Author: Phil Morey, SL in Accounting and Finance.
We’ve all been there, sitting in front of a stack of coursework essays with that sinking feeling that we are marking a ghost. A couple of years ago, the tell-tale signs of generative AI were easy to spot: the hallucinated references, the ‘off-topic’ content and the oddly robotic narrative. Now, however, AI-produced content has become virtually indistinguishable from high-quality, student-sourced work. Even when our suspicions are raised, the difficulty in proving a case of academic malpractice remains a significant hurdle for staff.
At the same time, the Newcastle University Education Strategy (2024) is driving us toward making our students more ‘work-ready.’ Students are increasingly choosing universities based on how well they bridge the gap between a degree and a graduate job. So, we are left with a fundamental question: How do we achieve the sort of educational development in our students that they and we increasingly demand, whilst maintaining the integrity and security of our assessments?
The Solution: The ‘Pre-Seen’ Hybrid Exam
The solution I implemented for the PG Corporate Finance module (NBS8335) was a move away from the traditional take-home essay and toward a 50/50 hybrid exam. This wasn’t just about moving the essay into an exam hall; it was about redesigning the preparation for it. The assessment on this module needed to change, partly due to increasing use of AI generated content but also to ensure that the assessment tested the breadth of knowledge contained in the course alongside depth in a particular area of focus.
The assessment was split into two distinct sections for the new hybrid exam:
- Section A (The Breadth): 50 marks of unseen, general knowledge questions to ensure a solid grasp of the entire syllabus.
- Section B (The Depth): A 50-mark essay where the titles and three specific Financial Times (FT) articles were pre-released to students four weeks in advance.
This model relies on what I call the ‘Invigilated Filter’. It allows students the space to research and brainstorm – even using AI as a tutor – but requires them to walk into a room and perform that synthesis by hand, from memory, under the watchful eye of an invigilator. As Dawson (2020) argued with respect to essay ‘mills’, in a world where assessment fraud can bypass many of our traditional assessment methods, securing the environment where the final performance occurs becomes paramount.
The ‘Fuzzy Mix’ of Pedagogies
Critics of traditional exams often point to them being overly didactic or mere tests of memory (Race, 2014). However, in a content-heavy module like Corporate Finance, there is a necessary level of didactic pedagogy – you cannot analyse a capital structure if you don’t know the formulas.
The module employs a multi-modal pedagogical framework, recognising that no single teaching mode is sufficient to bridge the gap between complex financial theory and professional practice (Archer and Breuer, 2016). This involves a didactic approach to lectures, ensuring the students navigated a content-heavy syllabus, coupled with active and authentic pedagogies through the pre-release of essay titles and FT articles and the encouragement to use generative AI as part of the preparation phase.
This approach is a ‘fuzzy mix’ whereby we use the lecture to transmit the rules of the game and the assessment to encourage active learning, placing the students in the middle ground between these different pedagogical approaches (Peseta et al, 2016). By pre-releasing the materials, we require students to act as practitioners using ‘live’ data. This is constructive alignment (Biggs and Tang, 2011) at its most practical: if we want students to think like professionals, we must provide professional ‘boundary objects’ like the FT to work with. Helliar et al. (2000) have long championed the use of ‘live’ news projects in finance, noting that they increase student engagement by making theoretical messiness tangible.
Data Gathering
To gain a comprehensive understanding of the student experience, a mixed methods approach was adopted, following an explanatory sequential design (Creswell and Creswell, 2018). The use of both questionnaires and focus groups allowed for data triangulation, increasing the reliability and depth of the findings (Saunders, Lewis and Thornhill, 2019). Data gathering was conducted after students had sat the exam and before exam results were released, in order to understand the ‘raw’ feeling of the students’ experience, untainted by their final grade (Robson and McCartan, 2016).
Initial Findings: Control, Inclusivity, and AI
The feedback from our post-exam student questionnaire (n=12) and two focus groups provided a fascinating look at the human side of this transition.
The questionnaire was designed to identify how closely students associated with specific statements about the assessment, their skills development and their use of generative AI. The outputs from the questionnaire were then used to help shape the focus group questions, following an explanatory sequential design where the quantitative findings inform the qualitative instrument development (Ivankova, Creswell and Stick, 2006).
1. The “Control” Factor
The questionnaire showed that students felt remarkably more in control of their performance (score of 4.45/5) compared to traditional exams. However, the focus groups revealed an initial shock regarding transparency.
“Initial feelings were….confused a bit because it took me a while to get my head around the fact that like we were getting given what was actually being put in the written exam.” (Focus Group 1, participant 3)
Once this feeling of disbelief subsided, preparation shifted from guessing to refining. As one student noted, the pre-seen format encouraged a deeper, more iterative engagement with the material:
“It took me a minute to read and relate the article to the syllabus but as I spent more time….it all began to relate and make sense. It definitely helped me to understand concepts….having to apply it to the real world FT article.” (Questionnaire, respondent 3)
2. Inclusivity and the International Experience
A profound benefit emerged regarding our international cohort. For non-native speakers, an unseen exam is often a test of reading and processing speed as much as financial acumen. By pre-releasing the FT articles, we removed the ‘linguistic load’ (Abedi, 2006) One student explained how the UK marking system and the unseen tradition for exams were barriers they could now navigate with more confidence.
“Knowing how the exams are here, because we’re international and all that….I had that feeling that if I’m preparing for this, I have something in my hand that I can actually go and write.” (Focus Group 2, participant 1)
This extra time allowed students to move past the translation phase and into the analysis phase, aligning with the goals of inclusive assessment (Boud and Falchikov, 2007),
3. AI as a Scaffold: The Brainstorming Partner
The data confirmed our hope: students used AI to scaffold their learning rather than bypass it. They used it to clarify complex concept links within the FT articles (score of 4.09/5), but the invigilated closed book exam ensured the final work was their own.
“I read….the FT article, and then I… asked AI to….give me a summary…. so I could…get an overall kind of understanding of what was happening. And then…. went back and forth with AI, just making sure I was understanding things in the right way.” (Focus Group 1, participant 3)
The Friction: ‘Synthesis Anxiety’
It wasn’t a perfect transition. We identified a novel type of stress, ‘synthesis anxiety’. Because students prepared so thoroughly, they struggled to fit their vast research into a 60-minute handwriting window.
“I definitely liked this exam the most….but I think I would have probably benefitted more with more preparation [time] as well…. because I had so much I wanted to say.” (Focus group 1, participant 1)
Reflections for the Future
This format isn’t a ‘cure-all’, but it is a powerful tool in our academic toolkit. It preserves the integrity we need for professional accreditation while giving the students space to behave like professionals during their revision.
Perhaps the biggest takeaway is that there are ways to engage with AI that provide positive outcomes for both students and academics. By changing the format of the assessment, we allow students the freedom to use modern tools during private study while ensuring that what they produce in their summative assessment is a true reflection of their own hard-won knowledge.
References
Abedi, J. (2006) ‘Psychometric issues in the accessibility of assessments for students with limited English proficiency’, Educational Assessment, 11(3-4), pp. 227-246.
Archer, A. and Breuer, E.O. (2016) Multimodality in Higher Education: Showcasing Creative and Scholarly Works. Leiden: Brill.
Biggs, J. and Tang, C. (2011) Teaching for Quality Learning at University. 5th edn. Maidenhead: Open University Press/McGraw-Hill.
Boud, D. and Falchikov, N. (2007) Rethinking Assessment in Higher Education. London: Routledge.
Creswell, J.W. and Creswell, J.D. (2018) Research Design: Qualitative, Quantitative and Mixed Methods Approaches. 5th edn. Los Angeles: SAGE.
Dawson, P. (2020) Defending Assessment Security in a Digital World. London: Routledge.
Helliar, C.V., Michaelson, R., Power, D.M. and Sinclair, C.D. (2000) ‘The use of live projects in finance teaching’, Accounting Education, 9(4), pp. 367-376.
Ivankova, N.V., Creswell, J.W. and Stick, S.L. (2006) ‘Using mixed-methods sequential explanatory design: From theory to practice’, Field Methods, 18(1), pp. 3-20.
Newcastle University (2024) Education Strategy. Available at: https://www.ncl.ac.uk/about/strategy/education/ (Accessed: 10 February 2026).
Peseta, T., et al. (2016) ‘The conundrum of the ‘middle ground’ in higher education pedagogy’, Teaching in Higher Education, 21(3), pp. 245-257.
Race, P. (2014) The Lecturer’s Toolkit: A Practical Guide to Assessment, Learning and Teaching. 4th edn. London: Routledge.
Robson, C. and McCartan, K. (2016) Real World Research. 4th edn. Chichester: Wiley.
Saunders, M., Lewis, P. and Thornhill, A. (2019) Research Methods for Business Students. 8th edn. Harlow: Pearson.