The Impact of Book Reading on Children with Low and Typical Language in Preschool Classrooms

• In Event: Vocabulary Instruction in Early Education Settings: Investigating Curriculum Support, Pedagogical Techniques, and Child Language Outcomes

Abstract

This paper examines the growth of receptive and expressive vocabulary in low and typical language Head Start preschoolers over an academic year in relation to teacher speech during shared book reading. Differences in preschool oral language skills have long-lasting effects on academic achievement (Cunningham & Stanovich, 1997; Dickinson & McCabe, 2001). Amount, richness (Weizman & Snow, 2001), and syntactic complexity (Huttenlocher, et al., 2002) of talk are factors found to influence child language growth. Shared book reading provides a context for language development in young children (Wasik & Bond, 2001). A recent meta-analysis found different effects of book reading on expressive vocabulary (d= .62) and receptive vocabulary (d= .45) (Mol, Bus, & de Jong, 2009). Teachers who utilize language promoting strategies during book reading promote growth of children’s expressive and receptive vocabulary (Wasik, Bond, & Hindman, 2006). However, children with higher initial vocabularies show greater growth than their peers with lower vocabularies (Ewers & Brownson, 1999; Penno, Wilkinson, & Moore, 2002).
Four hundred eighty-nine children enrolled in Head Start were randomly assigned to 52 classrooms in six clusters, which were assigned to one of three conditions: business as usual, treatment curriculum, and treatment curriculum plus intensive language support. After assignment, children were categorized as low language (PLS total score < 75) or typical language (PLS score > 75)(PLS-III; Zimmerman, Steiner & Pond, 1992). All children participated in whole-class readings that were conducted according to curricula specifications. Children were assessed at the beginning and end of the year with the Peabody Picture Vocabulary Test (PPVT-III; Dunn & Dunn, 1997) and the Expressive Vocabulary Test (EVT; Zimmerman, Steiner & Pond, 1992). Teacher talk was recorded and transcribed for analysis using the CLAN program (MacWhinney, 2001), and coded for syntactic complexity, lexical richness, and amount of talk. Multi-level regression models that controlled for classroom risk factors and initial vocabulary scores were developed to predict end of year receptive and expressive vocabulary scores for the two groups of children.
Mean growth in the raw scores for expressive (low language = 12.4; typical language = 13.8) and receptive vocabulary (low language = 17.2; typical language = 15.1) was observed, however no significant differences in growth between the groups were found. For children in the typical language sample, teachers’ syntactic complexity, lexical richness, and amount of talk were significant predictors of end of year receptive, but not expressive vocabulary. For children in the low language sample, lexical richness and amount of talk were significant predictors of end of the year expressive vocabulary, but not for receptive vocabulary. Curriculum condition, gender, and age of child were not significant factors for either group.
At-risk children are not a linguistically homogenous group that uniformly benefits from shared book reading experiences. These results indicate that children in the same classrooms at the same time were experiencing language input in different ways, underscoring the importance of taking different aspects of verbal input into consideration for better understanding language growth for at-risk preschoolers with varying levels of vocabulary knowledge.

Authors

• Erica M. Barnes, Vanderbilt University

• David K. Dickinson, Vanderbilt University