Data visualization is not just about the data. It starts with visual storytelling and grammar. Yes, there is a basic grammar of graphics – selecting the right chart form and choosing the right color to display the data clearly. In writing, we start with a compelling story – a clear message – and then we apply proper grammar and create beautiful prose. The invention of the word processor didn’t turn everyone into a writer.
"It takes both data strategy and design to turn volumes of information into clear, concise and intelligent graphics"
The question I often get is: What software should I use? Of course, all data visualizations are implemented with some sort of software. Let me be clear – that’s the final step and not the first. If we start with applying all the tricks that a charting software package offers, we can easily lose sight of the message. When I studied data visualization at Yale under Edward Tufte, a pioneer in data visualization, more than two decades ago, we started with pencil sketches. We focused on concepts. That’s still the right approach.
With writing we learn the alphabet long before we are asked to write an essay. However, with charting, we are given this magic program on our computer and we are expected to make graphics that are intelligent with no training. That’s where trouble begins.
It takes both data strategy and design to turn volumes of information into clear, concise and intelligent graphics. So what is data strategy? Data strategy is about identifying a message. It’s about telling a good story. For example, a heat map of mortgage delinquencies can tell us who borrowed money, where do they live and can they pay it back. This story helps homeowners make informed decisions about the housing market.
In any story, content comes first. The three essential elements of good information graphics are:
“Rich content brings meaning to a graphic. Inviting visualization interprets the content and highlights the essence of the information for the reader. Sophisticated execution brings the content and the graphics to life.” - excerpt from The Wall Street Journal Guide to Information Graphics.
So how much content to display? A critical step to a successful piece of data visualization is distilling massive amounts of complex raw data into insightful information. For example, a school report with charts demonstrating the student’s performance can help teachers customize the lesson plans. When these graphics are done well, they connect with our audience and motivate them to action.
An information graphic is not a form of self expression. It always has a targeted audience. From my days as graphics editor at The New York Times and head of graphics at The Wall Street Journal, I learned my audience is the most important group of people in my work. Having millions of readers means having millions of critics on a daily basis. Understanding my audience helps shape my graphic. In fact, some graphics are literally a matter of life or death. For example, a lab report with clear illustrated graphics about blood pressure and cholesterol level can motivate patients to exercise more and change their diet. However, if the graphics are overly complicated, the report is not only useless, it is a disservice to the patients. We have to take the perspective of our audience and approach the graphic solution with conviction and empathy, as well as accuracy.
Likewise, the choice of color is about clarity rather than style. Color should only be used to differentiate the hierarchy of information. Everyone can perceive the difference in shading, but they may not differentiate certain colors. “Color combinations such as red/green or blue/yellow…can be similar in value or lightness. The lack of contrast in lightness makes it virtually unreadable for color-blind users.” - excerpt from The Wall Street Journal Guide to Information Graphics. And yet, we see these faulty color combinations all the time in graphics. The key is to develop a charting color palette with the least number of colors, but with multiple shades of the same hue.
Illustration from The Wall Street Journal Guide to Information Graphics.
According to the National Institutes of Health, about 1 in 10 men have some form of colorblindness. Your senior manager may be one of them. Do you want your message to get lost?
Many designers and technologists over-complicate the design and concept of data visualization. As one of my mentors Alan Siegel – a longtime leader in branding and a driving force behind the plain English movement – says, “Simplicity works – in business, in government, in life.” My philosophy of data visualization is straightforward: Know the content, distill the information and put my audience first.
The views expressed here are her own and do not necessarily represent those of the Federal Reserve Bank of New York.