Information Sources for Word Meaning in Children's Picture Books

• In Event: Poster Session 01
  In Poster Session: PS 01 Topic - Language, Communication

Abstract

Picture Books are an important tool for promoting early language acquisition (Horst & Houston-Price, 2015). The range of mechanisms by which they support acquisition are thus of great interest. In this study, we investigated the relative and joint contributions of verbal and non-verbal context in in supporting word learning from picture books. We employed a modified version of the Human Simulation Paradigm (Gillette et al., 1999) where naïve adult participants identified the identity of common nouns (e.g., “bag”, “face”, “school”) from the illustrations in which those nouns appeared and/or the linguistic context in which those nouns occurred. Of interest was the relative contributions of illustrations and linguistic context to noun identification, and whether these contributions shifted with the age of the picture book’s target audience.

Methods

Adult participants (N = 48) were randomly assigned to the Image, Text, or Combination conditions. Participants completed 40 trials in which they had to identify a “mystery noun” (e.g., “bag”, “face”, “school”, etc.). Depending on condition, participants either saw the illustration in which the mystery noun appeared (Image condition; Fig 1B), the text surrounding the mystery noun (Text condition; Fig 1C), or both illustration and text (Combination condition; Fig 1D). Each trial’s mystery noun was unique and came from commonly-read storybooks geared either for infants and toddlers (i.e., publisher’s age range between 0-5 years) or preschoolers and young children (3-8 years).

Results and Discussion

Our study revealed several findings on the sources to word meaning in picture books. First, performance in the Image condition was very low (M = .16, SD = .05), demonstrating high degrees of referential uncertainty in the illustrations. Second, participants in the Text condition (M = .32, SD = .11) outperformed those in the Image condition, p < .001, highlighting the value of linguistic information in picture books (Fig 2A-B; see also Cameron-Faulkner & Noble, 2013; Montag et al., 2015). Third, not surprisingly, participants who saw both the illustration and text performed the best (M = .61, SD = .10), p’s <.001. It is noteworthy however that performance in the Combination condition was greater than the combined performance of the Image and Text conditions (see Fig 2C; p < .001), suggesting that the text and illustrations augment and amplify the information conveyed by each (see Schwarz, 1982). Finally, the contributions of illustrations and text change with the target audience, suggesting the possibility that how picture books support word learning may change with development. Whereas the role of illustrations weakened with target audience age, the role of linguistic context strengthened, F(1,28) = 6.41, p = .02 (see Figure 2D).

Picture books and picture book reading are widely regarded as an invaluable tool for learning (Hudson Kam & Matthewson, 2017; Wagner, 2017). Understanding how different elements of picture books support word meaning identification advances our understanding of how picture books may support learning, of the nature of the word learning input more generally, and may ultimately help educators and practitioners in utilizing picture books for intervention purposes.

Authors

• Sadie MacDonald, University of Connecticut
  Non-Presenting Author

• Brianna Kinnie, University of Connecticut
  Non-Presenting Author

• Audra E. Logan, University of Connecticut
  Non-Presenting Author

• Zoe Green, University of Connecticut
  Non-Presenting Author

• Sara Johnson, University of Connecticut
  Non-Presenting Author

• Sumarga H. Suanda, University of Connecticut
  Presenting Author