Automated Program Repair of Uncompilable Student Code
This work addresses the issue of discarded learning data in student modeling for CS1 education, enabling more comprehensive analysis of coding processes.
The study tackled the problem of uncompilable student code in CS1 courses by using large language models (LLMs) like GPT-5, Claude 3.5 Haiku, and Gemini 2.5 Flash for automated program repair, finding that all models could produce compilable repairs but varied in preserving students' control flow and structure.
A significant portion of student programming submissions in CS1 learning environments are uncompilable, limiting their use in student modeling and downstream knowledge tracing. Traditional modeling pipelines often exclude these cases, discarding observations of student learning. This study investigates automated program repair as a strategy to recover uncompilable code while preserving students' structural intent for use in student modeling. Within this framework, we assess large language models (LLMs) as repair agents, including GPT-5 (OpenAI), Claude 3.5 Haiku (Anthropic), and Gemini 2.5 Flash (Google), under high- and low-context prompting conditions. Repairs were evaluated for compilability, edit distance, and preservation of students' original structure and logic. We find that while all three LLMs are capable of producing compilable repairs, their behavior diverges in how well they preserve students' control flow and code structure, which affects their pedagogical utility. By recovering uncompilable submissions, this work enables richer and more comprehensive analyses of learners' coding processes and development over time.