BugTor: A Feedback-Driven Static Analysis Tailored for Novice Java Programmers
摘要
A common challenge for novice programmers is interpreting runtime error stack traces and manually identifying logic and style errors in their Java code, which can lead to inefficient debugging and poor coding practices. This study presents BugTor—a static analysis and feedback-generation tool designed to help novice programmers identify and resolve a wide range of common programming errors. Our preliminary evaluation results reveal that BugTor effectively identifies targeted errors and provides formative feedback, helping students pinpoint issues in their code and offering guidance on how to fix them. BugTor also provides students with insights into their performance through weighted error densities compared to submission-wide averages. For instructors, BugTor highlights patterns in common student difficulties, supporting data-driven curriculum improvements.