What is it about?
This study analyzes how LLMs can be integrated within curriculums equitably that aid, not hurt, novice Computer Science students’ development. We conducted a study on over 70 students from UCI’s introductory programming classes where they answered questions using iGPT–our developed in-house “interactive” chatbot asking clarifying questions to guide students to the answer rather than giving it to them immediately, simulating interactions between an instructor/TA and students.
Featured Image
Photo by Desola Lanre-Ologun on Unsplash
Why is it important?
Our paper shows significant statistical improvement in average question pass rate using our “interactive” chatbot as opposed to default ChatGPT, indicating the benefits of "interactivity" in LLM-student interactions and the possibility of integrating LLMs into the education space for the benefit of students' overall education.
Perspectives
"Interactive" chatbots or LLMs that simulate the behavior of a teacher show promise in the education field and can be a healthy alternative to the outright ban of AI use in the classroom. A strict prohibition could incentivize undocumented AI usage, despite labelled as academic dishonesty, and exacerbate the harm LLM reliance can do to a student's education. Our approach offers a compromised solution to the issue of AI in education.
Tyler Yu
University of California Irvine
Read the Original
This page is a summary of: Bridging Novice Programmers and LLMs with Interactivity, February 2025, ACM (Association for Computing Machinery),
DOI: 10.1145/3641554.3701867.
You can read the full text:
Contributors
The following have contributed to this page
