While generative AI like ChatGPT is often viewed by students as a more knowledgeable tutor, a recent study flipped this dynamic by allowing sixth-grade math students in Florida to teach the AI instead. The experiment was co-authored by Old Dominion University professor Dr. Jinhee Kim as part of her research into human-AI interaction in the classroom.
“Usually, we think about AI as a tutor guiding students’ learning process,” said Dr. Kim, a professor in the Instructional Design and Technology program. “But I wanted to try something else, so I collaborated with the University of Florida to design AI that the students could teach — a less-knowledgeable peer.”
For the study, Dr. Kim and her colleagues designed a “teachable agent” that asked students for help with math problems. Observers noted that the students were highly engaged and excited by the tool. In subsequent surveys, students reported that basic concepts became clearer as they explained them to the AI. They also felt a greater sense of responsibility and ownership over their learning when they were in charge of teaching an AI that asked for detailed, non-intimidating guidance.
This project is one facet of a broader research agenda for Dr. Kim and graduate assistant Kay “Rita” Detrick. Their work includes creating an emotional support AI for students, measuring physiological responses during AI interactions, and evaluating how academic performance and AI literacy affect student engagement with the technology.
A central theme of their research is the critical need to teach AI literacy alongside the technology itself. Detrick, who studies educational psychology, noted that the current application of AI in schools is often reactionary. “A lot of educators don’t have and aren’t getting training on how to use it or how to integrate it,” she said.
One recent project highlighted this need. The researchers evaluated graduate students for whom English is a second language as they used a custom AI for academic writing. While the students had high expectations for the AI’s ability to analyze sources and generate content, they also recognized their own crucial role in applying logical reasoning and questioning the AI’s assumptions and conclusions.
“We, humans, are going to be the leading users of AI, not have AI leading us,” Dr. Kim emphasized. “The user should be critical enough to know when and how to utilize AI.”
Detrick added that the key is whether a student “thoughtfully considers, critiques, and revises” AI-generated content, rather than simply copying and pasting it.
To help build this literacy, Dr. Kim is developing a virtual classroom where AI can simulate classroom scenarios to help train future teachers on instruction and management skills. Detrick believes this points to a future where AI tools are co-created by a collaborative ecosystem of developers, educators, and students.
Ultimately, the researchers argue that educational goals must guide technological innovation. “When it comes to AI in education,” Dr. Kim concluded, “I think we really need to have a good balance and interplay between technology and the noble goals of education.”
Source link
