Let's Make a Community [of Practice]: Using Community-Based Participatory Design to Support Interdependence
Elaine Czech, Dan Bennett, Grace Jane Stangroome, Vanessa Aisyahsari Hanschke, Amy Ingold, Paul Marshall, Oussama Metatla · 2026 · Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems (CHI '26) · doi:10.1145/3772318.3790423
Summary
Accessibility and HCI research on dementia has moved away from the medical-model framing of 'recipients of care' toward dementia-friendly communities, interdependence, and person-centred support. But Czech and colleagues note a gap: most of this work designs interventions without examining how existing community networks - memory cafes, churches, museums, Alzheimer's Society programs, the NHS - shape whether a technology is adopted, resisted, or ignored. The paper analyses the first phase of a year-long Community-Based Participatory Design (CBPD) project in a mid-sized UK city through Wenger's Communities of Practice (CoP) framework, which treats groups as social structures defined by shared histories, practices, and identities, and locates belonging in three modes: engagement, alignment, and imagination. Five two-part workshops (four in-person at a History Museum, Art Museum, City Farm, and Church, one virtual on Zoom) were held with 14 participants: six people living with early-stage dementia, six spousal informal carers, and two dementia program coordinators. Activities included an 'I Like, I Wish, What If' exercise, technology prompt cards, and co-design sheets structured around When/Where/Who/Type of Technology/Emotion. Approximately 16.5 hours of audio was transcribed and thematically analysed using a combined inductive/deductive approach with Wenger's modes of belonging as the deductive frame. The authors explicitly reflect on non-participation - who was absent, why - and propose that CoP offers a roadmap for designing technologies that support person-to-person interdependence rather than individual independence.
Key findings
Participants enacted all three modes of belonging during the workshops. Engagement surfaced through shared altruistic histories (Jo and Gareth as former social workers, Thomas and Dave as dementia advocates) and through participants actively recruiting others into activities. Alignment was supported by running workshops in familiar community spaces and by shared use of Zoom across several couples. Imagination manifested when participants described the conditions under which they felt welcomed (Gareth: 'a comfortable place for me is where I'm with friendly people, and they look at me') and when they proposed interactive boards, dementia-inclusive signage, and universal technology designs. External boundaries reproduced marginalisation in ways designers need to attend to. Transportation (loss of driving licence after diagnosis, unreliable buses) cut participants off from activities; pandemic-era cuts had not been reversed and social programs remained 'regenerating'; generational stereotyping assumed all people with dementia were elderly, excluding Dave (diagnosed young); and dementia navigators assigned by the state were described as 'about as useful as a chocolate teapot'. Applying Wenger's four forms of non-participation (compromise, cover, strategy, practice), the authors argue recruitment for participatory design must actively plan for each mode: run virtual sessions to offset transport barriers, use familiar venues to reduce 'cover', and design activities whose age and cognitive framing does not entrench stigma. The concrete design implications include peer-scaffolding technologies that build confidence to engage, cross-platform UI symmetry (iOS/Android/Zoom), pre-session virtual walkthroughs to familiarise unfamiliar venues, and system-level identification of the communities of practice bordering a research program (here, NHS engagement failed because the required ethical approval added 6-12 months, so alignment could not be sustained).
Relevance
For accessibility practitioners this paper reframes 'dementia-friendly' from a property of a space or product into a property of the interdependent community that sustains use of that space. A shipping app, museum exhibit, or care service that is individually 'accessible' can still fail if it doesn't align with the Alzheimer's Society, transport, faith, and peer-carer networks that decide whether anyone actually attends. The four-mode non-participation lens (compromise, cover, strategy, practice) is a practical diagnostic for why a pilot user base looks the way it does, and the five CBPD recommendations (familiarisation walkthroughs, peer-scaffolded engagement, cross-device symmetry, initiator acknowledgement, reflexive anti-stigma practice) are directly applicable to any participatory design program working with older adults, disabled users, or marginalised groups. Limitations matter: the sample is small (14 people), predominantly white, from affluent UK neighbourhoods, all in early-stage dementia, and already active in Alzheimer's Society programs. Minoritised ethnic groups who are diagnosed later in the disease trajectory are absent. The CoP analysis is retrospective rather than design-proactive. Findings on non-participation are by definition inferred, since the people who didn't participate were not there to speak. Generalisation beyond CBPD into product-ship accessibility work should be cautious.
Tags: dementia · aging · participatory design · community-based participatory design · interdependence · inclusive design · social model of disability · caregiving · research methodology · accessibility theory