Discourage academically dishonest use of AI in online courses

Teaching practices that help students connect

Now that we understand that forging relationships with students and connecting them to your course is the best approach, we can turn to eight practices that both discourage cheating and also help students connect more fully with their own learning. You will notice that these practices do involve some effort on your part, but communicating about the value and plan of your course and your assessments will pay off in the long run.

1. Use more frequent, low-stakes assessments instead of high-stakes assessments.

When students believe that a single exam or assignment will determine their success or failure in a course, they may decide it is safer to cheat using AI than to risk a low grade. Instead of one or two assessments being large determinants of students’ grades, you should consider two alternatives. 

First, consider smaller, more frequent assessments that carry less weight in students’ final average. This works best if assessments are iterative and help students practice similar analytic or thinking skills again and again. Second, and perhaps even more effectively, consider small assignments that build toward a larger project or paper: students turn in pieces of work throughout the semester and revise those pieces using instructor and/or peer feedback to create a final, stronger version.

With both these approaches, students perceive individual assignments as lower-stakes, feel supported by the instructor, and see assessment as a way to learn rather than a performance that might sink their grade. This plan for their learning should be communicated to students. Another benefit to these approaches is that the instructor will learn more about the students by seeing their writing and thinking develop across essays; this personalizing of assignments also minimizes cheating. Finally, instructors can provide structure for students to use feedback in ways that require them to integrate new ideas into the next step of their work. Assignments in these course designs are personal, meaningful, and designed to help students, not threaten them.

Examples of effective approaches to assessments

2. Help students see the value of assessments.

When students see assessments and assignments as just a hoop to jump through, they don’t see the need to engage in the work those assessments require. It is important to share with students how you believe that doing their own work on the assessment will benefit them. It is even more important that you then give students time to generate in writing their own ideas about how particular assignments will help them develop and learn. Optimally, this writing should be saved as part of students’ work on their assignments so that students (or you) can return to these initial ideas as they complete or make progress on their writing assignment. It can be helpful to structure this reflection so that students make connections between the writing they’re doing and their academic and non-academic worlds. Students can share their ideas with each other and with you by posting on discussion boards.

Examples of communicating the value of assessments

3. Help students understand that you’ve designed your course to help them learn and develop.

As you design course activities that build toward the essay you want them to write or the test you want them to take, make sure that students have multiple opportunities to practice using the skills that assessment requires and that you explicitly communicate the value of that practice to them. Before students are scheduled to complete a course assessment, create an announcement in Brightspace to remind them that they have spent time practicing what they need to succeed on the paper.

It’s wise to communicate the preparatory steps that you’ve planned as an instructor in your syllabus, but you can also point out to students throughout the course how the work that they are doing is iterative and builds toward the work of your assessments.

Examples of communicating iterative work

4. Help students prepare—and feel prepared—for assessments.

The research on cheating is clear: students cheat when they feel unprepared for assessments. The best way to work against this feeling is to structure self-regulatory preparatory work right into your class. Students will feel more prepared (and be more prepared!) when they make a plan for completing online assessments. 

It is important to note that not all students know that they should do this, and some students who understand the value of planning their work don’t know how to do it. You can help students respond more effectively to assignments by having them take some time when they receive an assignment to read it carefully, reflect on its value, and make a plan for how they will approach it. This work should be done in writing as a homework assignment that students submit to you.

Examples of prompts that can guide students to plan in helpful ways

Read the assignment description for the upcoming paper carefully. Then respond to these prompts and submit your responses to me in the paper planning assignment: 

• In what ways will this assignment draw on my strengths? How will I use these strengths to be successful?
• What do I expect will be most challenging as I complete this assignment? What resources (e.g., teacher, peers, course materials, outside resources, etc.) will I use to help me work through this challenge?
• What are two things I will do this week (by [date]) to begin working on this assignment?
• If I am not able to complete this assignment on time, what options will I discuss with my instructor (choosing a late submission date, choosing to submit part of the assignment on time, other options that work for your course)?

This kind of analysis and planning helps ensure that students read assignments more carefully, and it will also surface questions early on so that you can more effectively guide students through their work. Have students write in response to these prompts in an assignment where both you and they can read and respond to their initial ideas. You can provide feedback—either global or individual—that guides and supports students.

5. Help students work productively with your feedback on assessments.

Students don’t always realize that they can change their approach to their work and improve the outcome. When students receive a lower-than-expected grade on the first essay or test in your course, they might panic and decide that they will cheat on the next assessment. Help these students by having them analyze their efforts and the feedback you provide so they can be more successful. After students have received feedback on an assessment or assignment, have them respond in writing to prompts that help them use your feedback to analyze their efforts and make plans for future work. It is helpful to have students submit these responses in an assignment where they can continue to reflect on their progress.

Examples of prompts students can use to make the most of your feedback and feel hopeful about their next assessment

• How long did I spend preparing for this assignment? Was that enough? Do I need to spend more time on the next assignment?
• What did I do while I worked on this assignment? Did I (provide students with a list of helpful strategies here)?
• Given the feedback I received, what are two new strategies (from the list above or others) that I will try when I work on my next assignment? 

Students can analyze their work and adjust their strategies throughout the semester. Toward the end of the course, have students articulate their discoveries about the most effective ways to work, study, and prepare for online assignments. Ask students to describe how they will continue to use these discoveries throughout their college career and in their lives outside of school.

6. Allow students to focus course assignments on their real interests.

When students see completing assessments as a path toward their own goals or interests, they are simply much less likely to cheat. In addition, they are much more likely to dig deeper, work with greater focus, and explore their own thinking more fully. 

If your course design involves a paper or project that develops over the course of the semester, have students choose a topic of interest to them and explore why and how they will learn and benefit from that focus at the outset of the semester. Alternately, if your course design involves a series of writing assignments that do not build toward a final paper, students can also help finalize the focus of those papers or you can design those papers so that there is a personal investment for students. Finally, if your assessment is an online exam, students can append to their exam a description of other things they’ve learned that weren’t tapped by your items. In addition, exam items can be developed with students’ interests and values in mind.

Examples of personalized assignments

7. Create assessments that are specific to your course and your discipline.

A key practice that makes it more difficult for students to use others’ work in lieu of their own is creating assessments that are very specific to your discipline, your course, and the goals that you have for student learning. Answers to and examples of such assessments can’t be easily be found online or generated by AI. Inform students that the paper assignment you’ve created is unique not just to your course, but to this iteration of your course. This is another opportunity to frame the assessment as a meaningful and manageable measure of their thinking and learning.

Examples of assessments that are specific to your course and your discipline

8. Have students analyze and evaluate writing generated by AI and make critical decisions about their own writing.

Academic writing is a way to tackle problems in our disciplines; every text we ask students to craft should be part of the disciplinary thinking our courses aim them toward. When the writing we assign is meaningful to students and is part of a larger plan to help them develop their writing and thinking skills, they are less likely to cheat. 

We can also strengthen their resolve to resist shortcuts like generative AI by having them thoughtfully evaluate the writing it generates. Students can compare outlines, paragraphs, or short essays generated by generative AI in relation to your assignment rubrics or to previous students’ strong work. If you have students compare examples, don’t initially reveal which example was written by a student and which was produced by generative AI. Students will be able to explore the differences between these examples and note critical weaknesses in the AI example. Alternately, AI produced outlines and writing could be critically analyzed by students and they can annotate, edit, or revise the text. 

As you consider how you can help your students become critical of using generative AI to produce writing, be sure your assignments are still aligned with the learning goals that you have for your students. A word of caution here: while some instructors rush to embrace generative AI as a teaching tool, it is unproductive to structure activities that are unrelated to the disciplinary thinking you want students to practice.

After students share their ideas with their peers and have concretized what is strong, effective writing and what is not, you can reveal that they’ve been analyzing AI-produced writing. The next and perhaps most important step is to have students individually write down the specific weaknesses they see in this writing, what disciplinary thinking AI can’t do, and the value of doing their own writing and thinking. Students can identify specific writing and thinking moves that they plan to make on their next writing assignment. When students turn in that next assignment, you can ask them to point out one or two key writing or thinking moves they made that they feel has made their writing particularly effective.

Examples of analytic prompts to guide students in an evaluation of AI-generated text

Resources

• Aaron, L., Abbate, S., Allain, N. M., Almas, B., Fallon, B., Gavin, D., Gordon, C. B., Jadamec, M., Merlino, A., Pierie, L., Solano, G., and Wolf, D. (2024). Optimizing AI in higher education. 2nd ed. SUNY Press. https://sunypress.edu/Books/O/Optimizing-AI-in-Higher-Education
• Bowen, J. A. and Watson, C. E. (2024). Teaching with AI: A practical guide to a new era of human learning. Johns Hopkins University Press.
• Lang, J. M. (2013). Cheating lessons: Learning from academic dishonesty. Harvard University Press.