This timely project explores how UBC students are using generative artificial intelligence (GenAI) in their academic work, and how factors such as motivation, belonging, identity, and support shape that use. Using UBC Vancouver as a living lab, the project will identify patterns of student AI engagement and examine how courses and workshops can foster more ethical, equitable, strategic, and learning-oriented use of AI. We interviewed the project team to learn more about their work.
Interviewees:
- Melissa R. Hunte, Assistant Professor, Department of Language & Literacy Education, Education
- Warren Code, Associate Director, Skylight (Science Centre for Learning and Teaching), Science
The team collectively shared their insights and answered our questions.
Q: How will you distinguish between learning-oriented AI engagement and learning-displacing AI engagement?
Patterns of GenAI engagement based on students’ reported use and learning-related experiences will be examined using indicators such as GenAI-use frequency, purpose, timing, task type, motivation, perceived autonomy, perceived competence, social connectedness, perceived AI literacy, ethical use, academic ownership, and perceived impact on learning. Conceptually, learning-oriented engagement will be characterized by selective, intentional, and reflective use, stronger academic ownership, higher perceived AI literacy, clearer ethical awareness, and greater confidence in using AI without absolving responsibility for learning. Learning-displacing AI engagement will be indicated by patterns such as high-frequency use combined with lower academic ownership, lower confidence completing academic work independently, weaker ethical boundaries, and lower perceived value of assignments or program learning.
Q: Could you elaborate on how to measure the effectiveness of the AI-awareness intervention in shifting students towards more learning-oriented AI engagement?
We will evaluate the effectiveness of AI-awareness interventions in two phases. First, we will compare the AI-engagement profiles of students who have participated in AI-awareness workshops, courses, or modules with one another and with those who have not. This comparison will allow us to examine whether prior exposure to AI-literacy education is associated with more learning-oriented patterns of GenAI use.
Second, we will partner with existing AI-awareness workshops and courses across campus to collect pre- and post-intervention survey data from participating students. Students will complete the AI-use and motivation survey before and after participation, allowing us to examine whether their AI-use beliefs, motivations, and behaviours shift over time. The comparison between students with and without prior AI-awareness training will show whether exposure to AI-literacy education is associated with more learning-oriented AI engagement patterns at the population level. The pre-post design will provide stronger evidence about whether specific workshops or courses are linked to changes in students’ AI-use profiles over time. This will help identify which forms and levels of AI-awareness education are most effective, for whom, and whether additional or more targeted supports are needed.
Q: In what ways will your study reveal which student groups are most susceptible to learning-displacing AI engagement patterns?
The study will reveal which student groups may be more susceptible to learning-displacing AI engagement patterns by examining how AI-use profiles vary across demographic, academic, and relational factors. After identifying distinct GenAI-use profiles, we will examine whether profile membership differs by variables such as program level, faculty or department, language background, gender, age, academic belonging, social connectedness, academic self-efficacy, and academic ownership. This would allow us to identify whether particular groups of students are more likely to belong to profiles characterized by learning-displacing AI engagement.
Q: How will the findings from this study directly inform university policy and pedagogy regarding learning-oriented AI engagement in the classroom?
While certain clusters of attitudes and behaviours have emerged from the large amount of research on GenAI use by students in higher education in the past few years, it is still not well understood how or to what extent these may change in response to AI-related training. Our findings will directly inform university policy and pedagogy by identifying what types of AI-related support are most effective, for whom, and at what level of the institution. At the course level, these results can help instructors develop clearer, more learning-oriented guidance on responsible AI use. At the faculty or program level, findings can inform whether students would benefit from more structured AI-literacy supports embedded early in their programs. The survey and profile structure can also provide a foundation for future research that links self-reported AI-engagement profiles with additional forms of evidence, such as learning analytics, interviews, assignment-level AI-use reflections, and observations of students completing academic tasks with AI assistance. This future work could help validate the profiles, deepen understanding of students’ AI-use behaviours, and examine how different forms of AI engagement relate to academic performance, learning processes, and student support needs.
The study can also show how university culture, students’ sense of belonging, and social connectedness may influence GenAI use. If students with lower belonging or weaker academic confidence are more likely to rely on AI in unproductive ways, then responsible AI policy must be connected to broader policy, pedagogical, and student-support strategies. In this way, the project will help UBC develop policies and teaching practices that are technologically responsive, equitable, relational, and learning-centered.