Design hybrid instructional chatbot systems that balance breadth and depth via multi-agent architectures or adaptive interaction modes¶

Strategy · All strategies
Evidence · no claims cited

Description¶

The authors recommend treating the depth-efficiency trade-off not as a limitation but as motivation for hybrid systems. They suggest "a multi-agent architecture, in which agents are specialized for complementary roles" — one for rapid idea generation and broad strategy exploration, another for structured guidance and reflective questioning — or a single agent with adaptive modes, explicit control options, or progressive disclosure so instructors can align system behavior with their immediate goals.

Design Implications¶

Context¶

Requirements¶

• Ability to detect or let users select when concise breadth-oriented suggestions versus deeper reflective guidance is appropriate

Constraints¶

• The study sample comprised self-selected higher education instructors, which may limit generalizability to K-12 settings, informal learning environments, or institutions with different infrastructures

Target Learners¶

• Higher education instructors
• Instructors in K-12, professional training, coaching, and tutoring settings (proposed extension)

Target Learning Goals¶

• Balancing rapid access to teaching ideas with reflective instructional development

Related Strategies¶

• Adaptive-resolution simulation: allocate more LLM calls to educationally consequential moments and compress routine transitions
• Enhance student development agents with finetuned or small task-specific models and richer memory architectures rather than prompting alone

Examples¶

-

Key Sources¶

• Isabel Molnar, Peiyu Li, Si Chen, Sugana Chawla, James Lang, Ronald Metoyer, Ting Hua, Nitesh V. Chawla. (2026). TeachingCoach: A Fine-Tuned Scaffolding Chatbot for Instructional Guidance to Instructors. https://arxiv.org/abs/2603.18189