Effects of Extended Use of an Age-friendly Computer System on Assessments of Computer Proficiency, Attitudes, and Usability by Older Non-Computer Users
Joseph Sharit, Jerad H. Moxley, Walter R. Boot, Neil Charness, Wendy A. Rogers, Sara J. Czaja · 2019 · ACM Transactions on Accessible Computing (TACCESS) · doi:10.1145/3325290
Summary
This paper examines the impact of 12 months of extended use of PRISM (Personal Reminder Information and Social Management), a computer system specifically designed for older adults at risk of social isolation, on computer proficiency, attitudes toward computers, and perceived usability. The study placed computers with the PRISM software into the homes of adults aged 65 to 98 who had minimal or no computer skills and no computers in their homes — a population at significant risk for social isolation and digital exclusion. PRISM was designed with age-friendly principles: simplified interface, large fonts, minimal jargon, curated content areas (email, calendar, games, photo sharing, community resource directory, and a classroom feature), and in-home face-to-face training over several days. The study drew from a larger randomized controlled trial (N=300) with 172 participants in the PRISM condition, of whom 132 had complete data. Participants were predominantly female (73%), with a mean age of 76.4 years, living alone (72%), and ethnically diverse (41% Hispanic, 32% non-Hispanic Black, 27% non-Hispanic White). A growth mixture model applied to usage data over the 48-week period identified two statistically distinct subpopulations: more-frequent users (59%) and less-frequent users (41%). Critically, these groups did not differ at baseline on computer proficiency or attitudes — the differentiating factor was fluid cognitive abilities, with more-frequent users having significantly higher cognitive scores.
Key findings
After 12 months, more-frequent users showed significantly higher computer proficiency (CPQ), more positive computer attitudes across all three subscales (interest, comfort, efficacy), and higher usability ratings compared to less-frequent users — despite starting from equivalent baselines. Path analysis revealed that system use strongly predicted perceived usability (PSEQ), which in turn predicted computer proficiency, which then predicted computer attitudes. Age did not directly predict system use, challenging assumptions that the oldest participants would use the system less. However, age negatively predicted usability perceptions, which mediated its effect on proficiency and attitudes — meaning age's impact operated through usability concerns rather than disinterest. The bivariate change score models showed that total usage predicted changes in proficiency, and changes in proficiency predicted changes in all three attitude subscales, supporting a causal pathway from use to proficiency to attitudes. Notably, 18 of 90 participants (20%) still did not feel comfortable using PRISM after 12 months, and the calendar feature received the lowest usability ratings, indicating room for improvement even in a purpose-designed system. Hardware difficulties were rated more problematic than software difficulties.
Relevance
This research provides longitudinal evidence that well-designed, age-friendly technology can meaningfully change older non-computer users' self-assessed proficiency and attitudes, potentially serving as a gateway to broader technology adoption. For accessibility practitioners and organizations designing technology for older adults, several findings are directly actionable. First, usability is the critical mediator: perceptions of usability drive initial engagement, which drives proficiency, which drives attitudes — meaning investment in user-centered design and iterative usability testing for this population has cascading benefits. Second, cognitive abilities predict usage frequency, suggesting that systems for older adults with lower cognitive abilities need to be "extremely usable" and fault-tolerant, minimizing demands on perceptual speed and memory. Third, in-home, face-to-face training was essential but insufficient for 20% of participants, highlighting the need for ongoing support structures. The study's diverse, low-income, socially isolated sample makes its findings particularly relevant to addressing digital equity — these are precisely the older adults most likely to be excluded from technology's benefits and most in need of social connection tools.
Tags: older adults · computer proficiency · social isolation · usability · age-friendly design · technology adoption · digital divide · longitudinal study · cognitive abilities