In ancient times among Celtic peoples, a seanchaí (anglicized shanachie) was a combination of a historian, poet, teacher, musician, and diplomat. The synergy of these talents was used to create and tell stories, often to illustrate a particular concept or idea.
In a similar way, our work in Data for Good today draws on diverse skills in both science and the arts to tell a story directing attention to an important issue and informing ways to address it. One of the most important skills for telling stories with data is effective visualization. Combining statistical science with visual arts and communication, data visualizations are essential for communicating our scientific findings to guide effective action.
Keep in mind the following for telling your story with data visualizations:
- Have a clear purpose for each visualization. One way to implement this is to annotate the outline for the paper or presentation, specifying a visualization for each point.
- Keep it simple. Avoid crowded visualizations—each image should focus on just one story point.
- Make sure there is a legend and caption and all axes in graphs are clearly labeled and described.
- Underscore important information, helping the reader see the focal point.
- Get a review. Ask a person unfamiliar with the content to look at the visualizations, ask questions, and offer comments.
A good place to investigate different visualization ideas is the leading papers in your area. While a literature search is often seen as looking for peer-reviewed content, it should also capture the most effective ways for communicating the subject.
In my own D4G experience, the US Census Bureau does a great job at producing clear, simple, information-rich visualizations—often on the same D4G topics I am investigating. They even have a resource library and gallery of data visualizations you can search for ideas, techniques, and inspiration.
When a data visualization is influential in developing your understanding of a topic and how best to present it, it is appropriate to give an acknowledgement in the paper or presentation. While citations capture specific knowledge content important to the work, contributions that influence in a general way can be recognized in the acknowledgements. As always, being generous with credit is the best practice.
One significant quality to make—or break—a data visualization is the use of color. In a visualization, it becomes a vital part of telling the story. Use color like words to provide information, direct attention, and communicate subtle, nuanced meaning. Learn the color wheel and use it when selecting colors.
Related ideas in a visualization benefit from similar colors. For example, using warm colors for wealth metrics and cool colors for population numbers. Complimentary colors, which are opposite on a color wheel, are particularly good for distinguishing contrasting variables.
Consistent use of color to signify the same items throughout a paper or presentation teaches the audience the color language you are using and makes the data story easier to follow. For example, one could use orange for predicted values and blue for actuals in predictive analytics—high contrast because the colors are complementary and consistently used throughout the work to mean the same thing.
Maps are one of the most effective tools for telling the data’s story. Maps are easily understood by all audiences, providing a powerful way to bridge the gap between data and understanding. Always include a legend and add notes on the map to focus attention on the most important information. Two maps side by side with study data in one and additional information in the other to provide context can be effective.
An excellent example of this is a pair of maps created by Alex Najibi. One shows the locations of high-definition surveillance cameras deployed in Detroit, Michigan, as part of a crime reduction program. A paired map has Census Bureau racial demographics, indicating more surveillance cameras in areas with a higher percentage of Black, Indigenous, and people of color population.
Learning practices for effective data visualizations can go a long way toward clarifying and strengthening the message you want to communicate, bringing out the D4G full story for maximum impact.
David Corliss
With a PhD in statistical astrophysics, David Corliss works as a data scientist in industry. He serves on the ASA Board as a Council of Chapters representative and is the founder of Peace-Work, a data for good nongovernmental organization.
Leave a Reply