Cross-Lagged Effects of Spanish and English Skills of Dual Language Learners in Head Start Programs

• In Event: Poster Session 03
  In Poster Session: PS 03 Topic - School Readiness/Childcare

Abstract

Introduction
Children learning English while continuously acquiring their home language are considered dual language learners (DLLs) in the U.S. (Ballantyne et al., 2008). DLLs include 29% of 3- to 5-year-old children enrolled in Head Start (HS), a federally funded early education program primarily serving preschoolers from low-income backgrounds. There has been increasing understanding that the development of one language may benefit the development of another (linguistic interdependency hypothesis; Hammer et al., 2014), potentially due to a common proficiency across languages (Cummins, 1979). Yet, there is a limited body of research empirically testing the hypothesis, especially for young DLLs who start their learning in school settings with differing proficiency in English language. Differing English skills in the beginning of preschool may lead to different language development processes, entailing different language-related experiences and support. This study aims to fill these literature gaps by examining the following:
1. Are there cross-lagged effects between Spanish and English vocabulary skills in DLLs during preschool and kindergarten years?
2. Are the patterns of cross-lagged effects between Spanish and English vocabulary skills different between DLLs with differing initial English proficiency?
Method
Participants. This study used the Head Start Family and Child Experiences Survey 2009 (FACES 2009). Participants included 252 Spanish-speaking DLLs whose primary spoken home language was Spanish, as reported by parents, and who started HS at age 3.
Measures. English and Spanish receptive vocabulary skills were assessed using the Peabody Picture Vocabulary Test (PPVT-4; Dunn et al., 2006) and the Test de Vocabulario de Imagines Peabody (TVIP; Dunn et al., 1986), respectively. Depending on the English proficiency measured by preLAS (Duncan & DeAvila, 1998), FACES identified DLLs who had sufficient English proficiency to receive assessments in English. Using this information obtained at HS entry, we grouped children into emergent bilinguals (limited English proficiency) or bilinguals (sufficient English proficiency).
Analysis. Mplus was used to for path analysis. First, a full sample, cross-lagged path analysis was conducted across four data points (HS entry, HS year 1, HS year 2, and kindergarten). Then, we ran a multi-group, cross-lagged path analysis to examine differences between emergent bilinguals and bilinguals. The Maximum Likelihood was used for parameter estimation. All analyses considered the center-level clustering and a sampling weight.
Preliminary Results
Cross-lagged path analysis showed that Spanish and English vocabulary skills did not reciprocally predict each other. Multi-group analysis determined that Spanish vocabulary skills in HS Year 1 predicted English vocabulary skills in HS Year 2 for bilinguals (β = .18, p < .05), but not for emergent bilinguals. Additionally, significant concurrent associations between Spanish and English skills were found in HS Year 2 for both groups.
Conclusion
We found limited evidence of reciprocal associations between Spanish and English receptive vocabulary. Only bilinguals with initial English proficiency showed Spanish vocabulary predicting their subsequent English vocabulary during preschool. This suggests that cross-linguistic transfer may occur after children obtain a certain level of proficiency in the second language. DLLs may need support for the development of both languages during early childhood.

Authors

• JinHee Hur, California State University, Bakersfield
  Presenting Author

• Ji Young Choi, Iowa State University
  Non-Presenting Author

• Seung-Hee Claire Son, University of Utah
  Non-Presenting Author