A Tailored Generative Artificial Intelligence Chatbot for Reflective Learning in Introductory Programming

Abstract

The rapid development of Generative Artificial Intelligence (GAI) chatbots is affecting higher education teaching and learning, especially in introductory programming courses. Introductory programming courses are considered difficult with high failure rates. Introductory programming is often structured or adopts problem-based learning (PBL), with a focus on solving programming problems. GAI chatbots can generate the code that students typically learn in introductory programming courses, allowing students to get solutions for their programming assignments. Students and lecturers have concerns that using GAI chatbots in learning leads to plagiarism, over-reliance, and a reduced learning process. To overcome the challenge of GAI, educators must foster students’ essential learning skills such as Critical Thinking, Problem-Solving and particularly Reflective Thinking. Reflective Thinking enables students to navigate the vast expanse of current knowledge and information by constructing their own body of knowledge, which supports problem-solving. Developing students’ Reflective Thinking skills requires personalised support and feedback, a feature that GAI chatbots can assist with. This research utilised a tailored GAI Chatbot, Buddy, as a learning tool to support students’ Reflective Learning in introductory programming course in a controlled environment. Observation data suggests participants who are using Buddy demonstrated an increased in tendency to seek explanation and generate hypothesis behaviours, which associated with increased Reflective Thinking behaviours. Therefore, this study provides evidence that a tailored GAI chatbot can facilitate Reflective Dialogue to trigger Reflective Thinking in learning.