Leveraging AI Chatbots to Enhance Student Understanding of Electric Circuits

Abstract

This work-in-progress (WIP) examines the potential of AI-powered chatbots, especially ChatGPT, as instructional tools to support student learning in an introductory electric circuits course. AI chatbots are becoming increasingly prominent in educational settings due to their capabilities in providing immediate, natural language-based explanations that support both procedural and conceptual understanding. In this WIP, electrical engineering students engaged with ChatGPT to solve problems related to fundamental topics such as Voltage and Current Division, as well as Nodal Analysis. Through structured assignments, students compared traditional problem-solving methods with chatbot-assisted approaches. Key findings indicate mixed outcomes, with students demonstrating improved comprehension of conceptual topics but facing challenges in accuracy due to prompting errors and limitations in ChatGPT's analytical processing. On average, 82 % of students expressed positive feedback on ChatGPT, with 48 % reporting improved confidence and understanding in electric circuits. However, for Nodal Analysis, only 22 % of students who provided accurate prompts received correct solutions from ChatGPT that closely matched their hand calculations.

Ähnliche Arbeiten