QUARTZ makes qualitative data visualizations accessible for blind researchers

QUARTZ (Qualitative Understanding via Accessible Representation and Visualization) is a new web-based system from researchers Omar Khan and JooYoung Seo, accepted at ASSETS 2026, that tackles a long-overlooked accessibility gap. While quantitative charts have seen multimodal accessible solutions, qualitative data visualizations—concept maps, network graphs, Sankey diagrams, and coding stripes—remain entirely inaccessible to blind and low-vision (BLV) researchers. These non-linear, semantically rich structures are central to qualitative inquiry, yet no prior work addressed their accessibility. QUARTZ provides screen-reader-accessible, multimodal representations that allow BLV users to navigate and comprehend these complex diagrams. Using the Rapid Iterative Testing and Evaluation (RITE) method, the team conducted a user study with 8 BLV participants who completed 12 tasks across all four visualization types.

The study exposed accessibility barriers unique to qualitative visualizations: non-linear navigation breakdowns (where users lost context moving between connected nodes) and semantic comprehension gaps (where spatial relationships like hierarchy or flow weren't conveyed). Through iterative co-design, these barriers were resolved, producing empirical evidence and concrete design guidelines for an underexplored visualization domain. QUARTZ advances the infrastructure needed for BLV researchers to independently participate in qualitative research, from coding interviews to building theories. The paper includes design recommendations that can inform future accessible visualization tools, potentially extending to other non-linear data representations in education and data journalism.