smith (now Waddell) Language Academy (K-8)

(3) Chatham County: Siler City Elementary (K-5)

(4) Durham County: Southwest Elementary (PK-5)

(5) Greene County: Snow Hill Primary (K-2), West Greene Elementary (3-5), Greene County

Middle School (6-8)

(6) Winston-Salem/Forsyth County: Ashley Elementary (K-5)

Charlotte-Mecklenburg schools have whole school dual language programs. The other

district programs include at least one dual language class per grade. Most programs are

Spanish-English and have close to an equal number of non-English language and native

English speakers at each grade level.

Students who attended school at least 90 days and took an EOG test were included in the

study. Both dual language and non-dual language students (in schools with dual language

and schools without) are included in the sample for comparison purposes.

Page 2 of 3

Analysis Students are divided into three mutually exclusive groups (Thomas-Collier disaggregation

groups) with subcategories for analysis:

(1) English learners (LEP students) (N=12,179)

(2) Language minority students who are not English learners (N=7,097)

(3) Non-language-minority native English speakers

a. African Americans (N=37,435)

b. Whites (N=36,682)

c. Multi-ethnic (N=4,337)

In addition, separate analyses of two groups of interest from the total sample are:

(1) Students of low socioeconomic status (as measured by participation in free and reduced

lunch services)

(2) Students with exceptionalities (who qualify for special education services)

Mean Reading and Math scores are analyzed separately by grade. Statistical significance

(determined using confidence intervals) and practical significance (determined by measures

of effect size) are reported.

Significant

Findings

The distributions of Reading and Math scores in the six-district study sample are

approximately the same as statewide figures for each grade; therefore, the conclusions from

the study sample are likely to represent the state. Findings in favor of dual language

programs are consistently both statistically and practically significant.

 Overall, Reading and Math scores of students in two-way dual language education

are higher for all dual language students, regardless of ethnicity, socioeconomic,

LEP, or special education status. In most cases, by the middle school years, two-way

dual language students, regardless of subgroup, are scoring as high in Reading and

Math as non-DL students in the next highest grade who are typically one year older.

In some cases, the DL students are scoring as much as two years higher than their

non-DL peers.

 Both English learners (LEP students) and African American native English speakers

initially show very large achievement gaps when compared to White native English

speakers, regardless of subgroup. Achievement gaps are also present among students

of low-socioeconomic and special education status. However, when these groups

participate in dual language programs, their achievement gaps are smaller than the

gaps of the same groups who are not in dual language classes.

 Overall, low-socioeconomic-status (SES) students score much lower in Reading and

Math than students who are not low-SES. Many African American students and LEP

students are low-SES. Low SES may be a primary factor affecting differences in

Reading and Math achievement for these student groups. However, based on three

years of analyses (school years 2008, 2009, and 2010), we find that two-way dual

language programs significantly address the achievement gap for low-SES students.

 Students with exceptionalities (special education students) typically score

significantly lower in Reading and Math than students without exceptionalities. But

there is substantial evidence in the findings from analyses of school years 2009 and

2010 that special education students in two-way dual language classes who took the

End-of-Grade tests outscore special education students not in dual language classes.

Page 3 of 3

Researchers’

Conclusions

Findings strongly suggest that there are qualities to North Carolina’s two-way dual language

programs that confer greater educational gains in Reading and Math compared to non-dual

language education. After three years of studies, we conclude that two-way dual language

education is an effective way to improve the Reading and Math scores of all North Carolina

students.

There are higher percentages of Hispanics and Whites participating in North Carolina’s two- way dual language programs than African Americans. Given the large number of African

Americans in North Carolina’s student population, the lower Reading and Math scores of

African American students statewide, their under-representation in two-way dual language

programs, and the advantages that two-way dual language education provides African

American students, we recommend efforts to increase African American enrollment in dual

language programs.

Low-socioeconomic status may be a driving factor in lower achievement among English

learners and African American students. Given that low-socioeconomic-status (SES) students

in dual language programs score significantly higher in Reading and Math than low-SES

students not in dual language programs, two-way dual language programs can address the

achievement gap for many English learners and African American students.

In Reading, low-SES African American students attending dual language classes are scoring

at least one grade higher than their low-SES African American peers not in dual language as

early as 4

th

grade. Furthermore, in Math, low-SES African American dual language students

are more than two grades ahead of low-SES African American non-dual language students as

early as 5

th

grade. These findings indicate that two-way dual language programs are a school

reform model for addressing the achievement gap for formerly “at risk” groups.

Low-SES students are 41% of the total enrollment in dual language programs in this Year 3

study, illustrating that families of low income are well represented in this database. The fact

that dual language programs are successful in raising the scores of this “at risk” group lends

even greater significance to the higher achievement of students attending dual language

programs.

In both Years 2 and 3 of these studies, exceptional (special education) students in two-way

dual language classes have outscored exceptional students not in dual language classes.

Exceptional students deemed ready to take the EOG tests may benefit from placement in dual

language classrooms, possibly because of the cognitive stimulation afforded by mastering the

curriculum in more than one language and by acquiring a second language in addition to

English. This intriguing finding deserves further investigation in future analyses with larger

samples of students with exceptionalities in dual language programs.

In summary, after analyzing three separate years of data, we find that dual language classes

increase the Reading and Math achievement of all students regardless of subgroup, and are a

substantially effective means of addressing North Carolina’s large achievement gaps for

limited-English-proficient students, non-language-minority native-English-speaking African

American students, students of low-socioeconomic status, and special education students.

Longitudinal studies of student cohorts over multiple school years, beginning with the

addition of school year 2011 and 2012 data, will expand on these findings in the next report.