Predictors of Second-Language Reading Performance

Abstract

This study identified individual, instructional and socio-cultural factors that are reliable predictors of limited-English-proficiency (LEP) children's reading performance in English. The reading comprehension section of the Metropolitan Achievement Test was used to assess the reading skills of 91 LEP students (K-12) with different linguistic backgrounds in the U.S. school system. Information about the independent variables was gathered from a special questionnaire. The data were analysed by means of multiple regression analysis. The results of regression analysis showed that in this sample, grade, Vietnamese language background, percentage of English spoken at home, higher level paternal occupation, and a relatively high number of students in the English-as-a-second-language class had positive effects on the reading skills of the subjects. With these data it was impossible to separate the impact of length of U.S. residence from the length of ESL instruction in U.S. Intensity of ESL instruction showed a statistically significant negative effect on the reading scores. There were several predictors which did not have a significant effect on reading performance: gender, number of siblings and sibling position, hours of television watching, number of public library visits, length parents' stay in the U.S., parents' levels of education and social status, and mother's staying at home. A central theme of this study is that single-variate approaches are inadequate to study the causes of reading performance; rather, several predictors should be employed to measure true relationships between L2 reading performance and independent predictors.