Cheaters will face new obstacles next year.
Gavan Watson, vice-provost (teaching and learning), told Senate on March 28 that associate deans at the University are currently reviewing the use of artificial intelligence (AI) on high-stakes exams. He didn’t say when the review will be complete.
As of August 2023, the University hasn’t banned generative AI in the classroom across the board—only at instructors’ discretion. Currently, students in classes that permit the use of ChatGPT must cite the material generated by AI. Failing to cite material from generative AI constitutes a departure from academic integrity.
“If we’re concerned about deviations from academic integrity, one clear way we can get around that is by assessing individuals in person, but that doesn’t mean there aren’t technologically enhanced ways to ensure this,” Watson said to Senate.
READ MORE: Artificial Intelligence may have a place in academia
Findings from the review have the potential to increase the number of in-person or remotely protected exams, Watson said.
Final assessments could be adapted to allow ChatGPT, Watson said in an interview with The Journal. However, students aren’t learning anything if they’re using ChatGPT to answer multiple-choice exam questions and may miss the opportunity to demonstrate foundational knowledge they’ve learned in class.
“If there was a foundational or fundamental threat to generative AI, it’s not that students use it. That’s not the threat. It’s that students would use it in a way that would offload cognitive work,” Watson said.
The goal isn’t to eliminate specific assessment types, like online multiple-choice exams, where ChatGPT can be used effectively, Watson explained. The review will suggest alternative assessment methods where AI is less effective.
Watson came across an academic article describing students using ChatGPT on multiple choice exams. The empirical evidence, along with anecdotes from students who used AI on tests—and an article in The Journal describing students’ reliance on ChatGPT—all motivated Watson to act.
READ MORE: Cheaters tell all: How students are cheating with ChatGPT
AI could prevent students from getting the most out of their degrees and developing the skills needed to succeed in the workforce, Watson said. Assessments’ integrity within programs is important for the University’s confidence in granting degrees.
“We want the general public [to know] a Queen’s grad is able to do all of the things that we say a Queen’s grad is able to do,” Watson said.
Watson anticipates the review will produce a blend of solutions maintaining the integrity and fairness of assessments at the University.
Tags
All final editorial decisions are made by the Editor(s) in Chief and/or the Managing Editor. Authors should not be ed, targeted, or harassed under any circumstances. If you have any grievances with this article, please direct your comments to [email protected].