Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

ContextQ: Generated Questions to Support Meaningful Parent-Child Dialogue While Co-Reading

Griffin Dietz Smith, Siddhartha Prasad, Matt J. Davidson, Leah Findlater, R. Benjamin Shapiro

TL;DR

ContextQ tackles the challenge of scalable, high-quality dialogic reading prompts for home literacy. It uses a rubric-informed, self-supervised LLM question generation pipeline with a CEGIS-inspired feedback loop to avoid fine-tuning data, consisting of a question generation module and a co-reading tablet interface. An ablation study and a qualitative evaluation with 12 parent-child dyads show that ContextQ increases meaningful conversational turns and fosters deeper, life-relevant discussions, with parents often mediating between the app and child. The approach promises to close gaps in early literacy education by empowering untrained caregivers to cultivate richer dialogue during shared reading.

Abstract

Much of early literacy education happens at home with caretakers reading books to young children. Prior research demonstrates how having dialogue with children during co-reading can develop critical reading readiness skills, but most adult readers are unsure if and how to lead effective conversations. We present ContextQ, a tablet-based reading application to unobtrusively present auto-generated dialogic questions to caretakers to support this dialogic reading practice. An ablation study demonstrates how our method of encoding educator expertise into the question generation pipeline can produce high-quality output; and through a user study with 12 parent-child dyads (child age: 4-6), we demonstrate that this system can serve as a guide for parents in leading contextually meaningful dialogue, leading to significantly more conversational turns from both the parent and the child and deeper conversations with connections to the child's everyday life.

ContextQ: Generated Questions to Support Meaningful Parent-Child Dialogue While Co-Reading

TL;DR

ContextQ tackles the challenge of scalable, high-quality dialogic reading prompts for home literacy. It uses a rubric-informed, self-supervised LLM question generation pipeline with a CEGIS-inspired feedback loop to avoid fine-tuning data, consisting of a question generation module and a co-reading tablet interface. An ablation study and a qualitative evaluation with 12 parent-child dyads show that ContextQ increases meaningful conversational turns and fosters deeper, life-relevant discussions, with parents often mediating between the app and child. The approach promises to close gaps in early literacy education by empowering untrained caregivers to cultivate richer dialogue during shared reading.

Abstract

Much of early literacy education happens at home with caretakers reading books to young children. Prior research demonstrates how having dialogue with children during co-reading can develop critical reading readiness skills, but most adult readers are unsure if and how to lead effective conversations. We present ContextQ, a tablet-based reading application to unobtrusively present auto-generated dialogic questions to caretakers to support this dialogic reading practice. An ablation study demonstrates how our method of encoding educator expertise into the question generation pipeline can produce high-quality output; and through a user study with 12 parent-child dyads (child age: 4-6), we demonstrate that this system can serve as a guide for parents in leading contextually meaningful dialogue, leading to significantly more conversational turns from both the parent and the child and deeper conversations with connections to the child's everyday life.
Paper Structure (35 sections, 3 figures, 4 tables)

This paper contains 35 sections, 3 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Dialogic Reading as an Educational Practice
  4. Systems to Support Dialogic Reading
  5. Methods for Question Generation
  6. ContextQ
  7. Question Generation Module
  8. A Rubric for Dialogic Questions
  9. Question Generation Architecture
  10. Question Quality Evaluation
  11. Co-Reading Interface
  12. Iterative Design Testing
  13. Evaluation Study
  14. Participants
  15. Procedure
  16. ...and 20 more sections

Figures (3)

  • Figure 1: The question generation module utilizes a question synthesizer and suitability recognizer to produce high-quality dialogic questioning prompts. Dotted lines denote fixed inputs.
  • Figure 2: ContextQ presents dialogic questions to parents via a tablet-based reading application.
  • Figure 3: Top: For each interface, the proportion of parent-led dialogic interactions that began with each question type. Bottom: Proportion of dialogic interactions in each interface that touched on a particular topic. Meta conversations were discussions about the reading experience (e.g., do you like the story). Note: conversations could have more than one topic so percentages for each interface do not sum to 100%.