LLMs Integration in Software Engineering Team Projects: Roles, Impact, and a Pedagogical Design Space for AI Tools in Computing Education
Project Overview
The document examines the incorporation of generative AI (GenAI) tools, like ChatGPT and GitHub Copilot, in undergraduate software engineering education, particularly in the context of team-based projects. It highlights students' experiences and the dual impact of these tools on their teamwork, learning, productivity, and self-efficacy. While GenAI has been shown to enhance coding skills and overall productivity, the study also raises concerns about challenges such as over-reliance on AI and the risk of masking students' actual abilities. To address these issues, the document proposes a design framework for GenAI-based educational tools that aims to optimize the benefits of these technologies while minimizing potential drawbacks. Overall, the findings suggest that while generative AI can significantly enrich the educational experience, careful implementation and monitoring are essential to ensure that students develop their skills effectively.
Key Applications
GenAI tools like ChatGPT and GitHub Copilot
Context: Undergraduate software engineering team project module
Implementation: Students were allowed to use GenAI tools for coding tasks within a structured team project over a semester.
Outcomes: Increased confidence, improved teamwork dynamics, enhanced productivity, better learning through scaffolding.
Challenges: Over-reliance on AI, potential for generating buggy code, hiding of actual programming skills.
Implementation Barriers
Implementation Challenge
Students may rely too heavily on GenAI tools, leading to a lack of understanding of fundamental programming concepts.
Proposed Solutions: Encourage foundational learning before using AI tools; incorporate practical usage training and AI literacy in the curriculum.
Assessment Integrity
AI-generated code may not accurately reflect students' true abilities, complicating assessment and evaluation.
Proposed Solutions: Implement transparency measures in AI-generated outputs and ensure students verify AI-produced code.
Project Team
Ahmed Kharrufa
Researcher
Sami Alghamdi
Researcher
Abeer Aziz
Researcher
Christopher Bull
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Ahmed Kharrufa, Sami Alghamdi, Abeer Aziz, Christopher Bull
Source Publication: View Original PaperLink opens in a new window
Project Contact: Dr. Jianhua Yang
LLM Model Version: gpt-4o-mini-2024-07-18
Analysis Provider: Openai