AI-Powered Coding of Elementary Students’ Small-Group Discussions about Text

Presenter:  Carla Firetto

What new possibilities open up when researchers use AI to support large-scale qualitative analysis? In this study, we used AILYZE to identify argumentation in 371 student discussions. Using theory-informed prompts and with human oversight, the AI produced codes comparable to humans and had sufficient evidence of validity. Our approach shows how AI can extend, rather than replace, researcher expertise, enabling rigorous, scalable analysis that would have been difficult to achieve with human coding alone.