Dissertation Defense: Xinying Hou

The School of Information is pleased to announce the oral defense of Xinying Hou.

Title: Advancing Personalized, Engaging Learning and Assessment in the AI Era

Date: Friday, June 12, 2026

Time: 1:30-3:30pm

Location: Founders Room (LCSIB 3470) and Online via Zoom

Zoom: https://umich.zoom.us/j/93273199175

Barbara Ericson, serving as committee chair, will preside over the oral defense. 

All are welcome to attend!

Abstract:

Conventional learning environments often leave students disengaged and their varied needs unaddressed. Emerging technologies offer opportunities to foster deeper engagement and personalization in education at scale. This dissertation advances the design of engaging and personalized educational experiences within technology-enhanced learning environments through a series of empirical investigations and design explorations, organized into three approaches. The first approach drew on learning sciences theories to inform active learning design. I incorporated adaptive mixed-up code puzzles to scaffold code writing for undergraduate programming novices, and paired them with prompted self-explanation questions to support code comprehension. This approach did not yet address personalization. The second approach introduced personalization through a single AI-powered system to support more engaging learning experiences. I designed and evaluated CodeTailor, an interactive system that leveraged a large language model to deliver real-time mixed-up code puzzles tailored to students' incorrect code. The evaluation examined how active puzzle support compared to passively delivered solutions and how personalized support compared to common solution-based alternatives. The third approach extended these ideas to more complex personalization through multi-agent AI architectures. I designed a four-agent architecture for conversation-based formative assessment and developed LadderPuzzle, a dual-agent architecture that delivered personalized, non-code mixed-up puzzles to make solution planning an explicit and deliberate practice step. In summary, this dissertation contributes to the design and effectiveness of technology-enhanced learning environments, supporting learners as they build foundational knowledge and skills through engaging, personalized experiences.