Leveraging ChatGPT for Adaptive Learning through Personalized Prompt-based Instruction: A CS1 Education Case Study

Abstract

In this research paper, we discuss our attempt to teach high school students introductory programming with Python using a custom learning platform that leverages ChatGPT to generate personalized learning materials based on each student's educational background. The platform features topics and subtopics, each supported by prompts for Explanation, Example, Exercise, and Exercise Solution, with a context-setting prompt tailored to individual students' backgrounds while respecting their privacy. The case study brought up compelling insights. Students exhibited heightened engagement, and the lecturers transitioned from being traditional instructors teaching content to becoming mentors who guide students on what to do next, clarifying misunderstandings and addressing potential questions. Furthermore, students gained hands-on programming experience during the learning process, eliminating the traditional post-class experimentation phase. This innovative approach not only enhances traditional CS1 education but also suggests a broader application of Large Language Models (LLMs) for personalized learning across diverse fields, providing tailored instruction and fostering engagement.