Using AI for Learning in the Community College Classroom
By Nancy Murray
This is the first post in a series on using AI as a learning tool. The second post explains how the authors worked with students to examine whether AI belongs in a community college composition classroom.
“Is it cheating?” he asked. “Would you have failed me if you knew?” It felt like one of those pivotal teaching moments where my answer would be weighed heavily.
The fact that my English 101 student brought it up led me to believe that he was not “cheating” but grappling with an ethical question, one that became a hot topic at the community college just one year later when ChatGPT was offered to the public. Is it cheating to use artificial intelligence to write your essay? As I did not see myself as a moral authority, my student and I presented the question for class discussion.
Most students in my classroom, and I, did not know what ChatGPT was at the time, so the offending student had already demonstrated that he was ahead of the curve, but some of his classmates thought it was “just wrong” without being able to explain why. Others had solid, but unsubstantiated, objections. “You won’t learn if you do it that way,” they said. “We’ll lose our ability to think critically.” I wondered if there was truth to these concerns. I’d heard my colleagues make similar arguments. They seemed determined to “catch” students who dared to try the new technology, but I felt more inclined to be excited about the possibility of catching my students using it to support their learning. I decided to create a teaching unit where we could all learn what the technology was together, and where I could measure the student experience and learning in the process through surveys after each essay, polls before and after assignments, and written reflections at the end of the units. My intention was to compare results between essays written with and without artificial intelligence to determine if students who used ChatGPT were, indeed, learning.
As with most of my teaching, I had a lot to learn before I could start. Fortunately, there were AI articles published every day in the news, academic journals, tech magazines, business magazines, and more. I read them all with interest, but it was clear to me that I would not fully understand until I tried it, so I subscribed to ChatGPT-3.
The first question I asked was the very question I was grappling with. Is it ethical for students to use AI to complete their schoolwork? In less than 5 seconds I got a two-page breakdown of points to consider and the following conclusion: “Using AI to complete schoolwork raises significant ethical concerns related to academic integrity, fairness, and personal growth. While AI can be a valuable tool in education, students should use it responsibly and in conjunction with their own efforts and the guidance of their teachers.”
Fair enough, but if I were going to guide my students to “use it responsibly,” then I needed to know what that looked like. I asked ChatGPT to write a four-week teaching unit for an English 101 composition course that addressed the question. To my astonishment, what might have taken me weeks to complete was done in less than a minute with each week having a different angle to the question, an introduction, and time allotted for lecture, case study, debate, writing assignments, and reflection. I noticed that the unit was not a complete unit but rather a suggested outline for a possible approach. There were no specifics. For example, in the guest lecture section it simply said, “Find a guest who is knowledgeable on the subject to lecture for the class.” For the debates and discussions, it simply said, “Facilitate a debate or a discussion on the topic.” I found that ChatGPT was helpful for generating structure and ideas, but to make the lessons feel dynamic and engaging, and to be sure the information was accurate, I needed to rework my questions and review ChatGPT’s suggestions many times.
I spent two days filling in blanks, which required critical thinking and active reading. I found ChatGPT occasionally suggested readings, with citations, that didn’t actually exist. It didn’t always communicate clearly, either. When I asked it to write a lesson on how to explain an algorithm, it spat out a very confusing activity using a deck of cards that I had to go over repeatedly before seeing the point it was trying to make. Then I rewrote the instructions for clarity. It came to mind that students would have to unpack complicated sentences, too, in order to communicate their ideas using ChatGPT. They would need to engage in critical thinking to develop prompts for the best outcomes and to be sure the information from ChatGPT was credible and accurate.
The unit explores biases that might be programed into the technology and examines where ChatGPT gets the information to formulate its answers. The students’ study of the Predictive Language Model that ChatGPT uses parallels English 101 discussions about syntax, context, sentence structure, audience, and tone. In other words, writing my unit with ChatGPT provided copious opportunities to learn. It occurred to me that teaching with this technology could benefit students who are underprepared in reading and writing skills by providing short readings and real-time interactive guidance throughout the reading and writing process while also developing critical inquiry skills.
When I and my colleague deployed the unit, we didn’t see students eager to cheat, but eager to learn. They were dedicated to their own ideas while still appreciating the reduced anxiety around getting started, the help with reading and communication skills, and the valuable time saved by using ChatGPT. I saw students develop a taste for focused inquiry that enriches every learning experience. Together we realized that learning in the modern world requires the humanities to evolve with technology to foster innovation and growth. Doing so ethically requires open and honest communication about what we need and what we want to teach and learn.
In a companion post, my colleague and I will share survey results about what our students experienced and felt about the work they produced when using ChatGPT.
|Nancy Murray is an associate professor of English at the Community College of Baltimore County, Essex Campus.|