Redesign assessment toward supervised, oral, process-based and authentic formats to verify students' own reasoning under GenAI¶

Strategy · All strategies
Evidence · no claims cited

Description¶

Participants described moving away from unsupervised written essays toward more face-to-face, oral and closely supervised assessment in response to GenAI. Strategies named include invigilated examinations for essential skills, oral vivas with differing scenarios to observe clinical reasoning, rubrics written to identify AI use, and process-based marks such as "five marks for showing us a screenshot of their literature search terms from a database". The article presents these as academics' early adaptation efforts rather than tested interventions.

Design Implications¶

Context¶

Requirements¶

• Assessment tasks must continue to evaluate students' own reasoning, judgement and application of knowledge, per the article's recommendations section.

Constraints¶

• The article notes concerns that exams may be ineffective if students relied on GenAI throughout the rest of the course; these approaches also require careful design.

Target Learners¶

• undergraduate nursing students

Target Learning Goals¶

• verifying students' own knowledge and clinical reasoning
• academic integrity
• readiness for safe practice

Related Strategies¶

• Redesign assessment reflectively for possible AI use, using supervised tasks for lower-order skills and intentional AI use with reflection for higher-order work
• Pair AI-permitted take-home assignments with supervised modifications, oral defense, prediction, and transfer tasks to verify understanding
• Replace detector-centred assessment with a four-stage process-oriented model: declared starting point, versioned drafting, process artefacts, and targeted oral verification
• Redesign AI-era physics courses around an unaided written examination with mock exams, prepared tutorial discussion, and research-shaped tasks as bonus work with scaffolding
• Graduate AI literacy training should promote purposeful, task-oriented, critically supervised GenAI use rather than targeting overall use levels
• Adapt non-invigilated written assessments and their marking criteria in response to Gen-AI susceptibility
• Strategic steps to move beyond informal AI innovation: recognise redesign in workload models, embed innovation in promotion criteria, set institutional AI principles, redesign assessment, and fund scaling pilots
• Two-instrument assessment: AI-free coding quizzes plus verification-gated oral defenses of comment-stripped AI-assisted work
• Seven-step reach audit of a course
• Provide AI-free proctored spaces so some take-home work is verifiably unassisted
• Redesign assessment toward progressive, authentic, and reflective methods rather than intensive paperwork

Examples¶

-

Key Sources¶

• Dabkowski, Elissa; Missen, Karen; Allen, Louise; Whitehead, Dean; Worn, Ryan. (2026). 'Co-pilot won't teach you to be a nurse': nursing academics' perspectives on GenAI use in undergraduate education. International Journal for Educational Integrity. https://doi.org/10.1007/s40979-026-00220-2