Improving Low-Vision Chart Accessibility via On-Cursor Visual Context

Paper Title

Improving Low-Vision Chart Accessibility via On-Cursor Visual Context

Publication Info

• Topic area: Enhancing chart accessibility for low-vision individuals using visual context overlays.
• Keywords: Low vision, chart accessibility, assistive tools, focus+context, overview+detail, Dynamic Context, Mini-map, visual impairments, usability, accessibility design.

Background and Problem

• Problem / challenge: Charts are largely inaccessible to low-vision individuals (LVI) due to their reliance on magnification tools, which create a trade-off between accessing fine-grained details and maintaining global context. Existing assistive tools (AT) fail to adequately address this issue.
• Significance: Improving chart accessibility for LVI is critical for information equity, enabling independent data interpretation and reducing cognitive and physical effort.
• Motivation and related work: Prior work has focused on non-visual access methods (e.g., screen readers, sonification) or tactile interfaces, often grouping blind and low-vision users together. However, LVI prefer using residual vision. Current magnification-based ATs lead to fragmented viewing experiences, requiring excessive panning and mental mapping. This study builds on these insights to explore visual context delivery for LVI.

Solution

• Proposed approach: Two interaction methods—Dynamic Context and Mini-map—designed to overlay critical chart context (axes, legend, grid lines, and overview) around the user’s pointer.
• Novelty:
  1. Introduction of Dynamic Context, inspired by focus+context paradigms, to project chart elements around the pointer.
  2. Adaptation of Mini-map, based on overview+detail principles, to provide a minified chart view with pointer position indication.
  3. Integration of both methods with existing AT to enhance usability and reduce effort.
  4. Design insights for future AT development to balance visual access and clutter.
• Procedure and key techniques:
  1. Conducted a formative study with five LVI to identify key contextual elements and challenges in chart reading.
  2. Developed web-based prototypes of Dynamic Context and Mini-map with customization options for LVI.
  3. Evaluated the prototypes in a comparative study with 22 LVI using the Mini-VLAT chart comprehension test across three conditions (baseline, Dynamic Context, Mini-map).

Results

• Concrete findings:
  • Dynamic Context significantly improved usability (SUS score: 76.36) and reduced physical effort compared to baseline and Mini-map.
  • Mini-map provided spatial orientation but was less effective for detailed tasks.
  • Both methods enhanced perceived access to charts, with Dynamic Context rated highest for focus and compatibility.
• Advantage over baselines:
  • Dynamic Context reduced reliance on panning and improved access to chart elements like axes and legends.
  • Mini-map offered limited benefits for orientation but was less preferred overall.
• Experiments / evaluation:
  • Participants (N=22) used the prototypes with their preferred AT in a controlled study.
  • Metrics included test scores, completion times, usability (SUS), task load (NASA-TLX), and qualitative feedback.
  • Dynamic Context centralized pointer activity and reduced navigation complexity, while Mini-map had minimal impact on strategies.
• Limitations and future work:
  • Visual clutter in Dynamic Context hindered users with double vision.
  • Mini-map’s small size limited its utility.
  • Future work includes automating chart context extraction, addressing higher-level context (e.g., trends), and conducting longitudinal studies to assess long-term usability.

Summary

This study addresses chart accessibility for low-vision individuals by introducing two interaction methods: Dynamic Context and Mini-map. Dynamic Context projects chart elements around the pointer, improving usability and reducing effort, while Mini-map provides spatial orientation. In a study with 22 participants, Dynamic Context significantly enhanced perceived access and usability, though it introduced visual clutter. Design insights emphasize the need for balancing visual access, reducing complexity, and adapting to user needs. These findings contribute to advancing assistive tools for LVI and have potential applications in other visual domains.