UC Irvine

This dissertation systematically estimated the differences in academic achievement trajectories based on children’s initial language status at kindergarten entry among Hispanic students. The dissertation also thoroughly tested the hypothesis that the academic advantage of bilingualism is operating through a cognitive channel using mediational analysis in a latent growth model framework. The major findings of this dissertation are as follows: 1. bilingual students with limited English proficiency have a much lower initial level of test scores in all the three subjects than their monolingual peers and thus are particularly in need of future interventions to narrow the initial achievement gaps. 2. lack of English proficiency at kindergarten entry does not hinder language minority students’ academic growth rate in reading, math, or science after controlling for family and school characteristics. 3. bilingual English proficient students have faster learning growth rates in reading, math, and science; bilingual students with limited English proficiency have faster growth rates in reading and math but not science. 4. Mediation analyses suggest that the growth rate of working memory mediates the growth rate in academic achievement for both bilingual English proficient and bilingual students with limited English proficiency, and the initial level of working memory and cognitive flexibility mediates the initial gaps between language minority students and monolingual students.