This project aims to develop and continually improve Coding Buddy, a Generative AI-enabled tool, to support students in introductory programming courses and test Coding Buddy’s impact on student learning.
Introductory programming courses are often challenging for beginners, as they need to grasp new concepts, learn syntax, and develop problem-solving skills. Traditional teaching methods may fall short of providing the immediate, personalized support that students need. Large Language Models (LLMs), one type of Generative AI (GenAI) technologies, have the ability to understand and generate human-like text, making them a promising tool for providing personalized support for student learning. For instance, ChatGPT has demonstrated capabilities in answering student questions, explaining programming code, and generating programming exercises. However, off-the-shelf AI tools like ChatGPT lack in pedagogical strategies and may have side-effects for beginning programmers. By developing Coding Buddy that is guided by relevant pedagogical strategies, this study aims to explore the extent to which GenAI can support learning in introductory programming and assess its impact on student learning outcomes.