Beyond the Basic Chart

Duke experts share best practices when creating charts and graphs

Presenting data can be as simple as plugging numbers into a bar or line graph, pie chart, or map. But the colors, details and storytelling can drastically alter how data is perceived.

“People don’t always treat their visualizations in the way they treat their writing,” said Eric Monson, a data visualization analyst with Duke’s Data and Visualization Services. “In writing, we’re very used to having it be a multi-stage process of a rough draft and editing and really crafting it to be as effective as possible.”

When creating charts and graphs, keep it simple. Avoid using effects such as 3-D shapes, drop shadows and gradients because they can distort the data.

“Everyone wants their work to be noticed, so they’ll try to make it really flashy to gather attention,” Monson said. “They just don’t understand that in the end, those effects get in the way of their message.”

Also, free up color. Try to use the same color for all data points and then use a different color to accent an important piece of data, said Duke data visualization coordinator Angela Zoss. Don’t use colors that are too bright or clash with the background of a presentation or publication.

And most importantly, tell a story. If a line chart is showing that Duke hiring peaks in July, make a notation as to why. In Duke’s case, July is the start of the new fiscal year. 

Zoss said to keep in mind the variables in your data that are most exciting, whether it be zeroing in on a time period or population segment.

“Always think from the beginning what the purpose of the visualization is,” Zoss said.

Want to learn more?Duke Libraries’ Data and Visualization Services Department can help with all phases of a data visualization project, from finding and cleaning data to using the right software and design. The department hosts free workshops during the year. Email askdata@duke.edu or visit library.duke.edu/data.