A randomized study of language of reading instruction for English language learners: First-year findings

For many years, there has been much debate about how English language learners should be taught to read. Three main alternatives have been proposed: Structured English Immersion (SEI), Transitional Bilingual Education (TBE), and Two-Way Bilingual Education (TWB). This paper reports first-year findings from a four-year longitudinal randomized experiment comparing these alternative approaches. Spanish-dominant kindergartners beginning in 2004 and 2005 are the subjects. In one set of comparisons, 390 children were randomly assigned to be taught reading using either SEI or TBE. In another, 285 children were randomly assigned to be taught using SEI or TWB. All classes were taught using Success for All, equalizing all factors other than language of instruction. Measures of English and Spanish language and reading are being given each spring, along with observations, interviews, and other data. First-year outcomes are as expected. In the comparison of SEI and TBE, children in SEI scored much better in English while those in TBE scored much better in Spanish. Children in the TWB classes scored much better in Spanish than those in SEI, but the two groups were identical on English measures. These and other findings will be presented and discussed.