Adapting Assessment in the GenAI Era: AN Undergraduate Physics Case STudy

Abstract

PROBLEM

The widespread adoption of Generative AI (GenAI) tools in higher education, while offering learning opportunities, can also compromise academic integrity and assessment authenticity. This necessitates a proactive approach with clear institutional policies to guide appropriate use of GenAI.

PLAN

Assessments for a 2nd year physics unit were evaluated for vulnerability to GenAI misuse to determine the appropriate level of use according to institutional policies, as well as evaluating the need for redesign to ensure authenticity and resistance to GenAI misuse.

ACTION

A risk matrix was applied to assessment tasks to examine the likelihood of misuse and potential impact on learning, informing GenAI use category designation and flagging highly vulnerable tasks for redesign. Adopting GenAI use categories maintained constructive alignment and reduced opportunities for inappropriate GenAI use. Clear guidance on acceptable use was embedded in briefs and rubrics, supported with workshops and educational tools like infographics.

REFLECTION

The structured assessment evaluation led to clear use designation and informed targeted redesign. A key finding is that adoption of new or adjusted institutional guidelines for GenAI use has a non-trivial workload and may trigger larger collaborative projects to ensure GenAI resistant assessments, ensuring students achieve the learning outcomes. Future work will explore the impact of varying GenAI use on assessment tasks on their performance, including quantifying unethical use.