It’s said that a picture paints a thousand words. And that’s probably one reason why charts, graphs, and other types of data visualization are commonly used as a means of helping people better understand or visualize numeric or other types of information.

Charts and graphs come in various flavors—bar charts, pie charts, line graphs, etc. And each type of graph or chart is more or less ideally suited to displaying specific types of data or information, depending on the nature of the data to be conveyed. The right choice, for any given situation, has to do with the type of graph or chart selected, as well as some of the detailed attributes of the chart or graph itself. In this article, we’ll look at how the design of the chart or graph can drive perception of the information, and consequently, how people respond.

The Design Tells a Story

For example, let’s say we want to show data about two different types of toothpaste. A comparison of the two toothpastes shows that, out of a population of 5,000 people, twice as many suffer from serious gum disease while using one toothpaste over the other. Specifically, 15 people using Toothpaste A suffer from gum disease versus 30 people using Toothpaste B.

This second version of the display is quite different from the first. Does it give you a different impression of the information?

It’s worth noting that there are three important pieces of (numeric) data to consider in this scenario: the number of people who have gum disease for each type of toothpaste (30 for Toothpaste A and 15 for Toothpaste B), as well as the total number of people who were considered (5,000). A key difference between these two versions of the display, is how each of them handles these three data elements.

In the second display, the difference between 15 and 30 people doesn’t seem like much. That’s because the design makes it very easy to see a comparison of the numbers 15 and 30, respectively, against the total population of 5,000. The total population of 5,000 is made very salient in the second design.

In the first example, however, the design drives focus towards a direct comparison between the numbers 15 and 30. The total population of 5,000 isn’t even graphically represented in the design. Rather, it’s a line of text beneath the graphic. Even though the very same data is portrayed across both types of displays, what people perceive from each of these has everything to do with how the design influences what they are comparing.

Each design tells a very different story. The design itself becomes the storyteller then by the way in which it highlights certain information and drives different types of comparisons. We can see that the design itself has the power to influence the meaning of the data, and consequently, how people respond.

Real World Implications

The implications are substantial. Graphs and charts are used ubiquitously across a vast array of industries, including healthcare, financial services and many other domains. Oftentimes, consumers of the information are using it to make critical decisions that affect quality of life for themselves and those that depend on them. For example, companies may use graphical illustrations to help people understand risk, and to try and motivate them to take appropriate action in light of that information.

In our toothpaste example, people’s perception of the incidence of gum disease would likely be very different depending on whether they had encountered the bar chart versus the pie chart version of the information. In the bar chart version, it’s easy to conclude that the level of risk is twice as much with Toothpaste A than Toothpaste B (or half as much with Toothpaste B than Toothpaste A). And “twice as much” (or “half as much”) can seem like a lot. But when put into a different perspective—via the pie chart—the actual numbers seem much less consequential.

Indeed, a research study found that a key factor in affecting risk-relevant behavior is how the design drives attention to the information. Meaning, whether attention is called to the total number of people being considered (the total population) or only the number of people harmed. Researchers concluded that when incident rates are small, displays that highlight only the number of people harmed produce the greatest effects on risk-avoidant behavior.

Spatial/Axis Design

We’ve seen how the chart or graph design can highlight certain data elements by what it makes most salient. Let’s consider another example where salience plays a role in decision outcomes.

In this scenario, a college student is considering two different scholarship opportunities. Each of the two scholarships has tradeoffs. One offers more money but requires more time in order receive the money, while the other offers less money, but requires less wait time.

We can see that in both versions, the actual data displayed is the same. But the spatial representation is different. In the first version, the amount of wait time is made salient by elongating the X axis (by showing more space between the increments of time), so that it becomes more prominent than the Y axis. In the second version, the Y axis, showing the amount of money, becomes most salient by incorporating more space between the increments.

To what extent might these kinds of design variations have an effect on students’ preferences? In a study that examined this, researchers found that the design did indeed have an influence and caused students to form very different impressions of the same data, depending on which design they had encountered. In general, students who viewed Version 1 of the graphic display preferred Scholarship A, while those who viewed Version 2 preferred Scholarship B.

A key finding from this study, is that the design itself plays an important role in how people perceive and make meaning of the data. That is, the design itself is a critical driver of preference outcomes.

The Design of Graphics

The design of graphics is important because of the way in which it drives attention, facilitates certain types of comparisons and motivates behavior. When designing graphics, it’s important to first be clear about the story you want to tell:

Getting the design just right is difficult. And sometimes, the only way to “get it right” is to design a couple of different versions and consider which one best accomplishes your objectives in a truthful, yet compelling, way. It’s even worth usability testing these variations to glean insight into how your target audience is perceiving and interpreting the information. How does the story they tell from looking at the design compare with the story you want to tell?

Conclusion

Findings from the two research studies referenced in this article show that the design of graphs and charts can have a significant impact on viewer perception of the information displayed, as well as on how people are able to glean meaning from the information. Indeed, the “story’” told by the graphic can be very different—depending on the design.

Perhaps too often in everyday life, not enough thought goes into how design details like these can influence and drive viewer perception, and consequently, people’s behavior. The design itself is the storyteller, and every business has a story to tell. The question is: Does your design tell the story you intend?

Comments
  • Damon Scrase

    I really enjoyed this article. I can’t say I’ve ever thought much about exactly how you specifically frame information. I can see this helping me in the future. Thank you!

  • Excellent article, most people think of graphs and charts just as “pretty pictures” not aware of the importance of how the dat is presented.

  • Sreenatha Reddy K R

    Nicely explained about Data Visualization. Thanks for sharing.

  • Nice piece. It parallels nicely with The Visual Organization: Data Visualization, Big Data, and the Quest for Better Decisions. Data visualization + good design = a better chance of solid business decisions.

  • Well-articulated examples. I would argue this piece is more about truth and accuracy in dataviz than storytelling.

  • good article keep update with this topic please and post something about in deep.

  • OCP

    Nice article Colleen. So true that data visualization is essential nowadays. Especially true of those trying to interpret analytics data (most commonly GA), exporting spreadsheets to find answers; when there are far quicker work-arounds by using behavioral analytics tools

  • Pingback: The Effect of Design on Decision-Making()

  • Pingback: Power Your Brand With the Persuasive Triangle & the Inverted Pyramid()