Abstract
Virtual reality (VR) is a promising educational tool for dementia care. However, its development is often dominated by expert-led models that overlook the context-based tacit knowledge of people with dementia (PWD) and their caregivers. We aimed to describe and evaluate a “co-design” process while acknowledging the reality of an asymmetrical power structure in which the university retains the final decision-making authority, in a context where strictly community-based participatory research is difficult to implement. We adopted a “multi-layered engagement strategy,” using a dementia café in Fukuoka, Japan, as our base. Within this framework, a steering committee comprising PWD, caregivers, experts, and university researchers co-created all three phases of the process, while external stakeholders contributed through consultations and testing at specific stages. We conducted thematic analysis of natural conversations during meetings and used Arnstein’s Ladder of Participation as an analytical framework for process evaluation. The results revealed that the depth of participation determines the type of knowledge gained. While explicit insights were obtained from external stakeholders during the consultation phase, the ongoing negotiations during the partnership phase yielded contextual and tacit insights that challenged the project’s underlying premise: “Conveying the ‘atmosphere’ of a safe and supportive space is more important for learning than simulating specific dialogues defined by experts.” Even under constraints that hindered full empowerment, this sincere collaboration generated “power with” between participants and researchers. By demonstrating the incorporation of tacit knowledge from the field into the core of a product with unequal power dynamics, this study presents a highly transparent and reflexive co-design model that addresses the challenge of fragmented participation in technology development.