|Title:||The Effect of Dual-Language Immersion on Student Achievement in the Portland Public Schools|
|Principal Investigator:||Steele, Jennifer||Awardee:||RAND Corporation|
|Program:||Evaluation of State and Local Education Programs and Policies [Program Details]|
|Award Period:||3 years (7/01/2012-6/30/2015)||Award Amount:||$1,694,560|
Co-Principal Investigator: Robert Owen Slater (American Councils for International Education)
Purpose: Dual-language immersion programs seek to help the growing number of language minority students learn English and achieve academically, while giving language majority students the opportunity to develop proficiency in another language. These programs differ fundamentally from traditional foreign language programs in two important ways: 1) teachers deliver regular curricular content through a "partner language" (such as Spanish or Russian), but do not generally teach the partner language directly; and 2) students receive instruction in the partner language as early as kindergarten and may continue to receive language instruction through high school. While the primary goal of dual language immersion programs is for students to learn a second language, these programs have potential to improve student achievement across curricular areas. The researchers will examine the effect of the Portland Public Schools' (PPS) dual-language immersion programs on student achievement in English, mathematics, and science, and on student engagement as measured through student attendance and behavior.
Project Activities: Using administrative student panel data from PPS, the researchers will compare student achievement, attendance, and behavior outcomes between students who applied to a prekindergarten lottery to enroll in a dual-language immersion program and received a slot in a program and those who applied but did not receive entry.
Products: The products of this project will be evidence of the efficacy of PPS's dual language programs for students in grades 3 through 8. The evidence will be directly provided to PPS through the participation of department personnel on the project and to the research community and other education practitioners and policymakers through peer reviewed publications.
Setting: The study will take place in the Portland Public Schools in Portland, Oregon.
Sample: The participants will consist of seven cohorts of Portland Public School students who entered kindergarten in the academic years 2004–05 through 2010–11 and participated in the pre-kindergarten lottery for a slot in a dual-language immersion program. Participants will be drawn from the oversubscribed lotteries and include approximately 2,031 lottery winners and 1,143 students who did not receive a place in a dual-language program.
Intervention: Portland Public Schools operates ten dual-language immersion programs—seven in Spanish, and one each in Japanese, Mandarin Chinese, and Russian. Slots in the district's language immersion program are allocated to applicants through a stratified random-assignment lottery system. For the Russian and six of the Spanish programs, up to half of the slots are reserved for native speakers of the language and the remaining slots are for native English speakers (both types are selected by lottery). The other three programs cater primarily to native English speakers. From kindergarten through grade 12, immersion students receive at least half of their content-area instruction (e.g., math, science, social studies, etc.) in the target program language.
Research Design and Methods: This study uses a randomized, lottery-based design to estimate the difference in achievement, attendance, and behavior between students who are admitted to the immersion program through random assignment and students in the same lottery subgroup stratum who apply but are not admitted. Teacher and administrator interviews and classroom observations will be used to characterize the instructional variation among immersion classes and between immersion and non-immersion classes. Financial data from the district's financial office will be used to describe the fixed and marginal costs of immersion relative to the traditional curriculum.
Control Condition: The control condition is business as usual. Immersion-program applicants residing within district boundaries who are not randomly admitted to an immersion program through the district's lottery system receive the district's standard, English-based curriculum.
Key Measures: The primary outcomes for student achievement in math, science, and English language arts will be measured with the Oregon Assessment of Knowledge and Skills for students in 3rd through 8th grade. These students take the OAKs math and reading assessment in all of these grades, the writing assessment in 4th and 7th grades, and the science assessment in 5th and 8th grades. The primary outcomes for student engagement will be examined using annual attendance rates and annual number of behavioral referrals. In addition, partner-language proficiency will be collected for the students in the programs using the National Online Early Language Learning Assessment through grade 6, and with the Standards-based Measure of Proficiency in grades 7 and up.
Data Analytic Strategy: The impact within each subgroup strata in both elementary and middle schools will be estimated using using precision-weighted estimates to obtain average effects. To address the issue of non-compliance with assigned status, the student's randomly assigned status (lottery winner or loser) will also be used as an instrumental variable to estimate the average effect of attending a program for individuals who complied with their assigned status. An analysis of the bounds of the treatment effect will be done to address the possibility of selective attrition. For the seven programs that stratify student selection by student native language, estimates of the differential effect of the immersion program on native English speakers versus native speakers of a program's target language will be made.