A Mixed-Methods Study on the Impact of AI-Assisted Coding in an Undergraduate Programming Course

다국어 초록 (Multilingual Abstract)

As the use of generative-AI tools continues to expand in programming education, it is increasingly important to examine learners’ perceptions and experiences to support sustainable instructional practices. This study aimed to exploratorily analyze undergraduate students’ perceived learning outcomes and subjective learning experiences within an AI-assisted coding environment. The partici-pants were 37 students enrolled in a programming course offered by the School of Software at H University in G Province during the first semester of 2025. Data were collected through validated questionnaires and reflective open-ended responses and analyzed us-ing paired-sample t-tests, correlation analyses, and inductive thematic analysis. The results indicated that pre–post score differences were reported across learning-related variables after students experienced the AI-assisted coding course, with learning engagement showing a comparatively larger magnitude of change. Qualitative findings further revealed that AI immediate feedback and task support were associated with experiences of psychological flow, while some students expressed concerns regarding the potential emergence of cognitive dependence during AI-supported learning. Despite limitations related to the single-group design and small sample size, this exploratory study provides insight into how AI-assisted coding shapes students’ learning experiences and high-lights cognitive considerations that should be addressed in the instructional design of programming education.