Reproduce “Designing AI Peers for Collaborative Mathematical Problem Solving with Middle School Students: A Participatory Design Study” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Wenhan Lyu, Yimeng Wang, Murong Yue, Yifan Sun, Jennifer Suh, Meredith Kier, Ziyu Yao, Yixuan Zhang
Researchers at William & Mary and George Mason University conducted a participatory design study to understand middle school students' perspectives on AI peers for collaborative math problem-solving. The work identifies specific student preferences and tensions, such as the desire for co-agency and scaffolded help over direct answers, guiding future AI peer design.
Key insights
- Students desired co-agency, shared effort, and mutual responsiveness from AI peers, perceiving rigid or didactic AI behavior as 'teachery' and detrimental to collaboration.
- While AI peers reduced task completion time, they increased perceived physical/operational demand and lowered students' social cohesion and task coordination compared to human-only groups.
- Participants strongly preferred a 'scaffold-first' approach, favoring conditional, progressive hints and integrated tools (e.g., Desmos) over immediate full answers, which they considered undermining to learning.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.