Individual Variables in Context: A Longitudinal Study of Child and Adolescent English Language Learners

Abstract

Millions of public-school students in the United States are identified as English language learners (ELLs), whose academic success is tied to their second language (L2) English education. Previous research in adult populations indicates that L2 proficiency is related to the contextual variable of the prevalence of one’s first language (L1) among their peers, called L1 density, which may also moderate the effects of individual variables such as age and exposure to the L2. Despite its substantial impact on the amount and quality of adult learners’ exposure to the L2, the variable of L1 density has received little attention in child and adolescent populations, even though it is unknown what role, if any, L1 density plays in L2 acquisition in a school context. Other outstanding questions concerning individual variables include the nature of the purported rate advantage of later starters and whether the similarity of one's L1 and L2 is related to L2 proficiency.The current study addressed these questions by analyzing longitudinal L2 proficiency assessment records of 10,879 ELLs in grades 1–12 in the United States. The assessment was WIDA's ACCESS for ELLs Online Test, a national, standardized test with scores for each of the four domains of listening, reading, speaking, and writing. Multilevel models were used to estimate the effects of several variables: age of enrollment in a United States school, length of enrollment, language similarity, and L1 density. In the fitted model estimates, age of enrollment had a small, positive effect. Length of enrollment had a sizable, positive effect but attenuated over time. ELLs enrolling at a later age progressed slightly slower than ELLs enrolling at an earlier age, contrary to the widely accepted notion that later starters enjoy a rate advantage. Little to no evidence was found for a relationship between test scores and language similarity or L1 density, or that the effects of age of enrollment or length of enrollment varied with L1 density. The results of this study give evidence for the following conclusions for ELLs in United States schools: an earlier age of enrollment is associated with greater gains in L2 proficiency over time, speakers of different L1s are not expected to become differentially proficient in L2 English, and ELLs’ levels of L2 proficiency are not expected to vary with how many of their peers speak the same L1.