STEP INTO THE WORLD OF TECHNOLOGICAL INSIGHTS AND INNOVATION

6 Ways You Can Improve Data Visualization

Sep 09, 2016

6_Ways_You_Can_Improve_Data_Visualization_IA.jpg

Charts and graphs have become embedded in our lives, in the way we manage our businesses, consume news, learn, and buy. Marketers, political parties, news agencies and business data analytics use charts because they command our attention. We are almost twice as likely to read a news story or blog article if there is a chart on the page.

In a new book, Good Charts: The HBR Guide to Making Smarter, More Persuasive Data Visualizations, Scott Berinato gives us the researcher’s view of why we react to charts the way we do.

The Power of Visual Messages 

6_Ways_You_Can_Improve_Data_Visualization_IB.pngHumans are visual beings. We absorb and understand a visual message in a few milliseconds that would take a paragraph of text or a long conversation to explain. We develop our understanding in the context of  
assumptions of which we may not be aware until we stop to think about them.

North is up; south is down. Up is good; down is bad. Time moves right. Right is good; left is bad. See how confusing the 9-box is when we reverse the directions.

Colors matter in more ways than we can imagine (See Figure 1). Green is good and red is bad, unless red is hot. If it is, according to Berinato, then blue is cool. Orange is dominant,[1] which is why online marketers often use it to induce us to take action.

6_Ways_You_Can_Improve_Data_Visualization_IC.pngIn our minds, we group items of like colors together, and assume contrasting colors show differences (See Figure 2). In this the example from Good Charts, we see that the similar blue colors deceive us into thinking that “Not at all  interested” and “Extremely interested are similar responses.” Not only that – the dominant orange color causes us to assume the “Moderately interested” response is the most significant.[2]

 

Other Visual influences

Color is the most powerful influence in visualization, but it is not the only one. The shape and relative size of objects influence our thinking, as does their position relative to each other. By changing the scale of a chart, you can present a very different picture. The best example we can think of is how news outlets use charts with very narrow range to dramatize tiny changes in stock prices.

With all of these things to consider and many types of charts, it is easy to understand why charts we see sometimes make little sense. You can follow all the directions in your charting program and fail, just as you can follow all the rules of training delivery and bomb. Tutorials tell you how to create a chart, but few explain in any depth how to build a good one.

chasma analytics-1

Recommendations

In our experience over the past 30 years, we have developed a set of practices that help us use charts to convey the right information to the right people.

  • Good charts begin with understanding your audience and their needs. Learn what information they require and how they will use it. Seek to give them relevant information in a way that makes sense to them.
  • Understand the purpose of your chart before you begin. Is it to persuade or learn? To support a decision or explore possibilities? To control a process, show value, or show distribution, or track changes over time?
  • Design your chart so the first reaction is the right one, and test it on your users. Make the main point or theme the center of the design. If you need to show other information like the number of respondents in a survey, move it away from the center of the chart so it doesn’t distract from the first reaction. After your users absorb the initial idea, they will be ready for amplifying information.
  • Avoid clutter. Make your design simple, and don’t include design elements that don’t add value. For example, gridlines and axis values can be distracting. In the examples below, the axis and gridlines obscure the main message (See Figure 3), which is rising participation.

6_Ways_You_Can_Improve_Data_Visualization_ID.png

  • Use color to convey your message, but limit them to as few as possible. Saturation level shows a progression of values (See Figure 4). Contrast emphasizes differences.

6_Ways_You_Can_Improve_Data_Visualization_IE.png

  • Recognize that people rarely know what they want until they see it. They also may not know of how information can help them. Don’t limit yourself what they request. Explore how you can use color and shapes to enhance your message.

One More Thing

Don’t assume that a chart is the best way to present information. Many people, especially Finance-trained executives, want to see summaries and crosstabs. They will grasp complex data if you format it well.

In our next article, we will dig into specific chart types and make a few recommendations on which ones to use in specific situations.

References:

1. Berinato, Scott. Good Charts: The HBR Guide to Making Smarter, More Persuasive Data Visualizations. Boston, MA: Harvard Business Review, 2016. Print.

Pixentia is a full-service technology company dedicated to helping clients solve business problems, improve the capability of their people, and achieve better results.

Previously:  Next up: 

Share

News Letter Sign up

Get in touch with us
phone_footer.png  +1 903-306-2430,
              +1 855-978-6816
 
contact-us.jpg