On Reading Books to Children

Parents and Teachers

Anne van Kleeck, Steven A. Stahl, Eurydice B. Bauer, Anne van Kleeck, Steven A. Stahl, Eurydice B. Bauer

• Publisher: Routledge
• Year: 2003
• eBook ISBN: 9781135643737
• Topic: Education
• Subtopic: Education General

On Reading Books to Children: Parents and Teachers brings together in one volume current research on adult book reading to children. The authors, drawn from around the world, are key researchers and eminent scholars from the fields of reading and literacy, child language, speech pathology, and psychology, representing multiple perspectives within these disciplines.

Chapters on the effects and limitations of book sharing are integrated with chapters discussing promising programs on storybook research. The reality of reading to children is more complex than it appears on the surface. The authors discuss some effects of and suggestions for reading to children that have emerged from the research. The ideas set forth in this volume will stimulate new lines of research on the effects of storybook reading, as well as refinements of current methods, yielding findings that enrich our understanding of this important arena of literacy development.

IV
Where Do We Go From Here?

13
Research on Book Sharing: Another Critical Look

Anne van Kleeck
University of Georgia

Which skills appear to provide preschoolers with important foundations for their later development of print literacy? How important is sharing books with preliterate children to their development of those skills? After nearly three decades of research, including a fair amount of controversy regarding the relative importance of book sharing to later literacy (Bus, van IJzendoorn, & Pellegrini, 1995; Dunning, Mason, & Stewart, 1994; Lonigan, 1994; Scarborough & Dobrich, 1994), research has made it increasingly clear that many of the different skills fostered during book sharing facilitate the later development of print literacy. Nonetheless, a number of conceptual and methodological problems have not been addressed to date. The goal of this chapter is to discuss persistent problems in this body of research, and to point to potential solutions for future research.

The conceptual and methodological wake-up call was first sounded by Scarborough and Dobrich in their 1994 review of empirical work on the impact of book sharing for preschoolers’ language and literacy development. Scarborough and Dobrich concluded that this impact was not terribly significant, and certainly far more modest than seemed to be generally accepted. Direct responses to their review by Dunning et al. (1994) and Lonigan (1994) took the stance that the Scarborough and Dobrich conclusion was premature. They pointed out a few misinterpretations, but mostly discussed various methodological weaknesses in the studies which Scarborough and Dobrich had reviewed, suggesting that the findings from these studies should be interpreted with caution.

While I interweave many of the points raised in all three of these articles, I focus on concerns that I have not seen adequately addressed in previous critiques of book sharing research. This discussion covers three broad areas: (a) aspects of adult–child interaction during book sharing that have been ignored in previous research; (b) the need to consider various characteristics of the books that are shared during these interactions; and (c) aspects of the nature and timing of measurements used in book sharing research.

ASPECTS OF ADULT-CHILD BOOK SHARING INTERACTION IGNORED IN PREVIOUS RESEARCH

In this section, I discuss three dimensions of adult-child interaction during book sharing. They are: (a) a consideration of how much challenging input (i.e., either new information or information that requires children to stretch their cognitive or linguistic abilities) is effective in a learning situation; (b) the possibility that book sharing interactions in mainstream culture families are as much (or perhaps more) about socializing children to verbally display their knowledge for adults as they are about fostering the development of literacy skills; and (c) the need to consider both the children’s and adults’ levels of participation in the interaction.

What Amount of Challenging Input Is Best for Learning?

Van Kleeck, Gillam, Hamilton, and McGrath (1997) designed a study to look at the relationship between levels of concrete and abstract language in mothers’ and fathers’ book sharing discussions when their children were between ages 3;6 (years;months) and 4;1, and their children’s gains on a formal measure of abstract language knowledge 1 year later. We found significant correlations between the amount of book sharing discussion at three of four different levels of abstraction (Levels I, II, & IV, with IV being the highest level of abstraction; see Table 13.1) and children’s subsequent gains at the highest level of abstraction (Level IV). While it made sense that the amount of parental discussion at the highest level of abstraction was related to the children’s gains at that level, why was the amount of input at the two lowest levels of abstraction (Levels I & II) related to gains at the highest level? We reasoned that one possibility may lie in the nature of a successful learning environment.

We had adapted our levels of abstraction from work by Blank and her colleagues that focused on the discourse of preschool teachers (Blank, Rose, & Berlin, 1978a). These authors had suggested that preschool teachers should aim to raise about 30% of their discourse to a level that would challenge children, while keeping the other 70% at a level that the children had already mas-tered, and which would therefore allow them to respond successfully. Amazingly, as a group, the 70 parents (35 mothers and 35 fathers) in our study provided an average of 37% of their book discussion at the two higher levels of abstraction (Levels III & IV) and 63% at the lower levels (Levels I & II).

These percentages, along with the significant correlations we had observed between parents who provided more input at Levels I, II (lower), and IV (highest) and children who made the greatest gains at the highest level of abstraction, led us to conclude that perhaps abstract language skills in children are fostered simultaneously in two very different, and seemingly opposite, ways. The first method keeps most of the interaction at levels the child has already clearly mastered, in order to create a climate in which the child feels competent and successful. The second method involves raising about a third of the interaction to levels that the child has not yet mastered, thereby creating challenges and opportunities for growth.

TABLE 13.1 Coding Categories for Levels of Abstraction

A similar finding was reported by DeLoache and DeMendoza (1987), but with children who were substantially younger than those in the van Kleeck et al. study. DeLoache and DeMendoza noted two levels of complexity in the information provided by mothers during book sharing with children who were 12, 15, and 18 months old. They found that 74% of the information provided was simple (almost all labels), while the remainder was more complex (factual information, dramatizations, references to the child’s experiences with related objects). They also noted, however, that the older children received significantly more complex input, with elaborations increasing from 12% to 23% to 42% across the three ages studied. Both the van Kleeck et al. (1997) and the DeLoache and DeMendoza (1987) studies suggest that the ratio of nonchallenging to challenging input is an important factor for future book sharing research to consider.

Verbal Display of Knowledge

There may be other, partial explanations for the fact that the majority of linguistic input during book sharing takes place at lower levels of abstraction, which children have already mastered or almost mastered. Another “agenda” may be at work—teaching children to “tell what they already know” or “verbally display their knowledge.” Scollon and Scollon (1981) noted that the practice of asking children to verbally display their knowledge is fostered by cultural beliefs, and is clearly not a universal phenomenon. In mainstream American culture, adults often ask a child to answer a question to which the adult already knows the answer. Such “known information” or “test” questions abound in parent–child book sharing routines, particularly with infants and toddlers.

This practice socializes young children into the cultural practices for displaying knowledge that they will later encounter in the classroom (Watson, 2001). Indeed, Reid (2000) noted that Mehan’s (1979) IRE model (teacher Initiation, followed by a child Response, followed by teacher Evaluation) is still the most common classroom participation structure, and that the most common type of teacher initiation is the “known information” question. Teachers are likely to think that children who display their knowledge are the children who have the knowledge. They are unlikely to recognize that children who do not respond simply may have been differently socialized in this regard (see Vigil & van Kleeck, 1996, for a discussion of the many other reasons children may not respond). The impact of this assumption on teachers’ perceptions of children’s competence, on teacher–child interactions (or lack thereof), and subsequently on children’s achievement is an empirically documented story with which we are all familiar.

The Scollons (1981) pointed out that the practice of having young children verbally display their knowledge is culturally determined. They noted that in American mainstream culture, the subordinate child is the “exhibitionist,” who is expected to “show off his abilities,” especially in school. For a child, such exhibition would include verbal displays of knowledge to adults. In other cultures, such as the Athabaskan societies studied by the Scollons (1981), the relationship is reversed: children are assigned the “spectator” role, while adults display their skills (p. 17). In such cultures, children are not socialized to verbally display their knowledge to adults.

Heath (1989) suggested that such variations in practice stem from different cultural beliefs about children’s development. In the United States, where mainstream adults typically believe that children are “raised” or “trained,” adults exhibit a more directive, pedagogical style of interaction. In other societies, where children are believed to “grow up” largely apart from parental intervention, adults “do not intervene with highly specific verbalizations of the ‘here and now’ or request recounts of shared events, except for societal ceremonial occasions” (p. 345). Such children might display their knowledge through actions, rather than verbalizations.

In subaltern cultural groups within the United States, verbal display may not be part of children’s socialization. Heath (1983) found that African Americans in the community she studied rarely asked children known-information (or test) questions, and when they did, it was in order to chastise. Heath (1989) also observed that many Mexican American communities refrain from known-information questions, except when teasing children (see Valdés, 1996, for more on the children of Mexican immigrants).

Verbal display of knowledge is rarely addressed in book sharing research. I have seen two exceptions: one is a study of book sharing (DeLoache & DeMendoza, 1987); the other—Goodman, Larrivee, Roberts, Heller, and Fritz, 2000—is worth mentioning for its methodology, although it actually focused on mother–child interaction during spontaneous play. The DeLoache and DeMendoza study looked at mothers interacting with 12-, 15- and 18-month-old children. The authors examined the mothers’ beliefs about their children’s knowledge of words taken from the book that had just been shared, and asked the mothers to indicate whether the child could produce, comprehend only, or was unfamiliar with each word. The mothers were significantly more likely to skip over pictures when they thought their child did not know the label for that picture (they asked for such labels only 8% of the time). They were also much more likely to ask the child for labels they thought the child could produce (49% of label requests), than for those that they believed the child could only comprehend (18% of label requests).

Goodman and her colleagues (2000) had mothers complete the McAurther Communication Development Inventory (Fenson et al., 1993), a measurement tool that asks a parent to indicate all the specific vocabulary items that their child knows. In comparing these maternal reports to the questions the mothers asked their 30-month-old children during play, it was found that the children already knew 94% of the words that they were asked to supply in response to their mothers’ questions. While Goodman et al. concluded that the mothers’ agenda appeared to be the maintenance of discourse with their children, and not the teaching of new vocabulary, their data would also strongly support the idea that the mothers were socializing their children to verbally display their knowledge. Interestingly, in a series of studies by Sénéchal, and her colleagues exploring the impact of book sharing on vocabulary development, the idea that children are being socialized to verbally display their knowledge is not discussed (e.g., Sénéchal, 1997; Sénéchal, & Cornell, 1993; Sénéchal, Thomas, & Monker, 1995).

The verbal display theory suggests a variety of possible refinements for future research. First, the possibility that the best learning e...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• Preface
• I: Book Sharing in Families
• II: Storybook Reading in the Classroom
• III: Storybook Sharing as Cultural Practice
• IV: Where Do We Go From Here?