Using automated assessment tools to analyze in-depth student learning of programming concepts

This dissertation contributes to a growing body of research on computer science pedagogy by investigating novice student learning in connection with automated assessment tools (AATs). Analyzing original empirical research alongside existing studies will engage field-wide assumptions—extending some and challenging others—particularly with regard to comparing instructor and student perceptions to quantitative data in evaluating AATs, studying student behavior in iterative programming modifications, the effectiveness of enhancing compiler messages, and chronological ordering of student programs. This culminates in specific assessment-driven improvements in AATs and their use in programming classrooms, along with recommendations for future research and applications.