Supporting Oral Language Development Through Interactive Dialogic Reading

• In Event: Using Conversational Artificial Intelligence to Promote Active and Engaging Learning

Abstract

*Objective*

The benefits of reading with children can be magnified through dialogic reading, which involves having real-time conversations around the story narrative. Utilizing artificial intelligence, we developed a conversational agent embodied by a smart speaker to serve as a dialogic reading partner for preschool aged children. To test the potential benefits of these interactions, we conducted an experiment to showcase how dialogic reading with a conversational agent may increase children’s story comprehension and engagement as compared to an adult reading partner. To extend this study, we now partner with Sesame Workshop to develop dialogic e-books that promote parent-child interactions during joint reading.

*Theoretical Framework*

Past research emphasizes the amplified benefits of interactive storybook reading for preschool aged children through structured conversation around story narratives, such as asking open-ended questions and providing feedback to children (Arnold et al., 1994; Whitehurst, 1992). Children’s exposure to effective reading strategies builds the foundation for learning to read, yet varies depending on the skills and availability of their caregivers, with unequal access contributing to disparities in oral language and literacy at an early age (Farver et al., 2013). With their ability to support natural speech interactions, conversational agents may serve as effective dialogic reading partners for young children who are still developing their literacy and fine motor skills (Lovato & Piper, 2019).

*Method*

Children’s baseline oral language skills were measured. Using a 2 (dialogic reading or non-dialogic reading) x 2 (agent or human partner) factorial design, 117 children were assigned to one of the four conditions. Children’s comprehension of the storybook was measured using a 10-item questionnaire. Children’s engagement while reading was coded from recordings of the reading sessions.

*Results*

A regression analysis suggested that dialogic reading, with a human partner or a conversational agent, led to a 0.51 standard deviation (SD) increase in children’s story comprehension score (p < .001), with no significant difference between reading partners (human vs. agent). Dialogic reading also resulted in increased overall engagement (β = 0.41, p < .05) and narrative-related vocalization (β = 1.11, p < .001).

Dialogic reading with an agent also led to larger benefits for children who did not predominantly speak English at home than those who only spoke English at home.

*Scholarly Significance*

A properly designed conversational agent reading partner can offer benefits comparable to that of an adult partner. The benefits of a digital dialogic partner include enhanced engagement, encouragement of verbal utterances, and promotion of comprehension. As smart speakers become increasingly affordable and accessible, these findings are promising for the prospects of this technology in supporting children’s language development, especially for families with limited time, language skills, or resources to engage children in dialogic reading themselves. We have initiated a project with Sesame Workshop to develop culturally and linguistically relevant bilingual e-books with conversational agents to support parent-child joint reading between Latino children and their families. These e-books enhanced by parent-child-agent trialogue will model dialogic reading strategies and enrich children’s language learning environments.

Authors

• Valery Vigil, University of California - Irvine
  Presenting Author

• Ying Xu, University of Michigan
  Non-Presenting Author

• Kunlei He, University of California - Irvine
  Non-Presenting Author

• Andres Sebastian Bustamante, University of California - Irvine
  Non-Presenting Author

• Mark Warschauer, University of California - Irvine
  Non-Presenting Author