The Pulse of Higher Ed

The New Assessment Challenge in Online Learning

Generative AI is complicating how we teach working adults in online and professional continuing education. I recently asked a group of educators who teach online about their biggest concerns regarding AI and assessment of student learning. They had many! These educators were worried about: 

• enforcing academic integrity when AI use by students is hard to detect,
• treating students fairly when some are using AI and some are not,
• trusting students to do their “own best work,”
• students failing to develop critical thinking skills because of AI overuse,
• and helping students to use AI effectively and ethically.
  

Do these concerns resonate with you? Have you heard them expressed by colleagues who teach online?

Students, Shortcuts, and the Conditions We Create

There are no quick fixes for the disruptions that generative AI has caused in online education, but faculty and instructional technologists do have agency in how they respond to these disruptions. As James Lang pointed out in his 2013 book Cheating Lessons: Learning from Academic Dishonesty, most students who turn to various forms of unauthorized aid do so because of environmental factors, not a lack of morals. When there’s an emphasis on performance instead of mastery, when there are high stakes riding on an assessment outcome, when students have extrinsic motivations for and low expectations of success, that’s when they’re most likely to lean too much on AI or other shortcuts.

Designing for Mastery, Not Just Performance

The good news is that we can design against these factors!

• We can be more transparent about the purposes of our courses and assignments, so students see more value in the knowledge and skills they’ll gain. 
• We can design more authentic assessments, drawn from real-world professional contexts that help students develop durable skills like communication and collaboration. 
• We can lower the stakes in our courses through alternative grading practices that prioritize feedback over grades and progress over perfection.
• And even in online courses we can create more embodied experiences for students, like this clever approach to teaching calculus online. 

From AI Anxiety to AI-Aware Teaching

None of these strategies alone will “solve” the problem of AI and assessment in online courses, but together they create conditions that motivate students to do the hard work of learning. We need students to engage in robust assignments in our online courses—the assignments that provide students opportunities for deep learning and provide us with meaningful ways to assess that learning. We can’t let AI take these assignments away from us! And we don’t have to, if we can be AI-aware in the design of these assignments. 

Practical Strategies for More Trustworthy Assessment

What does that look like in practice? My co-authors Annette Vee and Marc Watkins and I share concrete examples in our new book, The Norton Guide to AI-Aware Teaching, published this summer with W.W. Norton. While not written exclusively for online educators, the book offers practical strategies for instructors who teach online to respond to the challenges and opportunities that generative AI presents in their teaching. We argue that being “AI-aware” means:

• being clear on course learning goals and objectives,
• understanding something about how AI works, and
• exploring what students think about AI and how they’re using AI tools.

When we’ve done that, we’re in an excellent position to decide how to adapt our assignments and assessments to work in a world where students have ready access to AI. 

The book is full of examples of AI-aware instructional practices, illustrating the “carrot” approaches I mentioned above but also the “sticks” of more secure online assessments—providing students clear boundary lines for their AI use, implementing transparency measures like process tracking and oral exams, using AI detectors thoughtfully and appropriately, and building any interventions with students on a foundation of trust. 

We cover a lot of ground in the book, and no matter what approach an instructor takes to AI, from resistance or restriction to exploration or engagement, I think they’ll find examples of AI-aware guidelines and assignments they can easily adapt to their particular teaching context.

Keeping Opportunity Open for Adult Learners

Online education has opened up so many doors of opportunity for students in online and professional continuing education programs. We don’t want to close those doors because of generative AI. Keeping them open, however, will require changes to what and how we teach. We can learn from and with each other as we figure out those changes together, and I hope the new book will be a part of that collaborative work. Find out how to order the Norton Guide to AI-Aware Teaching here.

 

Derek Bruff is an educator, author, and higher education consultant. He is currently associate director at the Center for Teaching Excellence at the University of Virginia and host of the UPCEA-sponsored podcast, Intentional Teaching, as well as a Strategic Advisor for UPCEA Research and Consulting. To learn more about UPCEA Research and Consulting, please contact [email protected].