The last article in this series will examine the most common types of messages that require business graphs, along with the best types of graphs for each of these messages.
Express and Explain
Too often, data presentations try to impress rather than express and entertain when they should explain. When you don't know what you're saying, you can always hide the fact by dressing up the presentation in flash and dazzle. Generally, data presentations filled with meaningless fluff result less from not knowing what to say than from not knowing how to say it. Software that produces business graphs makes it far too easy to decorate your message with distracting visual content. Tufte calls this "chartjunk" visual content that provides no real information and is therefore distracting and sometimes downright misleading. The most common example of chartjunk is the inclusion of grid lines in graphs often, grid lines that are dark and heavy. Grid lines are rarely useful, and prominent grid lines are especially distracting. The purpose of a graph is not to provide a means to interpret the precise value of each bar, line, or data point. Instead, the purpose is to see the shape of the data, and from that shape discern meaningful patterns, such as trends and exceptions.
Take a look at the top graph in Figure 3. It was taken from the November 24, 2003 issue of Newsweek. Notice how the cute images, the abundant text throughout the graph, and even the background gets in the way of seeing the shape of the data and discerning its message. Now look at the alternative that I've provided on the bottom, where the fluff has been removed and the presentation has been designed to directly and clearly support the message.
FIGURE 4 Example of chartjunk in the graph on the top vs. the same data minus the fluff on the bottom.
When designing business graphs, more is definitely not better, unless the more is meaningful data that supports the message. The author Antoine de St. Exupery once wrote: "In anything at all, perfection is finally attained not when there is no longer anything to add, but when there is no longer anything to take away." These are wise words of a great communicator. In the third installment of this series, I'll examine a sequence of design steps that will help you remove all distractions from the message and then make sure that the most important parts of the message stand out above all else.
Know, Don't Guess
Effective graph design isn't always an intuitive process. When you create a graph, you design something that communicates through visual perception. Knowing something about how we perceive and interpret visual stimuli objects made visible through light, possessing a particular combination of attributes such as color, size, and location in space is a necessary conceptual foundation to effective graph design. Which visual attributes can reliably encode quantitative data? Which visual attributes can be used to make some things stand out above others? Which colors work well together? What are the limits of short-term memory, and how do limits apply to graph design? Why is it difficult to accurately compare the sizes of the slices in a pie chart?
Many years of solid scientific research have provided answers to these questions and many more that are relevant to graph design. You don't need to be a scientist yourself to become familiar with these concepts and how they can be used to present data effectively.
The common mistakes that I've touched on in this article are only a sampling of a much larger list. You can probably identify many more without straining your brain in the least. My purpose, within the constrained space of this article, is simply to get you thinking about the importance of effective data presentation and noticing how some of your own design practices might need improvement. Stick with me through the three remaining articles in this series to continue honing your skills in effective data presentation.
Stephen Few [[email protected]] is the founder of Perceptual Edge, a consulting firm that specializes in information design for analysis and communication. His new book, Show Me the Numbers: Designing Tables and Graphs to Enlighten, is now available from Analytics Press.
Tufte, E. R., The Visual Display of Quantitative Information, Graphics Press, 1983.
"The Information Cannot Speak for Itself," July 10, 2004.