Enhancing Children’s Self-Reporting in Chatbot Diaries through Rhyming Style

Shanshan Chen et al.

Children’s self-report is essential for research, education, and healthcare, yet existing methods such as surveys and diaries can be experienced as tedious and so lead to disengagement and low-quality responses. Chatbots have been suggested as a way to support children through conversational interaction, using age-appropriate language and an empathetic tone. Here we explore what could be suitable conversational styles for such chatbots. Specifically, we explore rhyme as a child-centered conversational style. We first conducted a co-design workshop with 35 children, which revealed preferences for short, playful, and soothing conversational patterns. Building on these insights, we designed a voice-based sleep diary in rhyming style and compared it to a prose style in a within-subjects study involving 40 children aged 8-12. Results show that rhyming prompts significantly improved response quality across question types and age groups, while maintaining high engagement even among children who preferred the prose style. We contribute proof-of-concept empirical evidence and design insights demonstrating how phonological scaffolding exemplified through rhyme extends the design space of capability-adapted chatbots beyond semantic simplification alone. While limited to short-term, lab-based sessions, this work provides initial evidence that conversational style can function as a design lever.