With more than 25 years as a university professor, I can’t recall a time with more upheaval in how teaching, learning and assessment takes place on Canadian campuses.
Although universities have always had to adjust to new forms of technology – such as scientific and graphing calculators, note-sharing apps and other digital tools – nothing comes close to what we have experienced since ChatGPT burst on the scene in late 2022. The plethora of available generative AI (GenAI) tools is dizzying, with approximately three quarters of Canadian students relying on them for their coursework.
GenAI: Two camps, one path forward
So, how should universities confront the challenges to academic integrity presented by GenAI? This is the question with which every professor, program director, dean and senior administrator is currently grappling across Canada and the rest of the world.
The choices are not infinite and there is little agreement amongst academics on the best approach.
Traditionalists argue GenAI should be banned and in-person exams should reign supreme to ensure the validity of grades assigned to students.
Conversely, AI adopters argue for its strategic integration into coursework – let students use GenAI but make sure what you are grading is their unique contribution. Essentially, prepare them to be AI literate so they can thrive in the modern world.
Ironically, both camps still make an implicit assumption of learning and avoid challenging the traditional role universities play as a transmitter of knowledge. Yet, it is precisely because knowledge acquisition is necessary, but insufficient, in preparing students for their futures that we need a new path forward – namely, to fundamentally change how we assess and evaluate students.
The path forward is not subservient to bots, nor bound by traditional practices that are ill-equipped to tap into a broad range of today’s necessary skills.
Assessment design and soft skills
My research examines the intersection of a diverse range of both cognitive and non-cognitive skills. The former relates to achievement in traditional content domains such as reading, mathematics and science. The latter focuses on a diverse set of soft skills such as collaboration, growth mindset (the belief that your achievement in any particular subject area can be significantly improved through effort and persistence), self-regulation and emotional intelligence.
What is critically important is that both cognitive and non-cognitive skills influence one another and significantly contribute to educational attainment and work success. Universities now need to redesign assessment to emphasize performative tasks that more closely connect to these learning outcomes.
What I am suggesting is a lot easier to accomplish in smaller classes. In my graduate courses, which typically include fewer than 20 students, they make individual or group presentations on a specific topic, answer questions and defend their positions – all in real time.
Admittedly, doing this in an undergraduate Psychology 101 class with 500 students is no small feat. Universities are increasingly running large undergraduate classes as a necessity, not as a preferred choice – particularly in first- and second-year courses. These undergraduate courses tend to rely heavily on multiple-choice tests, which are easy to administer and mark.
Sustainable shift and funding
The future is not all bleak, however, and there are many positive examples. University programs are increasingly making use of oral assessments, case studies, portfolios and small group breakout sessions in some undergraduate courses – particularly in health sciences, business and education studies.
Unfortunately, this is being done selectively, not to scale. If we want to effectively prepare students for their futures, we need a more consistent approach to address the challenges of GenAI. Ultimately, all programs across a post-secondary institution need to develop assessment policies that yield more valid performance indicators.
Accomplishing a significant and sustainable shift in assessment policy is a challenge given the steady decline in public funding for post-secondary education.
“Depending on the measure chosen, we are anywhere from $2-20 billion short of where we should be had we kept our spending levels” of the late 2000s and early 2010s, a recent report found.
Children’s privacy is now critical infrastructure in the AI era
In Ontario alone, tuition revenue has now surpassed government operating grants. Ontario students have the lowest per-pupil government funding level in Canada and are thus paying the majority of the relative cost of their university education. They are paying the highest tuition rates, while also sitting in some of the largest first- and second-year classes in Canada.
Unfortunately, the full-time equivalent student-to-faculty ratio across Canadian universities has been climbing for many years. Thus, it shouldn’t come as a surprise that more time-consuming forms of authentic assessment are constrained when fewer professors are teaching more students.
A fundamental shift in university assessment is needed to expand the knowledge, skills and dispositions that students acquire in post-secondary settings. Additional government investments to address rising tuition, increased class sizes and AI-influenced assessment reforms are warranted.
While universities are sites of personal development, they also play a critical role in promoting economic growth and innovation. Addressing provincial funding shortfalls will require a significant investment, but it’s an investment we can’t afford not to make.
