Hello, and welcome to a new semester! We typically run a Digital Education reading group. Perhaps someone will organize one for us this semester. In the meantime, this looks like a good opportunity to keep up with some relevant research. Note that a lot of
the people on the invite list are people that our research group (broadly speaking) collaborates with a lot – people at NCSU, CMU, UPitt, in particular. So they are good faces to put with names that you might hear about.
Dear All,
I hope you have all had a great summer.
We will have our first reading meeting for the Student Modelling and Personalization in Computer Science Group in the 24-25 academic year and our first reading meeting for the Splice LLM working group jointly held this
Friday, August 30th, at 10:00 am EST through this
Zoom
link. Please find the title and abstract of this week's presentation given by
Dr. Jaromir Savelka from Carnegie Mellon University below.
Title: Overview of Computing Education Research With Large Language Models
Description: The talk focuses on the impact and implications of large language models (LLMs) like ChatGPT on programming education. It explores how LLMs can revolutionize teaching and learning by interpreting and generating both
natural language and code, offering personalized support, and automating the creation of educational materials. However, it also highlights challenges, such as the risk of incorrect outputs, over-reliance by students, and the potential for reinforcing existing
biases. The presentation outlines ongoing research on the capabilities of LLMs in education, their impact on learning outcomes, and future challenges, including designing introductory programming courses that effectively incorporate LLMs while mitigating their
drawbacks.
I am looking forward to meeting you at the meeting.
Best,
Bita
--
Bita Akram, PhD
Assistant Professor
Department of Computer Science
North Carolina State University