Experiencing EVA Park, a Multi-User Virtual World for People with Aphasia
Julia Galliers, Stephanie Wilson, Jane Marshall, Richard Talbot, Niamh Devane, Tracey Booth, Celia Woolf, Helen Greenwood · 2017 · ACM Transactions on Accessible Computing · doi:10.1145/3134227
Summary
This study examines the user experience of EVA Park, a multi-user virtual world designed specifically for people with aphasia (PWA) to practice conversations. Aphasia, a communication disorder most commonly caused by stroke (occurring in 45% of stroke survivors), affects speaking, comprehension, reading, and writing abilities. People with aphasia often experience social isolation, shrinking social networks, and high rates of depression (over 50%). EVA Park was built on the OpenSimulator platform and co-designed with five people with aphasia to be both accessible and engaging. The virtual world includes realistic locations (town square, shops, café, restaurant, health centre) alongside fantastical elements (elephants to ride, mermaids, a giant rubber duck, a Doctor Who Tardis) designed to stimulate conversation and playfulness. Users control personalized avatars, communicate via microphone and headphones, and interact with environment elements through single clicks. Twenty participants with moderate aphasia used EVA Park for five weeks, accessing it from home with support from qualified speech and language therapists acting as "support workers." The researchers conducted thematic analysis of 37 video-recorded therapy sessions and 40 post-observation interviews, coding for affect, conversation types, miscommunication, immersion, social presence, initiative, and flow.
Key findings
Positive emotional experiences far outweighed negative ones: 165 coded incidences of positive affect versus only 40 negative, with the ratio improving over time. Pleasure, fun, playfulness, and joking were the most common positive experiences. Negative affect was primarily frustration related to word-finding difficulties inherent to aphasia, not the technology. Conversations were stimulated equally by virtual world activities (229 incidences) and real-life topics (222 incidences), demonstrating that the environment successfully sparked natural dialogue. Importantly, conversations about using EVA Park itself decreased over time as participants became proficient, while conversations about the virtual world's content increased. Social presence was highly valued: participants rated interactions with their support worker highest (mean 4.9/5 at both observations), and ratings for being with multiple avatars increased significantly from observation 1 to 2 (Z=2.44, p=0.015). By week 5, participants sought out others more often and avoided them less, and initiative-taking became more evenly distributed between participants and therapists. Critically, neither the technology nor the remote nature of interactions created significant communication barriers — language-related misunderstandings (mean <2 per session) were far more common than technology-related ones (<1 per session).
Relevance
This research demonstrates that virtual worlds can be successfully designed for and used by people with acquired communication impairments, challenging assumptions about technological accessibility for this population. For practitioners, EVA Park offers a model for creating engaging therapeutic environments that address both rehabilitation goals and the emotional/social consequences of aphasia. Key design insights include: combining realistic practice scenarios (ordering in a café, visiting a hairdresser) with whimsical elements that spark spontaneous conversation; using "sound parcels" to create private conversation spaces; providing gesture buttons for non-verbal expression; and including video screens showing real-world content as conversation starters. The co-design process with people with aphasia was essential to achieving accessibility. The findings suggest virtual worlds could deliver multiple rehabilitation models — from one-to-one therapy to group interventions and peer support. However, limitations exist: EVA Park has not been tested with people who have severe aphasia or significant cognitive impairments, and some fantastical elements might confuse rather than delight users with additional cognitive difficulties. On average, participants voluntarily spent 16.9 additional hours in EVA Park outside scheduled sessions, demonstrating genuine engagement beyond therapy requirements.
Tags: aphasia · virtual worlds · stroke rehabilitation · speech-language therapy · social interaction · user experience