Author: Dr Elaine Tan, Senior Lecturer in Innovation (Reflections on assessment in Combined Honours)
The viva voce has long been a fixture of the doctoral defence and of professional examinations in fields such as medicine, yet it remains an uncommon way to assess undergraduates (Pearce and Lee, 2009). Given the pressures now reshaping how universities assess, not least the impact of generative AI, that under-use may be worth reconsidering (Fenton, 2025). Concerns about the practicality of such an approach are well founded, with rising student numbers causing organisational headaches (Perkins, 2026), and the idea is often quickly dismissed as expensive and unworkable. This, however, is a reflection on practice where it has been implemented: an account of running the viva as the oral component of an undergraduate capstone module (Combined Honours’ Final Year projects) with a cohort of 60 and a team of five markers. It examines how a live conversation tests the researcher rather than the research: whether a student can justify their choices, explain their process, and communicate what they have learned. It also discusses the potential payoffs of this resource-intensive activity: opportunities for authentic and inclusive assessment, meaningful engagement with feedback, and active reflection on learning.
Format of the review
The format was a 20-minute “professional and personal review” with two markers present: the supervisor leading and the second marker following up. Framing the viva as a review rather than an examination signalled a dialogue (Fenton, 2025) rather than an interrogation. Students prepared three questions in advance: to share their main findings, to reflect on their learning, and to consider how the project might shape their future plans. They then faced unseen follow-ups probing why they had reached certain conclusions, which areas mattered most, and how ideas related. They could bring a single side of A4 notes for support.
The module had two separate assessment components. The first, already submitted, was a “final output”. Its format was self-determined: students chose their own audience and medium, which ranged from dissertation-style essays to a video documentary with reflective commentary, amounting to 6,000 words or equivalent. The module was also designed so that students could collectively determine the weighting of the components. This was settled by a class vote: offered a range from 15% to 50%, the majority chose a modest 30%, reflecting their greater comfort with writing and their wariness of an unfamiliar format. That figure proved well judged: high enough to make the review worth taking seriously, low enough to keep anxiety in check.
Redistributing attainment
As with previous explorations of such assessment methods, there was potential for a redistribution of attainment. Weaker writers often perform better when allowed to speak, and oral formats tend to lift overall attainment (Huxham, Campbell and Westwood, 2012). Some students whose natural strengths lie outside writing reflected with genuine insight during the review and raised their overall mark as a result. Conversely, a polished final output did not reliably predict a strong review: a small number of students communicated their findings well but offered limited reflection. This also aligns with the case for assessment for inclusion, which resists treating underperformance as a student deficit when the assessment design may itself be the barrier (Tai et al., 2023; O’Neill and Padden, 2022).
AI deterrent, not “AI-proofing”
The format predates current concerns about generative AI, but it answers them neatly. In itself, it is a deterrent, not AI-proofing. There is no way to stop a student using AI to produce a written submission (Kofinas et al., 2025), and detection tools are an unreliable basis for judging the integrity of submitted work (Bassett et al., 2026; Grundy, 2025). What a live review does is diminish the appeal of complete cognitive offloading: a student who has outsourced the thinking still has to account for their decisions aloud, grasp their own process, and extend their findings under questioning they have not seen. The path of least resistance shifts back towards doing the work well enough to talk about it. This aligns with evidence that authentic, scenario-aligned oral tasks discourage misconduct precisely because they are difficult to outsource (Sotiriadou et al., 2020). It is not a complete solution, but for vulnerable capstone assessments still submitted as extended writing, it is a potentially effective one.
Feedback that students actually used
One significant detail of how the reviews were organised: students received their written marks and feedback on the final output before the review, and prepared for the oral assessment by reflecting on them. The marks and feedback were returned within the standard 20-day turnaround period for the final-output component. This meant that, while many students had completed the work some time before the review, they also had space to reflect on it. A two-week gap between the release of feedback on the written component and the review allowed students to prepare. One issue, however, was the number of extensions students had requested (nearly 50% of the cohort), which compressed this preparation window.
This relationship to the feedback on the final output meant that, rather than treating the oral as a threat, students approached it as a chance to pick up further marks on work already submitted. The written feedback included points to consider for the review, giving students the opportunity to respond to the questions and comments raised on their earlier submission. The effect on feedback engagement was marked: comments that might otherwise have been skimmed and filed (or left unread, as is common with summative capstone feedback) were read closely and used in preparation. Students now had an immediate, concrete reason to act on the feedback, and the review created a meaningful opportunity for feedforward and reflection. Feedback became something to use rather than merely receive.
What would change next time
Predictably, with 60 students requiring 40 minutes each (a 20-minute review plus 20 minutes of feedback and moderation) and two markers present throughout, the exercise required 80 hours of staff time in total. Offering students a wide choice of booking slots stretched the exercise out, with some days holding only a handful of reviews; a little less flexibility would serve staff considerably better and make the process more efficient. Marking with two assessors and moderating immediately after each review, however, kept the load manageable. Recorded verbal feedback (carrying both markers’ comments) replaced written reports, which meant the whole process was accomplished comfortably within the 40 minutes per student, end to end. It should also be noted that the capstone process itself was streamlined compared with a conventional dissertation to accommodate the review: students received three rather than four hours of supervision meetings, and the written component was capped at 6,000 words rather than the usual 10,000, to reduce marking time.
There was also a tension between interrogating students’ knowledge and understanding of the project and the broader prompts intended to elicit reflection on their learning. Students’ use of feedback in preparation was also mixed: given the chance to respond, some did no more than restate the comments provided, while others offered more considered responses, reflecting on how they might apply and develop them. Marker disagreement surfaced occasionally, usually where examiners approached a project from different angles, and was generally resolved by the supervisor supplying context.
Many students had never been assessed by interview and did experience some anxiety. To address this, an optional workshop was held to support their preparation, but it drew thin attendance in a crowded assessment period, a reminder that scaffolding must be timed as carefully as it is designed. As earlier studies have found (Pearce and Lee, 2009), students in the cohort judged the viva surprisingly user-friendly, and almost all valued the chance to talk at length about projects they had chosen and cared about. For a capstone, letting students articulate their learning aloud encouraged engagement with feedback and supported meaningful dialogue.
References
Bassett, M. A., Bradshaw, W., Bornsztejn, H., Hogg, A., Murdoch, K., Pearce, B., & Webber, C. (2026). Heads we win, tails you lose: AI detectors in education. Journal of Higher Education Policy and Management. https://doi.org/10.1080/1360080X.2026.2622146
Fenton, A. (2025). Reconsidering the use of oral exams and assessments: an old way to move into a new future. Educational Researcher, 54(7), 430–436. https://doi.org/10.3102/0013189X251333638
Grundy, D. (2025). The unfairness of AI-flagged academic misconduct investigations in UK universities. Scholarly Insights: Newcastle (SI:N), 1(1). https://blogs.ncl.ac.uk/sin/2025/08/05/the-unfairness-of-ai-flagged-academic-misconduct-investigations-in-uk-universities/
Huxham, M., Campbell, F., & Westwood, J. (2012). Oral versus written assessments: a test of student performance and attitudes. Assessment & Evaluation in Higher Education, 37(1), 125–136. https://doi.org/10.1080/02602938.2010.515012
Kofinas, A., et al. (2025). The impact of generative AI on academic integrity of authentic assessments within a higher education context. British Journal of Educational Technology. https://doi.org/10.1111/bjet.13585
O’Neill, G., & Padden, L. (2022). Diversifying assessment methods: barriers, benefits and enablers. Innovations in Education and Teaching International, 59(4), 398–409. https://doi.org/10.1080/14703297.2021.1880462
Pearce, G., & Lee, G. (2009). Viva voce (oral examination) as an assessment method: insights from marketing students. Journal of Marketing Education, 31(2), 120–130. https://doi.org/10.1177/0273475309334050
Perkins, G. (2026). Resilient assessment in the age of AI: authentic design and the case for verbal examinations in business education. Assessment & Evaluation in Higher Education. https://doi.org/10.1080/02602938.2026.2644516
Sotiriadou, P., Logan, D., Daly, A., & Guest, R. (2020). The role of authentic assessment to preserve academic integrity and promote skill development and employability. Studies in Higher Education, 45(11), 2132–2148. https://doi.org/10.1080/03075079.2019.1582015
Tai, J., Ajjawi, R., Bearman, M., Boud, D., Dawson, P., & Jorre de St Jorre, T. (2023). Assessment for inclusion: rethinking contemporary strategies in assessment design. Higher Education Research & Development, 42(2), 483–497. https://doi.org/10.1080/07294360.2022.2057451