Poster #44 - Vocabulary Growth in Spanish-English Bilingually-Exposed Toddlers From 18 to 30 Months

• In Event: Poster Session 05

Abstract

Introduction. The Output Hypothesis (Swain, 2005) asserts that generating language (i.e., output), not only hearing it (i.e., input), leads to language competence. Yet, language acquisition research overwhelmingly centers on environmental factors, including exposure to caregivers’ language input, that influence children’s early development (Huttenlocher et al., 1991; Pearson et al., 1997; Place & Hoff, 2011). This is despite theoretical predictions that verbal interactions lead to language learning (Burner, 1985; Vygotsky, 1978) because they provide models of and practice with language (Snow, 1994; Tomasello, 1994). Accumulating research with monolinguals suggests that the number of caregiver-child conversational turns in toddlerhood predicts language skills 10 years later (Gilkerson et al., 2018; Romeo et al., 2018). Research with school-age bilinguals also shows positive associations between children’s reported output in a language and their concurrent skill level (Bedore et al., 2012) and subsequent growth in that language (Ribot et al., 2018). The current study adds to this past literature by examining the influence of child-level factors, including language output, on Spanish-English toddlers’ bilingual skills over time.
Participants. Participants included 47 Latino children (F = 24; M = 23). Children’s input consisted of equal amounts of Spanish and English, leaning toward more Spanish (M = 3.81; SD = .88; see below). Children were approximately 18- (M = 18.80; SD = 0.57), 24- (M = 24.84; SD = .66), and 30-months old (M = 30.92; SD = .78).
Method. Participants’ homes were visited longitudinally at child age 18-, 24-, and 30-months. At 18 months, children were video-recorded for 75-minutes. Transcripts of video-recordings generated child output as the number of English and Spanish tokens (i.e., total words; CLAN; MacWhinney, 2018) used per minute. Caregivers self-reported children’s input from “Only English” (5) to “Only Spanish” (1). They also completed the MacArthur-Bates Communicative Inventories (MCDIs; Fenson et al., 1994) to indicate children’s English (out of 396) and Spanish (428) words produced at each time-point. Conceptual vocabulary combined both languages, without translation equivalents. Additional child-level predictors were collected via MCDIs (comprehension) or researcher-developed questionnaires (gender).
Results. Spanish and Conceptual vocabulary grew faster than English (p’s < 0.05; Figure 1). Spanish and Conceptual vocabulary showed linear growth, while English showed significant acceleration over time (curvilinear-growth). Table 1 shows that children’s output was consistently associated with vocabulary growth on all three outcomes, even when controlling for input and other known predictors of language growth (comprehension, gender). Specifically, more opportunities for Spanish and bilingual output (English, Spanish) were related to higher initial (18-month) Spanish (p < 0.05) and Conceptual (p < 0.05) vocabulary, as well as their growth rates (p’s < 0.05). More opportunities for English output was associated with English vocabulary acceleration (p < 0.05). Overall, findings suggest that 1) bilingual children’s English will catch up to, and potentially surpass, Spanish after 30-months and 2) opportunities for children to practice language influences their vocabulary growth. Thus, to help children maintain their Spanish (and bilingual) language skills, researchers should focus on child-level predictors of language development (e.g., output), in addition to environmental influences (e.g., input).

Authors

• Jordan Sierra Perry, Loyola University Chicago
  Presenting Author

• Perla Blanca Gamez, Loyola University Chicago

• Maily Galindo, Loyola University Chicago

• Carla Jauregui