Teaching Prompt-Based Programming with LLMs: A 45-Minute Lesson with Guided Practice for End-User Programmers

Tran, Marwan & Price (2026) — cs.CY 📄 Full text (arXiv)

This study by Tran, Marwan & Price (2026) introduces and evaluates a 45-minute structured lesson on prompt-based programming, a new modality enabled by LLMs where users express computational goals through natural language. The lesson design incorporates guided practice principles and targets end-user programmers with limited formal training. Results show significant pre-to-post gains in prompt quality, code correctness, and self-efficacy, supporting the case that ai-literacy interventions need not be lengthy to be effective. The work connects to broader conversations about how llm tools change the skills required for programming — shifting emphasis from syntax to prompt engineering. By focusing on end-user programmers rather than CS students, the study expands the scope of stem-education research to include lifelong and professional learning contexts. It also raises questions about the teacher-role in an AI-mediated classroom, where instructors must now teach prompt design alongside or instead of traditional coding concepts. Future work could explore how such lessons scale across higher-ed institutions and whether the gains translate to real-world programming tasks.

Related Pages

• prompt-problems-nl-programming-mistakes — Understanding Student Perceptions, Mistakes, and Debugging Approaches when Solving Natural Language Programming Tasks

Citation

• voice-text-prompt-problems-computing-education — Modality choice in prompt construction