Beyond the Veneer of Competence: How Structured Prompting Fosters Knowledge Transfer in LLM-Augmented Education
Abstract
The integration of Large Language Models into technical education exposes a critical fault line between building genuine conceptual mastery and merely outsourcing cognition. We examined this tension by tracking 123 computer science students across three AI-supported assessments and one unassisted final exam, comparing a standard control group against a cohort trained in structured Prompt Engineering (PE).
Fields of Study & Subjects
Intelligent Tutoring Systems and Adaptive Learning
Teaching and Learning Programming
Text Readability and Simplification
Publication details
- Journal / source
- Proceedings of the 14th International Conference on Advances in Information Technology
- Publisher
- ACM
- Pages
- 1–7
- Publication date
- 15 June 2026