Title

Let's Read!/¡Vamos a leer! Caregiver Speech During Book Sharing in English- and Spanish-Speaking Families

Publication Date

4-9-2021

Document Type

Presentation

Department

Child and Adolescent Development

Disciplines

Bilingual, Multilingual, and Multicultural Education | Child Psychology | Language and Literacy Education

Publication Title

Society for Research in Child Development 2021 Virtual Biennial Meeting

Conference Location

Virtual

Abstract

It is well-established that caregiver speech during book sharing supports children’s language development. Studies have found that caregivers who produce more utterances during book sharing have children with larger vocabularies, increased motivation to read, and improved reading comprehension. However, book sharing consists of not only recited text (recited speech), but also caregivers’ spontaneous comments and questions to support understanding of the book (spontaneous speech). Previous studies have typically analyzed book activities during a controlled setting in the laboratory or at home with pre-selected books, or have relied on parent reports of book sharing frequency. And, they have focused primarily on preschool-aged children in predominantly English-speaking families. In this study, we investigated the quantity and quality of recited versus spontaneous speech during at-home, unprompted book sharing activities in English- and Spanish-speaking families from diverse socioeconomic (SES) backgrounds.
This study explores naturalistic book sharing episodes excerpted from day-long LENA recordings in English-speaking (n=22) and Spanish-speaking families (n=20) with 24-month-old children from diverse SES backgrounds (Table 1). To identify book reading episodes, we sampled six 10-min recordings for each family from their densest language interactions, based on highest adult word counts (AWC) consisting primarily of child-directed speech. From that top hour of dense language, trained listeners identified all book-sharing interactions and transcribed them verbatim, coding for recited and spontaneous speech. We then derived the frequency of word tokens and types (adjusted for recording length), and mean length of utterance (MLU).
Preliminary analyses in the English-speaking sample showed that recited speech contained more tokens/minute (recited: M=119.07, SD=109.08; spontaneous: M=88.69, SD=44.43), more types/minute (recited: M=43.10, SD=35.42; spontaneous: M=35.44, SD=15.39) and significantly longer MLUs (recited: M=6.11, SD=1.90; spontaneous: M=3.79, SD=0.92) than spontaneous speech, t(21)=6.10, p< .001. Moreover, those caregivers who produced more word tokens/minute, word types/minute, and longer utterances during both recited and spontaneous speech were from higher-SES backgrounds (rs=.39-54, ps<.01). This suggests that caregivers from higher-SES backgrounds may be reading books with more complex language than caregivers from lower-SES backgrounds. At the same time, those families who produced more recited word types/minute were also those who produced more spontaneous word types/minute (Figure 1, r=0.51, p<.02), even controlling for SES (r=.49, p<.02). One interpretation is that there is a feedback loop where choice of book determines features of recited speech, thereby influencing features of spontaneous speech. Alternatively, there is also evidence for continuity across caregivers that transcends book choice.
Analysis of data from the Spanish-speaking families is ongoing. In both datasets, we will further examine the role of caregiver questions and elicited child responses about book content, as well as links to child vocabulary size and skill in real-time language processing. By investigating both recited and spontaneous speech during book sharing interactions, we can better understand which features of caregivers’ speech support children’s development.

Share

COinS