Setting the Context
Just before Christmas the Higher Education Authority (HEA) released a Policy Framework “to guide educators, academic leaders, and professional staff in making informed, value-based decisions about how gen AI is adopted and integrated into educational practice”. The framework is very timely and focuses on the specific issues associated with generative AI usage in higher education in the context of teaching and learning.
The framework is intended to be a “tool to think with”, so that individuals or institutions can use it as a tool for reflection and structured decision-making. The framework is concise and is underpinned by core five principles:
- Academic integrity, transparency and accountability
- Equity and inclusion
- Critical engagement, human oversight and AI literacy
- Privacy and data governance
- Sustainable pedagogy
Dr Alan Wall, HEA CEO noted that the framework has the potential to provide “a strong foundation for sector-wide coherence while enabling institutions to exercise academic judgement and autonomy as they engage with rapidly evolving technologies”. This it most certainly can do, and I believe it also has the potential to inform sector-wide policy and implementation practices in other sectors of education, namely in Further Education and Training (FET) and in schools. While each sector is unique in relation to the use of GenAI in teaching and learning, there are areas of commonality and there is potential to tweak the principles for these other sectors also. This is something we plan to do in our new Erasmus+ Project, Generative AI for VET Educators (GenAI4VET), which commenced recently with partners from Ireland, Belgium, Germany, Poland and Slovenia.
Focusing in on Principle 1
While the framework is concise, it touches on a range of issues from procurement, AI literacy, staff and institutional AI competences to the cost of procuring and supporting these tools for all those working and participating in HE. The cost of procuring appropriate tools and of supporting staff and students to use these tools appropriately will be substantial, in a context of a sector that is already underfunded.
However, one of the areas that attracted attention in the media is assessment and the possibility that face-to-face interviews, or Oral Assessment Safeguard as the framework describes it, could become a routine part of third-level assessment practices.
This is an interesting suggestion and one that several institutions have been considering for some time. In 2024 H2 Learning published a series of podcasts focusing on the GenAI in FET and HE, where we heard from staff on how they were modifying their assessment approaches in response to the arrival of GenAI. Teachers in Coláiste Dhúlaigh College, DCU and in SETU shared how they were using GenAI in their teaching and in assessment, where staff are already asking students to explain what they did and why. This seems to resonate closely with the ideas in the HEA Framework, which shows that institutions are already actively considering a range of assessment approaches, other than written artefacts.
Apparently, Socrates held a low opinion of writing and written artefacts and he believed that
True understanding, he argued, requires face-to-face dialogue. This is not found on the written page; rather it is achieved through dynamic question-and-answer exchanges involving challenging and reframing ideas across a range of perspectives.
So the arrival of GenAI technologies is now necessitating a return to older forms of assessment, such as the face-to-face interview.
Socrates’ method of focusing on asking better questions appears to be a timely response to the challenge of GAI. To design the future of education in response to GAI, we may need to return at least partially to the oral traditions of education, when dialogue was valued more than written texts, and exams were conducted through verbal questioning – viva voce, meaning with living voice (Generative artificial intelligence and a return to dialogue in education).
Written artefacts, are only one mode of assessment, yet it has almost become the de facto approach in many courses. Yet many universities, such as Maynooth University, are providing guidance on a range of assessment approaches in the Age of AI, and these are worth checking out. There are pros and cons associated with oral assessment approaches, and this was previously captured by our TeachNet podcast, AI and Assessment, with SETU and DCU staff sharing their experiences of the impact of GenAI on assessment practices in 2024.
The big takeaway from these discussions was the extra workload that these changes in assessment approaches are having on staff, this takes time.
Final Thoughts
Thus, institutions need to review their existing assessment approaches and provide staff with time to design new oral assessments, and they will also need time to assess them. This is something that should be considered at the institutional level and not left solely in the hands of individual staff. Staff need to be inherently involved in any decisions to alter existing assessment approaches, but they need institutional support to change their practices and this needs investment.