How Students Are Using AI-Powered Hints in Programming Courses | The Research Blog

In a study conducted by JetBrains Academy, researchers explored how students interact with an AI-powered hints tool designed to aid in learning coding within programming courses. The tool, integrated into JetBrains Academy courses, provides personalized hints, leveraging static analysis and large language models (LLMs), to guide students towards solutions rather than simply pointing out errors. The study utilized process mining to analyze quantitative data from students' interactions and conducted interviews for qualitative insights, revealing distinct behavioral patterns. Students engaged with hints in varied ways: some selectively used hints to enhance personal understanding, while others combined multiple partial hints to construct solutions, indicating proactive problem-solving approaches. The research highlighted the importance of understanding student interactions with hint systems and suggested improvements for better support in adapting hints to individual learning strategies. The findings underscore the effectiveness of AI-powered tools in fostering critical thinking and adaptability in coding education.