On July 28, 2021, the Data Visualization Society hosted a Fireside Chat to discuss accessibility in data visualization. I moderated a panel of community members who represented a wealth of diverse perspectives. The panelists were: Frank Elavsky, Data Visualization and Accessibility Design Systems Engineer; Sarah L. Fossheim, Multidisciplinary Developer and Designer; Øystein Moseng, Head of Accessibility and Core Developer at Highcharts; and Léonie Watson, Director and Accessibility Engineer at TetraLogical.
What follows is an abbreviated summary of the chat. You can watch the full discussion on the DVS YouTube channel or at the end of this article.
What does accessibility mean in data visualization? Although it can be challenging to define precisely, the panelists emphasized that accessibility means thinking beyond the visual, despite the fact that our field is known as data visualization. This could mean, for example, sonifying data, as Carni Klirs did using COVID data from The New York Times, or as the podcast Loud Numbers has done. Or, as explained by Léonie, who is blind, it could mean ensuring that users like her have access to a “comparable experience” of the data—perhaps by including a data table that allows her to take in complex data more easily. User testing is always critical, as is involving people with varied disabilities in the creation of data experiences.
As for the future of accessibility in data visualization, the panel agreed that better browser support is essential, but so too is a perspective shift among designers: Thinking about accessibility can lead to visualizations that are more fun and interactive! Data visualization designers often approach accessibility from the perspective of compliance, which can mean dread and negativity, but accessibility is in reality an extension of what we are already trying to do, what we already think is fun to do—finding creative ways to communicate data through experiences.
Improving accessibility in data visualization does depend on people sharing their knowledge of the topic. If you know about accessibility, talk about it at conferences! Talk about it within your organization! Design for your future self, who may have different needs as you age (check out Adrian Roselli’s talk on Selfish Accessibility for more on this idea). Know that accessible designs often benefit more people than you realize, including those who may not “need” accommodations but appreciate them nonetheless. (This is the “curb cut effect.”)
And if you’re just getting started on thinking about accessibility, the panel had some tips:
- Set aside time in your workflow specifically for thinking about accessibility.
- Use existing tools to learn and improve your visualizations: for example, the Chartability framework developed by Frank, any of the tools linked on this DataViz Accessibility resource page, the free edX Introduction to Web Accessibility course, or the short accessibility tests developed by TetraLogical.
- Share your learning process to help others and to push the field to improve. (One example: Amy’s article on writing alt text for data visualizations. Another example: Chris DeMartini’s blog posts on his struggles with creating accessible visualizations.)
- Ask questions or start a conversation on the DVS Slack #topic-accessibility Slack channel! With enough people speaking up, with enough momentum, accessibility in data visualization will improve.
Check out the full conversation below: