Visualizing Urban Accessibility: Investigating Multi-Stakeholder Perspectives through a Map-based Design Probe Study
Manaswi Saha, Siddhant Patil, Emily Cho, Evie Yu-Yen Cheng, Chris Horng, Devanshi Chauhan, Rachel Kangas, Richard McGovern, Anthony Li, Jeffrey Heer, Jon E. Froehlich · 2022 · Proceedings of the 2022 CHI Conference on Human Factors in Computing Systems (CHI '22) · doi:10.1145/3491102.3517460
Summary
A design-probe interview study investigating how different stakeholders make sense of urban accessibility data through map visualizations. The authors built 24 map-based probes across seven map types — point, severity-weighted point, grid, heatmap, choropleth, street, and ego-centric isochrone — using Project Sidewalk's crowdsourced sidewalk-accessibility dataset for Washington DC (curb ramps, missing curb ramps, obstacles, surface problems). Twenty-five participants spanning five stakeholder groups — people with mobility impairments, caregivers, accessibility advocates, policymakers, and city department officials — took part in a three-part study: reacting to and interpreting the visualizations; performing find and compare sensemaking tasks (including comparing the 'accessible reach' of individuals with and without a mobility impairment via isochrones); and critiquing the map types and stating preferences. Through thematic and video analysis the paper characterizes the analytic tasks and data needs across stakeholders and organizes them into a multi-layered task model spanning macro goals (access to healthcare, jobs), mid-level strategies (navigability, connectivity, livability), and low-level tasks (sidewalk, POI, and transit accessibility). Its three contributions are that task and data characterization, an analysis of how individual differences shape sensemaking, and a set of ten design considerations for future interactive geovisual analytic tools supporting urban-accessibility communication, planning, policymaking, and advocacy.
Key findings
Sensemaking was driven by personal relevance rather than by a top-down overview: participants reasoned bottom-up from personally meaningful reference points, and the authors found that the classic 'overview first, zoom and filter, details on demand' model was not suited to these participants or tasks. Interpretation depended on three forms of familiarity — with accessibility (lived or professional), with maps, and with the city in question — and severity emerged as inherently subjective ('severity is in the eye of the beholder, or the eye of the traveler'), with accessible reach varying by an individual's mobility profile, pace, and functional status. A map's usefulness tracked how well it matched the user's existing mental model and the chosen unit of analysis (points vs grids vs neighbourhoods), surfacing the Modifiable Areal Unit Problem, where the aggregation choice changes both the insight and the trust placed in it. Street visualizations and ego-centric isochrones were the most preferred, for their experiential, on-the-ground feel. Trust rested on data provenance and analytic provenance — where the data came from and how scores were modelled — and on being able to drill down to the underlying data (the 'disclosure' principle). The headline implication is that no single map serves all stakeholders or tasks; the work argues for tools that support varied, sometimes conflicting needs and the construction of audience-specific views.
Relevance
A strong, recent, top-venue anchor for the state of accessible-mapping research — and a revealing one, because it is in a sense the inverse of accessible cartography for non-sighted users. It studies how to visualize accessibility data for (largely sighted) stakeholders making planning, policy, and advocacy decisions; by the authors' own admission in their limitations, the visualizations themselves were not designed to support people with different visual abilities, and accessible visualization is deferred to future work. The disability lens throughout is mobility impairment plus sighted professionals, not blind or low-vision users — so the population and the rendering problem that accessible spatial-cognition work centres on are precisely the gap this paper leaves open. Several findings transfer directly: the failure of 'overview first' for these tasks mirrors, from the visualization-research side, the modality-level point that a non-sighted user has no parallel channel for an overview and must build understanding outward from a chosen reference point; the documented weight of individual differences and subjective severity is empirical grounding for adaptation driven by user capability and preference; and the authors' open question — whether one configurable tool could serve conflicting stakeholder needs — is a natural hook for a CISNA-plus-capability-model answer. Useful both as field-critique evidence and as a framework to position non-sighted spatial-cognition work against.
Tags: accessible maps · spatial cognition · data visualization · geovisualization · sensemaking · urban accessibility · sidewalk accessibility · multi-stakeholder · design probe
Standards referenced: ADA