Poster #116 - Variation in Communication Style Across Spanish- and English-Speaking Parent-Child Dyads During Free Play

• In Event: 3-003 - Poster Session 09
  In Poster Session: PS 09 Section - Language, Communication

Integrative Statement

In parent-child engagement, both the quality (Hirsh-Pasek et al., 2015) and the quantity (Hurtado et al, 2008; Weisleder & Fernald, 2013) of language input are positively associated with vocabulary. However, in our recent study (Masters, Villanueva, Poulin-Dubois, Zesiger, & Friend, 2018), we found that Spanish-speaking dyads use fewer examples of rich symbolic engagement during free play relative to their English-speaking peers. Importantly, rich symbolic engagement is positively related to early vocabulary. What other characteristics of parent-child engagement can help to clarify this pattern in our Spanish-speaking sample?
We examined two communicative strategies common in speech to young children: Recasts/Expansions and Redirects. Recasts/Expansions are positively associated with vocabulary (Farrar, 1992; Nicholas, Lightbown, & Spada, 2001; Tomasello, 1992), whereas Redirects are negatively associated (Dunham, Dunham, & Curwin, 1993; Tomasello & Farrar, 1986). We expected that Spanish-speaking parents would engage in fewer recasts and expansions but more redirects relative to English-speaking parents. Such a pattern would help to explain the discrepancy between samples in the use of rich symbolic engagement.
Participants were 50 Spanish- and 50 English-speaking parent-child dyads from our original study, of which we present a preliminary sample of 34 Spanish (15 girls, Mage = 31;14, range = 28;18 to 37;6) and 35 English (20 girls, Mage = 30;7, range = 28;0 to 32;21). Dyads participated in 20-minute free play sessions and parents completed the MCDI:WS, a measure of expressive vocabulary. Conversations were transcribed to inter-rater agreement = .90 and coded to inter-rater agreement = .87.
The Spanish-speaking sample was characterized by lower parent education and child expressive vocabulary than the English-speaking sample (t(52.88) = 52.88, p <.001 and t(63.09) = 5.96, p < .001, respectively). However, the number of conversational turns did not differ across samples (Ms = 309.41 and 314.14, respectively, p = .858). Overall, Recasts/Expansions were significantly correlated with expressive vocabulary (r(69) = .489, p< .001). Redirects were not. However, Language group was significantly correlated with Recasts/Expansions (r(69) = .530, p < .001) and with Redirects (r(69) = -.238, p < .001) such that, consistent with our prediction, Spanish-speaking parents used fewer recasts and expansions but more redirects relative to their English-speaking peers.
In previous research, we found that Spanish- and English-speaking dyads engage differently during free play: both groups use rich symbolic engagement but this effect was attenuated in our Spanish-speaking sample. The present study clarifies these findings by identifying one source of this attenuation. Specifically, opportunities for rich symbolic engagement may be reduced in the Spanish sample by the presence of a high rate of parent redirects.
We plan to extend these preliminary results to report the results on a total sample size of 100 dyads (50 Spanish and 50 English). This will allow us to clarify the relation between the language spoken by dyads (English or Spanish) and characteristics of dyadic engagement as well as to calculate more stable estimates than these preliminary results allow.

Authors

• Allyson S. Masters, San Diego State University
  Presenting Author

• Anele Villanueva, San Diego State University
  Non-Presenting Author

• Diane Poulin-Dubois, Concordia University
  Non-Presenting Author

• Pascal Zesiger, University of Geneva
  Non-Presenting Author

• Margaret Friend, San Diego State University
  Non-Presenting Author