Use structured think-aloud sessions with a semi-structured moderator guide to capture student feedback for iterative AI tool refinement¶

Strategy · All strategies
Evidence · no claims cited

Description¶

The strategy runs think-aloud sessions in which students verbalize thoughts and expectations while using an AI tool in real time, guided by a semi-structured moderator's guide covering preparation, session flow, and post-session reflection. Moderators use minimal prompts such as "What do you think?" and "What do you expect to happen next?" to capture authentic feedback. Sessions in the case study included a 5-10 minute introduction, a 20-30 minute main task, and a 10-15 minute post-task reflection, with insights documented for subsequent development cycles.

Design Implications¶

Context¶

Requirements¶

• A structured moderator guide to ensure consistency across sessions, prepared moderators familiar with the tool, and recording arrangements

Constraints¶

• The article notes concurrent verbal reporting can disrupt the natural flow of thought, and that maintaining a second recording device added complexity for students in this study

Target Learners¶

• middle school students
• K-12 learners

Target Learning Goals¶

• usability evaluation of AI learning tools
• student voice in tool refinement

Affordances¶

Related Strategies¶

• Write Alouds
• Self Talk Logging

Examples¶

-

Key Sources¶

• Tzirides, Anastasia Olga (Olnancy), Galla, Michele, Cope, Bill, & Kalantzis, Mary. (2025). Thinking Through AI: Advancing Cognitive and Collaborative Research for AI in Education. https://doi.org/10.35542/osf.io/s8hqe_v1