Leveraging Large Data Sets for User Requirements Analysis
Maria K. Wolters, Vicki L. Hanson, Johanna D. Moore · 2011 · The Proceedings of the 13th International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS 2011) · doi:10.1145/2049536.2049550
Summary
This paper demonstrates a novel methodology for using large-scale demographic surveys to inform the design of accessible technology. Rather than conducting specialized usability studies with small samples, the authors repurpose the English Longitudinal Study of Ageing (ELSA) — a broad survey of health, wellbeing, and socioeconomic status of English older adults — to derive user requirements for spoken dialogue systems (SDS). SDS are voice-based interfaces used in telecare and home care settings, increasingly relevant as populations age. The authors propose a three-step audit framework: an inclusion audit to identify who could benefit from the technology, a technology audit to determine what platforms are available to potential users, and an exclusion audit to estimate barriers that might prevent successful use. Working with a final dataset of 3,408 participants aged 65 and older, the researchers analysed ELSA data across four key dimensions of ability: dexterity, vision, mobility, and literacy. The methodology is designed to complement more specialised disability surveys like the Disability Follow-Up Survey by providing broader socioeconomic context about the lives and technology access patterns of people with specific ability limitations.
Key findings
The inclusion audit revealed that roughly one in four older people surveyed (30.6% of those aged 65+) had at least one limitation in dexterity, vision, mobility, or literacy that could make them candidates for SDS. Literacy and mobility were the most prevalent issues, with prevalence increasing sharply with age — mobility problems roughly doubled between successive age groups. The technology audit found that landline telephones (97.2%) and televisions (98.95%) were near-universal among the target group, while computers (42%) and internet access (29.95%) were far less common. The target group was also significantly less likely to own mobile phones. The exclusion audit identified hearing as the most significant barrier: almost half the target group reported hearing problems, and around 40% reported difficulty understanding speech in noise. Neurological conditions (stroke, dementia, Parkinson's disease) were also significantly more prevalent in the target group but comparatively rare overall. Cognitive function scores for memory and fluency were significantly lower in the target group. The authors concluded that SDS should be delivered via landlines and that multimodal alternatives should use television rather than computers, given their near-universal adoption.
Relevance
This paper offers a practical, replicable methodology for accessibility practitioners and designers who need to understand the characteristics of their target user populations without commissioning expensive bespoke studies. The three-step audit framework — inclusion, technology, exclusion — provides a structured approach to requirements analysis that could be applied to any technology and any demographic dataset. The finding that literacy, not vision, was one of the biggest barriers challenges common assumptions about older adult accessibility needs. For practitioners working on voice interfaces, telecare, or any technology aimed at older adults, the paper highlights the critical importance of designing for hearing loss and the need to consider what technology platforms users actually have access to rather than assuming internet connectivity.
Tags: inclusive design · aging · spoken dialogue systems · telecare · user requirements · design exclusion · hearing · dexterity · mobility · literacy · older adults