AI in Graduate Education: At the Intersection of Learning, Research, and Knowledge Translation
In this post we’re considering the impact of generative AI on research-intensive graduate education. We hope our reflections and ‘Challenge’ will resonate with many research disciplines.
As expected, graduate education is undergoing a transformation in response to the ever-increasing access to powerful generative AI platforms. Universities across the globe have published guidance for the use of AI in graduate work, and Western is no exception. That is, here at Western, after completing many layers of consultation, we have established a principles-based guidance framework recommending accountability, integrity, transparency, privacy and data security, and inclusion in AI use. This framework acknowledges the complexity of AI integration while placing responsibility on students to understand the limitations and consequences of the tools they employ. Faculty members shoulder this same responsibility alongside the graduate students they supervise and mentor.
What makes graduate education, and particularly research thesis-based graduate education, distinctive is that it requires students to go far beyond didactic learning. Graduate students occupy a dual role as both learners and knowledge producers. As knowledge producers, they are expected not only to advance their fields through original research, but also to effectively share their specialized work with broad audiences of knowledge consumers. Western’s guidance appropriately recognizes this duality, requiring students to discuss AI use plans with their supervisors and committees while prohibiting AI use in the thesis examination process itself—a safeguard for both intellectual property and the independent expert judgment that defines scholarly evaluation.
So how can supervisors foster productive conversations about when and how AI tools can support research and knowledge translation workflows, without compromising students’ development of disciplinary expertise, methodological rigor, communication proficiency, and the critical thinking that they must demonstrate and defend throughout their work? In some cases, mismatched views of the utility of AI may preclude this question from even being considered. For example, the student may be enthusiastic about the potential for AI to enhance research productivity, while the supervisor is skeptical, or vice versa. Recognizing these different perspectives, we offer a two-part Challenge.
The Challenge
For AI pragmatists (i.e., those who consider the use and products of AI with both caution and consideration), ask your AI platform of choice to provide a conversation starter to explore the question of “how" do we foster productive conversations about when and how AI tools can support research and knowledge translation workflows, without compromising our students’ development of disciplinary expertise, methodological rigor, communication proficiency, and critical thinking that they must demonstrate and defend throughout their work?”
Ask the AI to provide several options. Review the options. Are they aligned with the guidance of your institution? Are they aligned with your perspective and your student’s perspective on AI use? Do they consider graduate program learning outcomes?
For AI skeptics (i.e., those who think the use of AI will more likely have a detrimental versus constructive impact on graduate research student training), copy and paste one of your published research abstracts into Microsoft Copilot or another LLM. Pay attention to the privacy policy and user agreement of the tool. Ask the AI to provide a summary of the abstract for a non-expert audience of your choosing. Review the non-expert summary. Was the overall accuracy of the work retained? Were any biases introduced? What level of knowledge and critical thinking skills are needed to assess the adequacy of the AI-assisted summary? How much would you need to revise the summary to be comfortable with the output? Do you think this AI-assisted approach to knowledge translation could be valuable for you and for your students?
Disclosure
Claude.ai was used to draft an outline for this post (not including the Challenge) after the concept was developed by the authors. The initial draft of the post was revised with Claude.ai for clarity, followed by further revision by the authors. Claude.ai and Copilot were used to test the two-part Challenge.Your Challengers: Nica Borradaile & Laura Murray
Nica is an Associate Professor in the Department of Physiology and Pharmacology, and the Associate Dean Graduate and Postdoctoral Studies for the Schulich School of Medicine and Dentistry. Nica’s research is in lipid metabolism as it relates to liver and vascular complications of obesity. She has approached the use of GenAI as a tool to assist and accelerate analyses of large datasets arising from high throughput measurement of changes in genes, proteins, and metabolites. Through her roles in graduate education, Nica contributed to the development of the SGPS Provisional Guidance for the use of Generative AI in Graduate Studies.
Laura is the Associate Dean of Graduate and Postdoctoral Studies in the Faculty of Health Sciences and a Professor in the School of Communication Sciences and Disorders at the University of Western Ontario. Her clinical work experiences as a speech-language pathologist have informed and inspired her research, which examines how cognitive deficits (i.e., attention, memory, executive functioning) interact with the language abilities of adults living with aphasia or other acquired neurogenic communication disorders. Her research has also focused on developing assessment and treatment strategies (including use of digital platforms) for acquired neurogenic communication disorders that consider these cognitive deficits and that can be used in clinical settings. Laura has extensively published and presented, to both academic and clinical audiences, in the fields of aphasia, right hemisphere cognitive-communication disorders, dementia, progressive language disorders, epilepsy, traumatic brain injury, and typical aging. She has taught within speech-language pathology and neuroscience programs, and has received several awards for her teaching efforts at both the undergraduate and graduate levels.
