Design Guidelines and Recommendations for Multimodal, Touchscreen-based Graphics
Jenna L. Gorlewicz, Jennifer L. Tennison, Hari Prasath Palani, Nicholas A. Giudice · 2020 · ACM Transactions on Accessible Computing · doi:10.1145/3403933
Summary
This paper presents comprehensive design guidelines for creating accessible graphics on touchscreen devices using vibration and audio feedback for people with visual impairments. Drawing on over eight years of research, the authors address a critical gap: while touchscreen devices have become ubiquitous, graphical content remains largely inaccessible to blind and visually impaired users. Traditional solutions like embossed tactile graphics require specialized equipment, are time-consuming to produce, and cannot be easily updated or distributed digitally. The guidelines cover four main areas: graphical element design (minimum line widths, spacing requirements), feedback assignment (mapping vibration to spatial information and audio to semantic content), hardware adaptations (tablet size, physical reference markers, stabilization), and user exploration strategies. The authors recommend specific parameters based on empirical testing: lines must be at least 1mm wide for detection and 4mm for comfortable tracing, with 4mm minimum gaps between objects. For vibration feedback, they suggest limiting to 3-5 distinct profiles and assign frequencies of 2.5 Hz for dashed lines, 10 Hz for dotted, and 50 Hz for solid lines. The paper demonstrates applying these guidelines to bar charts and geometry figures as case studies, implementing solutions on Android tablets with native vibration and text-to-speech APIs.
Key findings
A user study with 22 participants recruited from the National Federation of the Blind convention compared multimodal touchscreen graphics against traditional embossed tactile graphics. The key finding was that performance was statistically equivalent: participants achieved 80.09% accuracy on touchscreen graphics versus 81.83% on embossed versions, with no significant difference between mediums (p = 0.269 for bar charts, p = 0.718 for geometry figures). Participants performed better on geometric figures (85-86% accuracy) than bar charts (77-82%) across both mediums, suggesting that spatial reasoning tasks may transfer more readily to digital formats. Heat map analysis revealed efficient exploration strategies—users focused on corners and key points rather than tracing entire shapes. Participants appreciated the immediate audio feedback for labels on bar charts, though some found text-to-speech at 2x speed too fast. The study also identified practical challenges: double-tap gestures were difficult to trigger reliably, and strong finger pressure caused tablets to slide during exploration, requiring physical stabilization with rubber stickers.
Relevance
This research validates touchscreen devices as a viable platform for accessible graphics, with performance matching traditional embossed materials that have been the standard for decades. For accessibility practitioners, the specific technical parameters (line widths, vibration frequencies, spacing requirements) provide actionable implementation guidance. The finding that digital graphics can match physical tactile graphics is significant for educational accessibility—it means accessible STEM materials could be distributed instantly rather than requiring specialized embossing equipment. The guidelines have direct implications for developers creating accessible data visualizations, educational apps, or any touch-based graphical interfaces. However, limitations include the study's focus on relatively simple graphics; scaling to complex charts, maps, or diagrams remains an open question. The reliance on device-native vibration motors also means implementation varies across hardware. Future work extending these guidelines to more complex visualizations would significantly expand their practical application.
Tags: tactile graphics · touchscreen accessibility · haptic feedback · multimodal interaction · STEM accessibility · visual impairment · vibrotactile feedback