Working Papers
“Community Playtesting CardioCare Quest Among Adults with Disabilities at the Hozhoni Foundation”
Designing playful health technology for Indigenous adults with disabilities requires more than an accessible interface. This paper deploys CardioCare Quest, a community co-designed hypertension intervention, with adults at the Hozhoni Foundation in Flagstaff, Arizona. We characterize the cardiovascular risk profile of participants through pre-play measures of blood pressure, BMI, and resting heart rate; observe how players engage physically with the Dog Walking game through step counts and heart-rate response; and report shifts in hypertension knowledge and self-monitoring attitudes after open community play. The findings point toward design implications for playful health technologies that take seriously the lived experience of intellectual and developmental disability in rural Indigenous communities.
“An Interactive, Accessible Visualization Platform for Systematic Review of Playful Health Technology Research”
Systematic reviews in accessibility and serious-games research are typically static, text-heavy, and difficult to navigate for the disabled researchers most affected by their findings. We present an interactive web platform that organizes 215 peer-reviewed papers (1994–2023) along nine experiential design dimensions drawn from the Playful Health Technology framework. The platform pairs coordinated spectrum heatmaps, dot matrices, and filter views with full keyboard operability, screen-reader compatibility, audio-tactile sonification, and WCAG 2.1 AA contrast. In a pilot study with eleven HCI researchers, nine preferred the heatmap view over conventional bar charts, and participants completed exploratory synthesis tasks in roughly half the time of equivalent manual review. We argue that accessible visualization is not a downstream concern but a prerequisite for equitable knowledge synthesis in our field.
“Applying Accountability and Transparency to an LLM-Powered Synthesis of Game-Based Research in ASSETS”
Large language models are increasingly used to support literature synthesis, but their fitness for subjective, experiential annotation in accessibility research remains poorly understood. We deploy a four-model ensemble (GPT-5, Gemini Pro, Llama 3, Mistral Large) against 215 peer-reviewed papers on serious games for accessibility, comparing their answers to a human consensus baseline of 1,720 annotations from HCI researchers. Our retrieval-augmented prompting pipeline requires each model to self-report confidence, cite the source chunks supporting its answer, and iteratively expand its context through four decision modes. We find that models can match human annotators on some questions but fall short on more experiential ones, and that decision mode and confidence correlate meaningfully with answer correctness. We offer the pipeline as a tool researchers can use to apply LLMs in their own work while preserving fairness, accountability, and transparency.
Work in Progress
“Mapping Three Decades of Experiential Design in Playful Health Technology: A Systematic Review with Interactive Visualization”
“Interactive Visualization of Thirty Years of Accessibility and Serious-Games Research”