Associations Between Sentence Diversity and Common Language Measures in Young Dual Language Learners

• In Event: Poster Session 01
  In Poster Session: PS 01 Topic - Language, Communication

Abstract

Narrative language sampling is often recommended in addition to or instead of standardized language tests, and it is particularly useful for assessing the expressive language of children from culturally and linguistically diverse backgrounds (Castilla-Earls et al., 2020). The mean length of utterance in words (MLUw) and the number of different words (NDW) are commonly measured from both English and Spanish language samples (Castilla-Earls et al., 2020). Sentence diversity (SDv), or the number of unique subject-verb combinations, is a language sample measure that may better characterize emergent syntax than MLUw in young children with and without language delays (author, year). Given the importance of syntax development to social and academic participation (Owen Van Horne, 2020), the dynamic nature of bilingual language development (Bedore et al., 2012), and the growing number of children who speak Spanish at home (U.S. Census Bureau, n.d.), more measures of cross-linguistic syntactic knowledge are needed.
This study explored the within-language (Spanish, English) associations between SDv and measures of linguistic ability in preschool-age Spanish-English dual language learners (DLLs) with low expressive language skills. We addressed the following research questions: (a) For young DLLs, is SDv associated with receptive and expressive language skills as measured by standardized assessment? (b) Is SDv associated with expressive language skills as measured by MLUw and NDW? We hypothesized that, within both languages, DLLs’ narrative SDv would correlate with measures of expressive language (expressive vocabulary score, MLUw, NDW) but not receptive language.
We performed bivariate correlational analyses using a subset of data from a larger study (authors, date). The data included standardized assessment and narrative measures of language development in Spanish and English from 36 children (ages 59–68 months) who demonstrated balanced Spanish-English or Spanish language dominance. The standardized measures were the Expressive Vocabulary and Sentence Structure subtests of the English and Spanish versions of the Clinical Evaluation of Language Fundamentals—Preschool-2 (Wiig et al., 2004, 2009), and narratives were elicited using wordless picture books. The sample children had low Expressive Vocabulary scaled scores in both Spanish and English (Spanish M = 4.3, SD = 1.7; English M = 3.9, SD = 1.5). MLUw and NDW were calculated using SALT Software (Miller & Iglesias, 2019). To calculate SDv, bilingual coders identified the subject and verb of each active declarative sentence and counted the unique subject-verb combinations per narrative (author, year).
Results are in Table 1. Neither receptive nor expressive language scores correlated with SDv in either language for children in the sample. Partially confirming our hypotheses, NDW and MLUw correlated with DLLs’ SDv in both languages (also see Figure 1). These findings suggest that SDv is related to other narrative measures in English and Spanish but not to scores from standardized language tests. Longitudinal data are needed to determine if the addition of SDv measures improves predictions of long-term development in DLLs. Measuring SDv could be an important step toward improved and earlier access to services to support syntax development in Spanish and English for young DLLs.

Authors

• Natalie S Pak, Vanderbilt University
  Presenting Author

• Carol Scheffner Hammer, Columbia University
  Non-Presenting Author

• Monica Galindo
  Non-Presenting Author

• Pamela Hadley, University of Illinois at Urbana-Champaign
  Non-Presenting Author

• Ann P Kaiser, Vanderbilt University
  Non-Presenting Author