Mahdis Mousavi, Unsplash
By Doug Ward
A new survey from the Association of American Colleges and Universities emphasizes the challenges instructors face in handling generative artificial intelligence in their classes.
Large percentages of faculty express concern about student overreliance on generative AI, diminishment of student skills, decreased attention spans, and an increase in cheating. The sample for the survey is not representative of all faculty, but it captures many of the concerns I have heard from instructors over the past three years.
There are no clear paths for adapting to the challenges of generative AI, and we have to take a multi-prong approach and experiment as we move forward.
That doesn’t mean we have to start from scratch, though. Taking some steps now, at the beginning of the semester, will make things easier as the semester progresses.
What you can do
In another post, I offered some suggestions to help make generative AI feel less overwhelming. Here are some steps you can take to make class go more smoothly and reduce problems related to academic integrity. Most of these are drawn from approaches that instructors are taking now.
- Work with students on a class policy. You should have a class policy in your syllabus, and CTE offers several approaches to creating a policy. You should explain that policy to students, but you can also encourage students to ask for clarifications or suggest additional language. That can help improve student acceptance of the policy — and improve it.
- Use alternative forms of grading. One of the biggest problems we face with assignments is that students’ quests for grades tend to block out everything else, including learning. CTE resources on alternative grading can help you consider ways to put learning above grades.
- Explain the why of assignments. All too often, we give students work to do without explaining how it will help them, how it connects with other disciplinary work, or how they will use it in the future. Helping students better understand the importance of assignments and the role they play in developing skills can improve motivation.
- Talk about research into generative AI and learning. Research into generative AI and education is still relatively sparse, but one thing is clear: Generative AI can’t learn for us. True learning requires effort, struggle, and occasional failure. Avoiding that work now will create problems later. (Another blog post provides a deeper look into the research on generative AI in teaching and learning.)
- Doing more in-class work. Class time is precious, and we should use it for the things that are most important. Some instructors have found that having students write, code or create in class reduces student desire to use generative AI, especially with low-stakes assignments. In-class work also allows instructors to work with groups or individual students, answering questions and providing guidance.
- Schedule individual meetings. Oral discussions can help instructors gauge student understanding. If students can’t explain their process for writing, coding or creating, they probably haven’t done enough (or any) work. Meetings don’t have to be long or complicated. Sara Wilson, a CTE faculty fellow from mechanical engineering, for instance, meets with students after every assignment, requiring additional work if students can’t explain the process they used in completing programming assignments. She often has more than 100 students, using class time for the individual meetings.
- Have students create a log. Reflection is an important part of learning. Having students create learning logs can promote reflection and allow instructors to see students’ thinking and the approaches they use as they complete work. Again, it is important to explain the purposes and benefits to students.
Much of the problem we have had with generative AI in education comes down to intrinsic motivation and trust. The education system emphasizes grades over learning, and students often see coursework as an obstacle. We need to tap into their intrinsic interests and chip away at the systemic barriers that inflate the importance of grades and turn teaching into policing.
Building trust is a crucial part of that process. That includes using approaches that encourage and reward effort, providing opportunities for questions and discussion, allowing students to learn from failures without grade-destroying penalties, and building a sense of community in each class. All of that requires work from instructors and students, but it also provides long-term benefits.
Doug Ward is associate director of the Center for Teaching Excellence and an associate professor of journalism and mass communications.