P21 2026 - Kavalchuk: Securing Academic Integrity via Individualised Paper-Based Assessments by Laurine Hurley

alix_thoeming — 2026-09-10 06:20:48 UTC

Could you share more about how you protected against AI-assistance in the ePortfolio? I may have missed something key here, but if it is uninvigilated could students not still use a well-prompted GenAI tool to inappropriately generate their answers? Perhaps more information about how you identify the AI-generated submissions would answer my question.

2026-09-10 06:26:28 UTC

Thank you for sharing your work. Do you think the principle of this assessment individualisation could be adapted to a discipline such as health, where assessment focuses more on clinical reasoning and professional judgement than numerical problem solving? I have had a colleague do similar to you with image data set creation which has been successfu and I am interested in your thoughts. Thankyou

Thanks for sharing your work. I imagine it takes considerable time to set up the underlying problem and variations in the question bank in the first instance. I'm curious if you are adapting and adding variations to the data set with the use of AI?

genevievemurray4 — 2026-09-11 11:51:13 UTC

I am trying to envision adapting this to a legal setting by changing details regarding complex hypothetical factual scenarios. Your response to the previous question re the focus on process rather than outcome is something that I am definitely going to reflect on in adapting my assessment/rubrics.

genevievemurray4 — 2026-09-11 11:51:24 UTC