Treffer: Student Perceptions of Learning through Original and AI-Generated Python Programs from a Software Quality Perspective

This study explores student perceptions of learning to code by evaluating AI-generated Python code. In an experimental exercise given to students in an introductory Python course at a business university, students wrote their own solutions to a Python program and then compared their solutions with AI-generated code. They evaluated both solutions using a software quality assessment framework, focusing on the correctness, efficiency, understandability, consistency, and maintainability, which provided a guide to evaluating code beyond simply correctness of the solution. Research examines how students perceive and utilize generative AI, considering their motivations, outcomes, and experiences. Findings suggest that while students see significant potential in using AI tools to enhance their coding process and appreciate the efficiency and compactness of the AI-generated code, they often prefer their own solutions due to familiarity and features used. This research aims to inform future studies on student application of AI tools in learning to code and provides educators with a model for evaluating AI's impact on student learning.

As Provided