UC Irvine

We examined spelling errors patterns, cohesive ties, and syntax features in English and Spanish essays written by Spanish-English emergent bilinguals in Grades 1, 2, and 3 (N = 278; 51% female) enrolled in either English immersion or English-Spanish dual immersion programs. In Study 1, we addressed whether students made consistent spelling errors that could be due to crosslinguistic influence, and whether these spelling errors differed by grade level, English learner status, and instructional program while controlling for free and reduced lunch status. Spelling errors potentially due to crosslinguistic influence were coded. We found that students tended to make crosslinguistic errors only in one language and not in the other language (i.e., either Spanish-influenced errors in English or English-influenced errors in Spanish). Students in the dual immersion program made more Spanish-influenced spelling errors in English compositions while students in the English immersion program made more English-influenced spelling errors in Spanish compositions. Students in higher grades made less Spanish-influenced spelling errors in English compositions than students in lower grades. In Study 2, we investigated whether syntax features (mean length of utterance, number of verbs, subject agreement words, and subject agreement words accurately conjugated) differed by grade level, English learner status, and instructional program while controlling for free and reduced lunch status, whether syntax features in one language predicted its counterpart measure in the other language, and whether syntax features were related to writing quality within languages. Essays by students in Grades 2 and 3 had longer utterances, more verbs, and greater accuracy than Grade 1. Mean length of utterances and subject agreement accuracy were significantly associated with its counterpart in the other language, respectively. Syntax features were related to quality, but the magnitude of relations differed by English learner status and instructional program. Finally, in Study 3, we explored whether reference, conjunction, and lexical ties measured by quantity and accuracy were predicted crosslinguistically by counterparts in the other language, whether they differed by grade level, English learner status, and instructional program while controlling for free and reduced lunch status, and whether they were related to essay quality within and across languages. Students in Grades 2 and 3 used a greater number of cohesive ties and more accurately than Grade 1, English learners used fewer cohesive ties in English and Spanish essays than non-English learners, and dual immersion students wrote a greater number and more accurate Spanish cohesive ties than English immersion students whereas the opposite was found for English immersion students and English essays. Reference and lexical ties predicted writing quality within each language, and English cohesive ties predicted Spanish writing quality. Taken together, these results suggest that spelling errors by emergent bilinguals in primary grades are associated with instructional program and English learner status, which is related to spelling in the other language, and that written syntax features and cohesive ties predict writing quality for Spanish-English emergent bilinguals in primary grades.