Semester

Spring

Date of Graduation

2026

Document Type

Thesis (Campus Access)

Degree Type

MS

College

Statler College of Engineering and Mineral Resources

Department

Lane Department of Computer Science and Electrical Engineering

Committee Chair

Katerina Goseva-Popstojanova

Committee Co-Chair

Thomas Devine

Committee Member

Matthew Valenti

Abstract

Generative AI is rapidly transforming programming both as a tool for learning and as a system capable of producing code. This thesis examines that transformation from two complementary perspectives. First, it investigates how generative AI influences introductory programming education through a survey of 139 students across two semesters, analyzing patterns of use, perceived bene f its, motivation, self-efficacy, academic performance, and pedagogical implications using descriptive statistics, Mann Whitney tests, Spearman correlation, and logistic regression. Second, it evaluates the quality of Python and Java code generated by modern large language models on 1,651 programming tasks by analyzing the lines of code, cyclomatic complexity, code smells, failure proneness and vulnerability presence through descriptive analysis and inferential statistics. Across both studies, the findings show that generative AI cannot be adequately understood through surface-level outcomes alone. In education, it offers fast, accessible assistance but may reduce engagement and self-confidence under some conditions. In code generation, benchmark success does not fully capture software quality or security, and structural characteristics such as code smells, code length, and task difficulty help explain failure proneness and vulnerability presence.

Download

Share

COinS