Designing Accessible Visualizations for People with Disabilities
Visual representations are not accessible by everyone.
Visualizations represent data in easy-to-understand forms. Effective visualization is one that tells a story about the data, highlighting patterns and outliers, and making it easier for readers to comprehend the information.
Unfortunately, visualizations are not accessible and readable by everyone. This is a very common problem especially for people with disabilities (PWD). A badly-designed visualization can be extremely frustrating. For example, too much data can overwhelm readers, while inappropriate representations can confuse them. But an inaccessible visualization can be worse as it can exclude certain users. For example, color-blind users may not be able to read visualizations that heavily rely on the use of different colors.
Why is accessibility important? There are two main answers: accessibility supports the social inclusion of a wide range of people, including PWD, and, morally, it is the right thing to do.
In the paper “Sociotechnical Considerations for Accessible Visualization Design”, Alan Lundgard, Crystal Lee, and Arvind Satyanarayan discusses how to address the need for creating more accessible visualizations for PWD. Based on their research with blind students, the authors provide sociotechnical considerations for designing accessible visualizations. They emphasize the importance of involving PWD throughout the design process as they are better informed about their experiences and needs than anyone else. By having PWD as co-designers, researchers are not biased by their own assumptions and learn directly from the users. This way researchers can also move away from designing technologies and visualizations that are well-meaning but may have unintended consequences.
To motivate the sociotechnical considerations for designing accessible visualizations informed by the authors’ work, the background research of this paper brings together disability studies, research on tactile information systems, and participatory design methods.
When designing for PWD, researchers must distinguish between the two models of thinking about disability. The medical model ties to the person’s physical or psychological state and focuses on curing the disability. For example, a deaf person can use cochlear implants to manage their disability. The social model, on the other hand, considers impairments as physical or psychological abnormalities and describes disability to be the exclusion people face due to their impairments. The social model drives the adoption of assistive technologies and tries to address the infrastructural barriers experienced by PWD.
Tactile Information Systems
Braille displays, 3D models, and embossed maps are different tactile systems that are used for blind education. While designing tactile systems, researchers must understand how different users absorb and make sense of information conveyed to them. For example, while sighted people see the whole picture, blind users put together the information by sequentially touching individual parts of the graphic. Visualization designers must ensure that each tactile element is quickly distinguishable and information is easy to access sequentially.
Participatory Design Methods
Researchers must be attentive to the power dynamics between themselves and their study participants. Participatory research methods encourage researchers to work with their participants rather than on them. This way participants go beyond being study subjects as they become co-designers who contribute to different parts of the design process including problem defining, data collection, and analysis.
The authors present a set of sociotechnical considerations for designing accessible visualizations grounded in their work with blind students.
- Non-Intervention: Researchers should consider whether a technological intervention is necessary at all. For PWD, well-meaning interventions can often worsen the situation they are intended to help with.
- Research and Design: Researchers should carefully consider the tension between research and design when designing for and with PWD. While the goal of the research is to create new knowledge, it might not address the immediate needs of the users. On the other hand, design can satisfy those needs but the solution may lack research novelty.
- Participatory Methods: Researchers should involve users as equal participants throughout the design process and not just for evaluating the solution using user studies. Participatory approaches are key to inclusive design as they reduce the distinction between who is doing the research and for whom the research is being done. The research and design tools should be accessible to the participants.
- Communicating Expectations: Researchers should communicate clearly to their collaborators about the intentions, expectations, and capabilities of their project. This includes the project’s goals, intended duration, and available resources to support it. For academic publications, author credit and order should be discussed. For marketable prototypes, equitable compensation and intellectual property should be discussed.
- Time and Compensation: Researchers should be sensitive and respectful of all collaborators’ time. PWD often face additional barriers with mobility and access, making the time they spend on a design project especially valuable. They should always be compensated for their specialized skills such as the ability to read braille or use a screen reader.
- Accessibility Guidelines: Researchers should be familiar with various accessibility guidelines and adhere to them while designing technology. Best practices should be integrated into solutions as relevant to the design problem.
- Technology Access: Researchers must be mindful of technology access that PWDs have and use design approaches that fit within their constraints. A low-tech approach may be more readily available to users on a daily basis as compared to a high-tech approach which yields more sensational results.
- Technology Resolution: Researchers should ensure that technologies encode information effectively. This is especially important for web visualizations which may not be screen reader compatible and for braille as it requires standardized height and spacing, and cannot be resized.
Researchers should draw on these considerations for creating accessible visualizations. They should move away from design practices that don’t actively involve PWD in the process. This can lead to many exciting opportunities for collaboration between disability, assistive technology, and visualization communities.
This blog post is based on the following publication:
- Alan Lundgard, Crystal Lee, Arvind Satyanarayan. Sociotechnical Considerations for Accessible Visualization Design. In Proceedings of the IEEE Conference of Information Visualization, 2019.