Date
AI Overview
Why it matters
Rising enrollment and shortages of instructors and teaching assistants limit individualized student feedback in programming courses, while existing AI tools typically provide direct answers rather than step-by-step guidance.
Study at a glance
- Question: Can an AI agent provide step-by-step, context-aware tutoring guidance comparable to human tutors for programming assignments?
- Setting and data: Four college-level Java programming assignments evaluated through simulated student tutoring by the development team.
- Method: System development and prototype testing across four error types using simulated student prompts.
- AI tools: LeafTutor
Key findings
- LeafTutor provided targeted, step-by-step guidance and successfully diagnosed conceptual, structural, and implementation errors across three structured programming assignments.
- The tutor used instructor-provided assignment materials and lecture notes to generate grounded, context-aware explanations rather than generic advice or direct solutions.
- The agent struggled to pinpoint exact faulty code expressions for runtime errors when detailed assignment instructions were unavailable.
Summary drafted with AI from the full paper and may contain errors. See the full paper for details.
High enrollment in STEM-related degree programs has created increasing demand for scalable tutoring support, as universities experience a shortage of qualified instructors and teaching assistants (TAs). To address this challenge, LeafTutor, an AI tutoring agent powered by large language models (LLMs), was developed to provide step-by-step guidance for students. LeafTutor was evaluated through real programming assignments. The results indicate that the system can deliver step-by-step programming guidance comparable to human tutors. This work demonstrates the potential of LLM-driven tutoring solutions to enhance and personalize learning in STEM education. If any reader is interested in collaboration with our team to improve or test LeafTutor, please contact Pu Tian (pu.tian@stockton.edu) or Yalong Wu (wuy@uhcl.edu).
Research Categories
What is the application?
Who is the user?
Who age?
Why use AI?
Study design
Methods
AI tools
Publication Source
Publisher
arXiv
