The Role of Vocabulary Type and Cross-Linguistic Component Skills in Dual Language Learners' Expository Reading

• In Event: Disrupting the Status Quo in Vocabulary: Radically Rethinking "What," "Who," and "How" for Vocabulary Instruction

Abstract

Skilled expository reading is important for middle-school dual language learners (DLLs) as content area learning at this age frequently relies upon extensive independent reading. Expository reading is supported by vocabulary instruction focusing on key content, high-utility or academic vocabulary (August et al., 2014; Crosson et al., 2019; Lesaux et al., 2014; Snow & Uccelli, 2009; Vaughn et al., 2009). Knowledge of subject content vocabulary in particular is thought to support DLLs’ reading comprehension, not only by conveying key conceptual information but also by enabling fluent, automatic reading that frees the reader to engage with high-level text meaning (Carlo et al., 2004; Peng et al., 2018). Eye-tracking research with first language (L1) speakers has revealed that function words are quickly and easily read and more often skipped over in comparison to content words (Rayner, 1998). However, less is known about the relative importance of content vs. function words in supporting fluent reading for DLLs.
In the current study, we examined Spanish-English bilingual middle-schoolers’ English expository reading and in particular, aimed to understand the role of cross-linguistic reading component skills such as decoding, vocabulary knowledge, and syntactic/sentence integration skills in supporting fluent reading. Fifty-nine middle-school students (mean age = 12.9 yrs., sd=1.1) who were either L1 speakers of English (n=29) or Spanish (n=30) participated. Students read four expository passages from a reading inventory commonly used in schools while their eye movements were recorded. Additionally, they completed standardized assessments of decoding, vocabulary, syntax and comprehension in English (all students) and Spanish (L1 Spanish speakers only). We used nonlinear principal components analysis to reduce data dimensionality of eye measures that included first-pass and total fixation durations and counts, gaze durations, regression counts, saccade amplitudes and counts, and skipping rates. The resulting first principal component served as a proxy measure of silent reading fluency. We entered this variable in a generalized multilevel linear regression model with Woodcock-Muñoz normative reading component measures as predictors.
For both L1 English and L1 Spanish speakers, English decoding skills were positively associated with more fluent processing of both content and function words. For L1 Spanish speakers, both English and Spanish decoding as well as Spanish syntactic skills displayed this positive association. English vocabulary knowledge facilitated content word processing for L1 English speakers only. For L1 Spanish speakers, Spanish, but not English, vocabulary knowledge facilitated content word processing in comparison to function words. These findings suggest that while vocabulary instruction generally focuses on key content words, effective vocabulary instruction for DLLs may need to take into account the role of function words in fluent text processing. Furthermore, the optimal degree to which vocabulary instruction incorporates teaching of function word structures may vary according to both first and second language knowledge and skills.

Authors

• Sibylla Leon Guerrero, Harvard University

• Laura Mesite, MGH Institute of Health Professionals

• Veronica Whitford, University of New Brunswick

• Gigi Luk, McGill University