Kexin Bella Yang, Menghan Liu, Liyi Xu, Nikol Rummel, Vincent Aleven (2026) — Carnegie Mellon University; University of Washington; Ruhr-Universität Bochum. arXiv:2605.18761 [cs.HC] — Accepted at CSCW 2026, to appear in PACM HCI
Yang et al. (2026) tackle a fundamental tension in AI-augmented classrooms: how to balance teacher orchestration with student agency during dynamic transitions between individual and collaborative work. Using participatory speed dating with teachers and students, the study maps a three-stage design space (before, during, and after pairing) and proposes a hybrid-control framework for analytic-based orchestration tools.
The core recommendation — structured teacher guidance initially, with progressively increasing student autonomy — provides an actionable design principle for intelligent-tutoring systems that manage classroom-level coordination. This graduated autonomy model connects to scaffolding theory and extends it from individual learning support to social-organizational support. The work also contributes to the teacher-role literature by formalizing how human-AI co-orchestration can distribute agency without undermining either party's effectiveness.
By situating the research within real K-12 classroom dynamics and publishing at CSCW, the paper bridges the collaborative-learning and human-in-the-loop communities that have often addressed these questions separately. The three-stage framework (before/during/after) offers a structured lens for evaluating other orchestration tools, complementing the ai-tutor-behavioral-evaluation focus on individual tutor performance with classroom-level coordination metrics. For k-12 practitioners, the study validates concerns about AI systems that optimize for efficiency at the cost of student self-regulated-learning and teacher professional judgment.
Related Pages
- expert-cognition-dashboard — Provides cognitive infrastructure for AI Twins to support adaptive intervention
- intelligent-tutoring — related
- teacher-role — related
- student-experience — related
- collaborative-learning — related
- human-in-the-loop — related
- scaffolding — related
- ai-tutor-behavioral-evaluation — related
- k-12 — related
- self-regulated-learning — related
- ai-k12-evidence-base — related
Citation
APA: Kexin Bella Yang, Menghan Liu, Liyi Xu, Nikol Rummel, Vincent Aleven (2026). Balancing Teacher and Student Agency: Co-Orchestration Tool Design Supporting Real-Time Dynamic Pairing. arXiv:2605.18761. arXiv:2605.18761 [cs.HC] — Accepted at CSCW 2026, to appear in PACM HCI.