Exploring a design space of interventions that promote reflective use of conversational generative AI among university students
Year: 2026
Type: article
Abstract: The use of conversational Generative AI (GenAI), such as ChatGPT, in higher education raises concerns around cognitive offloading and academic integrity, particularly when students engage with AI-generated content with limited reflection. This thesis uses a Research Through Design (RtD) approach to explore how reflective support can be integrated into conversational GenAI interfaces for academic use. Through low-fidelity sketching, engagement with prior literature, and research team discussions, we formulate an initial design space around two dimensions: degree of externalization and level of friction. We instantiate this space through four contrasting design concepts and examine students’ perceptions of them in a qualitative elicitation study with 20 undergraduate students from diverse disciplines. Findings show that no single intervention is universally preferred; rather, receptiveness is shaped by perceived reflective abilities, academic integrity concerns, external accountability, disciplinary context, task characteristics, and workflow constraints. Building on these findings and prior literature, we extend the design space with the type of motivation as a potential third dimension and introduce a speculative mockup of a fifth design concept. Overall, this thesis contributes an initial design space, five contrasting design concepts, and empirical insights for designing context-sensitive reflective GenAI interventions in higher education.
Source: Mspace (University of Manitoba)
Author: SHAIARA, FAIRUZ
Language: English
Subfield: Education
Field: Social Sciences
Domain: Social Sciences
SDG: Quality education
Open Access status: green
OpenAlex ID: W7213569893