Mentors: Zhe "Jay" Shan, Ph.D., Chad Anderson, Ph.D.

Generative AI is increasingly integrated into programming education, providing real-time feedback and personalized guidance for novice programmers. This study examines its impact on learning to code through quantitative text mining of online programming communities and in-depth interviews with computer science students and recent graduates (N=25) at a U.S. college. Topic modeling and thematic analysis were used to identify key concerns and benefits. Findings indicate that while AI tools enhance debugging efficiency and conceptual understanding, they also introduce concerns about over-reliance, academic integrity, and misinformation. Participants found AI most effective when used alongside traditional learning methods that promote critical thinking. This study outlines strategies for responsibly incorporating GenAI into programming curricula, emphasizing structured AI integration and a hybrid approach that balances AI-driven assistance with hands-on problem-solving to ensure deep learning and long-term skill development.

Explore the Project