UC Irvine

The dissertation consists of three studies that are the process of designing, developing, and evaluating generative-AI-powered story-authoring platforms for children. The first study focuses on the formative study on how stakeholders in education (i.e., teachers, parents, and students) perceive and leverage generative AI platforms (i.e., ChatGPT and Stable Diffusion) for writing activities. I found that the GenAI systems could be beneficial in generating adaptable teaching materials for teachers, enhancing ideation, and providing students with personalized, timely feedback. However, there are concerns over authorship, students’ agency in learning, and uncertainty concerning bias and misinformation. I provided design strategies to mitigate these constraints by implementing an adults-oversight system, balancing AI-role allocation, and facilitating customization to enhance students’ agency over writing projects.

The second study focuses on co-designing a logic model in informing designing AI-based Writing Tutoring Platforms (AWTP) with educators, then designing and evaluating AWTP prototypes that focus on opinion writing. From the co-design process, we identified the platform's potential users, features, functionalities, and desired outcomes. With this insight, we created a prototype, AWTP. The usability study findings with the AWTP prototype suggested AWTP’s efficacy in improving students’ writing engagement by increasing their time spent in writing, total word count, and lexical diversity. Feedback study revealed AWTP’s potential efficacy in improving motivation in writing by reducing anxiety over writing for emergent writers.

The third study focuses on designing, developing, and evaluating story-authoring platforms, StoryAI, for narrative writing for children. From the usability study, I found StoryAI’s efficacy in students’ perception of writing competencies (i.e., planning, translating, and revising) as well as AI literacy (i.e., perception, confidence, and motivation).

Overall, the three studies provide convincing evidence for leveraging GenAI-based learning platforms to support children’s literacy development. The findings are intended to contribute to designing interactive intelligent systems for learners and educators, leveraging story creation to teach various subjects including writing on different topics (i.e., science, history) to empower children’s long-term academic success and development.