Concurrent Session Four
4 - 4:50 p.m. (3:30 -4:20 p.m. ADT)
Making Assessment Meaningful: Rethinking Learning, Process, and Community Connections
In an era where generative AI challenges traditional pedagogies and assessment methods, how do we ensure our teaching remains meaningful and our assessments truly reflect student learning? This panel brings together three educators from diverse disciplines—music and inquiry studies, psychology, and veterinary science—to engage in a candid, conversational exploration of authentic and process-focused assessment for first-year and transitional students. Drawing from practical classroom experience, the panelists will share how curiosity-driven projects, real-world community engagement, and innovative use of AI tools can make assessment both more relevant and more resilient. We will discuss the challenge of defining “authentic assessment,” navigating the loaded nature of educational jargon, and ensuring process takes precedence over mere product—especially when AI can generate answers instantaneously. Panelists will present case studies: from inquiry-based learning scenarios requiring collaboration with community clients, to scalable feedback mechanisms powered by custom-built AI tools in large-enrollment courses, and practices that emphasize student reflection, transferable skills, and lifelong learning over rote memorization. Together, we aim to address core questions: How do we prepare students for uncertainty and ambiguity beyond university walls? What does meaningful assessment look like now, and how can it adapt as technology and learner needs evolve? We invite session participants into a dynamic dialogue, hoping they leave with more questions than answers—and with practical ideas to adapt, adopt, and share in their own teaching. Join us as we explore process, possibility, and purpose in higher education assessment.
Presenter(s): Stacey MacKinnon, Associate Professor, Psychology
Affiliation: University of Prince Edward Island
Stream: Authentic Pedagogy and Assessment
Session Format: Panel
Designing assignments that reduce student reliance on AI
Generative AI has prompted many instructors to reconsider how assignments support student learning. This session explores how the Transparency in Learning and Teaching (TILT) framework can help instructors increase students' understanding of why they are completing an assignment, what is expected of them, and how the work contributes to their learning, helping to reduce students' reliance on AI. Participants will explore practical examples and assignment redesign strategies that can be adapted across disciplines.
Presenter(s): Carolyn Best, Educator Developer, CITL
Affiliation: 91´«Ă˝
Stream: The Learning Experience
Session Format: Presentation
Research in the age of Agentic AI : Understand, Adapt, Evolve
AI is doing much more than simply introducing new tools to university research. It is also beginning to influence how questions are formulated, how knowledge is explored, how evidence is interpreted and how research is communicated.
How should researchers respond to these changes? In which areas can generative and agentic AI genuinely support research, and in which areas is caution, critical judgement and human expertise essential?
This presentation invites you to explore the transformations that are already taking place throughout the research process, and to consider how these can evolve while maintaining rigour, accountability, and the vital role of human judgement.
Presenter(s): Florin Filip, Senior Advisor for Innovation, Creativity and Organizational Transformation
Affiliation: Universite de Moncton
Stream: SoTL and AI-Accelerated Research
Session Format: Presentation
Think Before You Bot: Blueprinting Better Bots
Have you considered creating an AI assistant for tasks that are repetitive, time-consuming, or frustrating? As AI assistants become easier to build, the challenge is no longer whether we can build them, but whether we are designing them well. In this session, we will look beyond the prompts you type and buttons you click, to examine an array of intentional design decisions that shape effective AI assistants. Using SyllaBot, Dalhousie’s new AI-assisted syllabus checker as a case example, we will explore a practical blueprinting framework that considers factors such as purpose, audience, tone, scope, inputs, outputs, guardrails, and human-in-the-loop components. Participants will apply the framework to begin a preliminary design plan for an AI assistant that makes a task easier or solves a problem for yourself, your team, or your students. This session is applicable to all AI platforms (e.g., CoPilot, ChatGPT) and types of AI assistants (e.g., chatbot, agent).
Presenter(s): Daniella Sieukaran (she/her) Senior Educational Developer (Program Development), Centre for Learning and Teaching
Affiliation: Dalhousie Univeristy
Stream: Workshop
Session Format: Workshop