Date
AI Overview
Why it matters
Understanding how students engage with generative AI during authentic learning activities remains limited, as existing studies rely heavily on controlled settings, single domains, or small-scale qualitative data.
Study at a glance
- Question: How do undergraduate students interact with generative AI tools during authentic coursework across different academic courses?
- Setting and data: Technion – Israel Institute of Technology, 821 undergraduate students across 6 courses, 2,078 chat logs, and 15,887 interaction turns.
- Method: Instruction-guided LLM annotation and descriptive analysis of authentic chat logs.
- AI tools: ChatGPT, gpt-5-mini
Key findings
- Student interactions concentrated in recurring patterns, with roughly half directed toward the task or domain, about 30% toward AI output, and about 20% toward students' own work.
- Most student requests occurred at the understand and apply cognitive levels, while evaluate and create formed a substantial minority and analyze appeared rarely.
- Interaction patterns differed across courses: mathematically oriented courses emphasized task-focused problem solving, whereas writing-oriented courses featured more references to students' own work and higher-level cognitive activity.
Summary drafted with AI from the full paper and may contain errors. See the full paper for details.
Generative AI tools (GenAI) are increasingly used by students during coursework, yet empirical understanding of how students engage with these systems in authentic learning contexts remains limited. Existing studies have largely relied on controlled settings, single-domain analyses, or small-scale qualitative data, leaving open how student-AI interaction unfolds across courses and forms of academic work.
We present a large-scale analysis of naturally occurring student-AI interactions collected from undergraduate students across multiple university courses and academic domains. The dataset comprises over 15,000 student-AI interaction units drawn from voluntary use of generative AI during real coursework.
To characterize these interactions, we analyze each student turn along two complementary dimensions, cognitive intent and interaction context, capturing whether requests are directed toward the task or domain, the student's own work, or prior AI output. Using instruction-guided annotation applied at scale, we examine how these interaction patterns are distributed overall and how they vary across courses.
Our analysis reveals that student-AI interaction is highly structured. Across courses, interactions concentrate in a small number of recurring patterns rather than exhibiting highly idiosyncratic use. At the same time, systematic differences emerge across courses, giving rise to distinct interaction profiles associated with different forms of academic work.
Research Categories
What is the application?
Who is the user?
Who age?
AI tools
Publication Source
Publisher
arXiv
