Does Your Viz Pass the Eye Candy Test?
The use case for data visualization is simple. Whether it's a single visualization, a dashboard, or even a data story, when layered upon the foundation of careful statistics and good data, a well-balanced data visualization can deliver meaningful, immediate, and actionable insights at a glance. When done correctly, using visualization for insight is intuitive and it's easy; it doesn't just give you the data you need, it shows you what you need to know about it, too. Through the careful distillation of images, color, and design, a well-crafted data visualization can leverage the visual horsepower of the human brain to spur decision-making, improve analytical research, or act as a beacon for effective communication. Good data visualization should be fast, informative, and--above all--valuable. This makes data viz a critical tool in the modern analyst toolkit.
Although based in data and statistics, it's hard not to think visually when it comes to data visualization. Half science and half art, visual information representation is an artful process, with the end results delivered via colorful and content-rich charts and graphs; carefully crafted dashboards; compelling, data-driven narratives; or even astute infographics.
But visualization comes with its own set of risks. Getting too carried away with the artful, visual aspects of data visualization can dilute--or worse, distort--the data's meaning. What could be extremely insightful presentations of data can become little more than pretty colors cleverly disguised as information visualization. While a good data visualization is often an aesthetically pleasing one, beauty itself does not equate to analytical prowess. Thus, any visualization attempt should not forget the cardinal rule of data visualizaton: Above all else, show the data. The key to curating a "well-balanced" data visualization isn't just providing a visual representation of data; it's providing the right kind of visualization for the data and visualizing it with intent by using the art to support the data's story and not overshadow it.
Here are three simple questions to ensure your data visualization passes the eye candy test:
Is it approachable?
First, make sure the visualization is straightforward and easy to understand by its intended audience. Then, capitalize on the fact that people perceive a design as easier to use if it includes design elements (color, shapes, etc.) to make it visually appealing and leverages pre-attentive features. This is visual design, the practice of removing and simplifying things until nothing stands between the message and the audience. In visualization, the best design is the one you don't see because you're too busy looking at the data.
Does it tell a story?
A visualization should tell a story about--or explain--its data. Therefore, visualizations require a compelling narrative to transform data into knowledge and emotion into action. Make sure your visualization has a single story to tell. Too often, people want to present all the data in a single visualization that can answer many questions, but effective visualizations are a one-visualization-to-one-story ratio. Your audience needs answers, not more questions.
Is it actionable?
Does the visualization provide guidance through visual clues that prompt action? Visualizations should leverage visual clues (colors to highlight or alert, or annotations to narrate important information) and dashboards should establish a visual hierarchy to direct the audience's attention. This is the "fitness" test: Before you even know what the numbers say, your audience's eyes should know where to go to decode the information being presented.
If you can answer "yes" to all three questions, chances are good that your visualization is a well-designed, meaningful, non-eye candy data visualization that leverages colors, shapes, and design to not only display but influence the way your audience receives insights into data. If it doesn't, this test should help you identify where you need to go back and spend a little more time perfecting your viz.
Based in the greater New York City area, Lindy Ryan researches and teaches business analytics and data communication at a major East Coast university, and is the author of The Visual Imperative: Creating a Culture of Visual Discovery. Follow her on Twitter @lindy_ryan.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||BIG DATA BASICS; visualization|
|Publication:||Big Data Quarterly|
|Date:||Mar 22, 2019|
|Previous Article:||Repeatable Machine Learning With Kubeflow.|
|Next Article:||The Real Dragon in the Room: What happened to the 500 million datapoints from the Starwood data breach?|