Lessons from Expert Focus Groups on how to Better Support Adults with Mild Intellectual Disabilities to Engage in Co-Design
Ryan Colin Gibson, Mark D. Dunlop, Matt-Mouley Bouamrane · 2020 · Proceedings of the 22nd International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS 2020) · doi:10.1145/3373625.3417008
Summary
This paper proposes and evaluates a two-stage procedure for making co-design workshops accessible to adults with mild intellectual disabilities (ID). The context is the development of a clinical AAC application to help people with mild ID communicate more effectively with their general practitioners (GPs) — motivated by severe health inequalities in this population, where a 2013 study found 42% of deaths among 247 ID patients across three UK hospitals were premature and 27.5% were attributable to inadequate care. Standard co-design techniques (focus groups, prototyping, think-aloud evaluation) rely on higher-order cognitive skills like abstraction, creativity, and metacognition that may be impaired in people with ID, yet there is a lack of guidelines for adapting these methods. The researchers first reviewed literature on co-design with the ID population (8 papers from PubMed, Google Scholar, and ACM), identifying common techniques and adjustments. They then conducted two focus groups with 12 experts (ID nurses, researchers, charity workers, a digital inclusion assistant, and notably one participant who herself has mild ID) across Dundee and Glasgow, Scotland. Each focus group (averaging 78 minutes) walked through four proposed design tasks — a focus group discussion, an image board activity, paper prototyping, and evaluation of an existing tablet app — with experts identifying accessibility barriers and recommending adaptations for each.
Key findings
The experts identified multiple barriers across all four activities. For focus groups: response bias (adults with ID are often "people pleasing" and give expected answers rather than their own views), complex concepts (words like "symptom" and the concept of time are cognitively challenging), and caregiver interference (some caregivers "take over or direct" participants, while others can helpfully identify inaccurate responses). For image boards: labeling images can introduce response bias (participants may just agree with labels rather than form independent judgments), and the heterogeneity of the ID population means different people respond to different image styles (photographs vs. line drawings vs. cartoons). For paper prototyping: low-fidelity paper was deemed more accessible than high-fidelity digital prototypes, but abstract UI elements like "Answer" and "Question" tags were cognitively challenging — experts recommended concrete objects only. For think-aloud evaluation: the simultaneous demands of performing an action and describing it were too cognitively complex — experts recommended a post-task walkthrough instead. Key adaptations recommended across all activities: use plain language throughout; employ sticky notes as concrete referents to aid short-term memory; provide diverse concrete examples to ease cognitive load; limit Likert scales to 5 points maximum with supplementary symbols; use multiple modalities (verbal, visual, tactile) since people with ID respond to information differently; clearly define the caregiver role (support without directing); let participants interact with technologies before evaluating them; configure accessibility settings on devices in advance; and use flip-chart paper turning to symbolize screen transitions in paper prototypes.
Relevance
This paper fills a critical methodological gap for researchers and practitioners who want to involve people with intellectual disabilities in technology design but lack guidance on how to adapt standard co-design methods. The two-stage process — literature review followed by expert validation of proposed tasks — is replicable across different design contexts and populations. The specific, actionable recommendations (shorten Likert scales, replace think-aloud with post-task walkthrough, use concrete rather than abstract objects, define caregiver roles explicitly) can be applied immediately to any co-design study involving people with cognitive disabilities. The finding that experts from different professional backgrounds contribute different types of insights (ID nurses identified clinical barriers, digital inclusion workers flagged technology-specific issues) argues for diverse expert panels. The health context underscores the stakes: poorly designed healthcare technologies for people with ID contribute to documented premature deaths. Limitations include the Scottish healthcare context (publicly funded NHS system), the focus on mild ID only, and the acknowledged risk of over-reliance on expert proxies rather than direct involvement of people with ID themselves.
Tags: intellectual disability · co-design · participatory design · augmentative and alternative communication · healthcare accessibility · research methods · inclusive design · design methodology