secondary research
REadings
Ai in higher education symposium
AI conversations: UAL webinars
There were a number of readings which gave me direction and context for my work
I attended several Webinars and one online Conference, many of these were focused on some functional elements of the use of AI, sharing how educators and institutions were using it, however underpinning all of them was a sentiment that AI is here and we’re not leaning into it enough to understand how to embrace and use it ethically.
There are discrepancies across higher education in terms of the willingness and capability to embed it and really understanding how and who should be leading this work.
The paper by Myke Healey on Approaches to Generative Artificial Intelligence A Social Justice Perspective gave me further insight into the challenge around the design and development of some LLM tools and the idea of ‘digital colonialism’ leading to me to reflect on the compelxity for universities such as UAL where social justice is one of our pillars and yet this technology is being adopted and used by students and teachers already. This was supported by ‘AI’s English Problem’ which explored the Western bias of LLM
“ChatGPT can also make up words, struggle with syntax and generate gibberish in many underrepresented languages, Andrew Deck writes for Rest of World”,
However despite all this, surely we need to lean into exploring capability and have informed discussions on the role of AI in educations. Oxford University’s paper ‘Beyond ChatGPT’ is a really helpful document but even there the conclusions are around the engagement of AI with students and less on how teaching staff should / could / would want to connect with it.
The Russell Group issues five principles following consultation and engagement with the sector
- Universities will support students and staff to become AI-literate.
- Staff should be equipped to support students to use generative AI tools
effectively and appropriately in their learning experience. - Universities will adapt teaching and assessment to incorporate the
ethical use of generative AI and support equal access. - Universities will ensure academic rigour and integrity is upheld.
- Universities will work collaboratively to share best practice as the
technology and its application in education evolves.
Beyond Chatgpt (2023) Centre for Teaching and Learning. Available at: https://www.ctl.ox.ac.uk/beyond-chatgpt (Accessed: 18 January 2024).
Point 1 – that universities will support students and staff to become AI literate led me to consider whether interventions such as this research could be part of such literacy training.
For thoughts on how secondary data informed by approach please see https://samb22.myblog.arts.ac.uk/2023/10/27/arp-research-methods/(opens in a new tab)
research analysis
Thematic clustering
Whilst the sample size for my research was not significant there was quite a lot of data from the conversations with tutors which needed to be segmented and themed. Braun and Clark outline that…
“Thematic analysis is a qualitative method for uncovering a collection of themes, ‘some level of patterned response or meaning’ (Braun & Clarke, Citation2006, p. 82) within a data set. It goes beyond word or phrase counting to analyses involving ‘identifying and describing both implicit and explicit ideas’ (Guest, MacQueen, & Namey, Citation2012, p. 10).
They go on to outline that it can be used for 2 – 400 participants so my sample size of 4 fits within this criteria. As this is action research we are not exploring statistically robust samples and want to understand what impact our intervention has had.
I used a method that Lindsay had outlined in the PG Cert Workshops of downloading MP3 files to Word for transcription. I then printed out the interviews and highlighted interesting comments or discussion points, using a different colour for each tutor. I then cut out those comments and clustered thematically looking for insight or topic areas.
I then used MIRO to mindmap the themes and additionally look for any commonalities or observations beyond the themes such as where differing opinions may have arisen, and where the conversations led beyond AI.
research findings
I mapped the key themes and then used a mind-mapping technique on MIRO to capture these visually
The research created positive and encouraging conversations with my colleagues on the role that AI may play in supporting teaching. It is far from simple and loaded with ethical conversations however there may be ways in which we can use it as a constructive tool to support.
The concept of a third marker in the room, a tool to test assessment criteria looking for gaps and providing ideas on how to structure are all interesting. None of these require student work to be uploaded to an AI. For tutors with neurodiversity such as dyslexia then the early indications are that this could prove to be immensely useful.
outcome presentation
I decided to do a story from a tutor’s point of view of the research conversations that would capture the key findings. I started writing a story but decided to shorten it as it felt too long and not effective enough in communicating the key findings. I wanted something I could read out in the class presentations so shortened to a poem/story.
However I feel the analysis I’ve done is also clear enough that if I wanted to do any follow on research that I would know which areas to dig into further
A tutors story
reflections on what next
I would be interested in exploring this topic further and engaging with digital learning team on what pilots or projects are being developed around the role of AI in supporting teaching. One of my colleagues on the course has also explored AI as her PGCert and tested an intervention in one of my teaching sessions. It was great to work together on that and participating in her project has continued to prompt thoughts for myself on this issue