Storytelling with Data

Home > Other > Storytelling with Data > Page 10
Storytelling with Data Page 10

by Cole Nussbaumer Knaflic


  The resulting visuals, though at the core were exactly the same, had a completely different look and feel because of these simple changes. As with many of the other decisions we make when communicating with data, the audience (in this case, my client) should be kept top of mind and their needs and desires considered when making design choices like these.

  Cultural color connotations

  When picking colors for communications to international audiences, it may be important to consider the connotations colors have in other cultures. David McCandless created a visualization showing colors and what they mean in different cultures, which can be found in his book The Visual Miscellaneum: A Colorful Guide to the World’s Most Consequential Trivia (2012) or on his website at informationisbeautiful.net/visualizations/colours-in-cultures.

  As another example on color and tone, I recall flipping through an airline magazine on a business trip and finding a fluffy article on online dating accompanied by graphs charting related data. The graphs were almost entirely hot pink and teal. Would you choose this color scheme for your quarterly business report? Certainly not. But given the nature and lively tone of the article these visuals accompanied, the peppy colors worked (and caught my attention!).

  Brand colors: to leverage or not to leverage?

  Some companies go through major undertakings to create their branding and associated color palette. There may be brand colors that you are required to work with or that make sense to leverage. The key to success when that is the case is to identify one or maybe two brand-appropriate colors to use as your “audience-look-here” cues and keep the rest of your color palette relatively muted with shades of grey or black.

  In some cases, it may make sense to deviate from brand colors entirely. For example, I was once working with a client whose brand color was a light shade of green. I originally wanted to leverage this green as the standout color, but it simply wasn’t attention grabbing enough. There wasn’t sufficient contrast, so the visuals I created had a washed-out feel. When this is the case, you can use bold black to draw attention when everything else is in shades of grey, or choose an entirely different color—just make sure it doesn’t clash with the brand colors if they need to be shown together (for example, if the brand logo will be on each page of the slide deck you are building). In this particular case, the client favored the version where I used an entirely different color. A sample of each of the approaches is shown in Figure 4.16.

  Figure 4.16 Color options with brand color

  In short: be thoughtful when it comes to your use of color!

  Position on page

  Without other visual cues, most members of your audience will start at the top left of your visual or slide and scan with their eyes in zigzag motions across the screen or page. They see the top of the page first, which makes this precious real estate. Think about putting the most important thing here (see Figure 4.17).

  Figure 4.17 The zigzag “z” of taking in information on a screen or page

  If something is important, try not to make your audience wade through other stuff to get to it. Eliminate this work by putting the important thing at the top. On a slide, these may be words (the main takeaway or call to action). In a data visualization, think about which data you want your audience to see first and whether rearranging the visual accordingly makes sense (it won’t always, but this is one tool you have at your disposal for signaling importance to your audience).

  Aim to work within the way your audience takes in information, not against it. Here is an example of asking the audience to work against the way that comes naturally to them: I was once shown a process flow diagram that started at the bottom right and you were meant to read it upwards and to the left. This felt really uncomfortable (feelings of discomfort are something we should aim to avoid in our audience!). All I wanted to do was read it from the top left to the bottom right, irrespective of the other visual cues that were present to try to encourage me to do the opposite. Another example I sometimes see in data visualization is something plotted on a scale ranging from negative to positive where the positive values are on the left (which is more typically associated with negative) and the negative values are on the right (which is more naturally associated with positive). Again, in this example, the information is organized in a way that is counter to the way the audience wants to take in the information, rendering the visual challenging to decipher. We’ll look at a specific example related to this in case study 3 in Chapter 9.

  Be mindful of how you position elements on a page and aim to do so in a way that will feel natural for your audience to consume.

  In closing

  Preattentive attributes are powerful tools when used sparingly and strategically in visual communication. Without other cues, our audience is left to process all of the information we put in front of them. Ease this by leveraging preattentive attributes like size, color, and position on page to signal what’s important. Use these strategic attributes to draw attention to where you want your audience to look and create visual hierarchy that helps guide your audience through the visual in the way you want. Evaluate the effectiveness of preattentive attributes in your visual by applying the “where are your eyes drawn?” test.

  With that, consider your fourth lesson learned. You now know how to focus your audience’s attention where you want them to pay it.

  chapter 5

  think like a designer

  Form follows function. This adage of product design has clear application to communicating with data. When it comes to the form and function of our data visualizations, we first want to think about what it is we want our audience to be able to do with the data (function) and then create a visualization (form) that will allow for this with ease. In this chapter, we will discuss how traditional design concepts can be applied to communicating with data. We will explore affordances, accessibility, and aesthetics, drawing on a number of concepts introduced previously, but looking at them through a slightly different lens. We will also discuss strategies for gaining audience acceptance of your visual designs.

  Designers know the fundamentals of good design but also how to trust their eye. You may think to yourself, But I’m not a designer! Stop thinking this way. You can recognize smart design. By becoming familiar with some common aspects and examples of great design, we will instill confidence in your visual instincts and learn some concrete tips to follow and adjustments to make when things don’t feel quite right.

  Affordances

  In the field of design, experts speak of objects having “affordances.” These are aspects inherent to the design that make it obvious how the product is to be used. For example, a knob affords turning, a button affords pushing, and a cord affords pulling. These characteristics suggest how the object is to be interacted with or operated. When sufficient affordances are present, good design fades into the background and you don’t even notice it.

  For an example of affordances in action, let’s look to the OXO brand. On their website, they articulate their distinguishing feature as “Universal Design”—a philosophy of making products that are easy to use for the widest possible spectrum of users. Of particular relevance to our conversation here are their kitchen gadgets (which were once marketed as “tools you hold on to”). The gadgets are designed in such a way that there is really only one way to pick them up—the correct way. In this way, OXO kitchen gadgets afford correct use, without most users recognizing that this is due to thoughtful design (Figure 5.1).

  Figure 5.1 OXO kitchen gadgets

  Let’s consider how we can translate the concept of affordances to communicating with data. We can leverage visual affordances to indicate to our audience how to use and interact with our visualizations. We’ll discuss three specific lessons to this end: (1) highlight the important stuff, (2) eliminate distractions, and (3) create a clear hierarchy of information.

  Highlight the important stuff

  We’ve previously demonstrated the use of preattentive attributes to draw o
ur audience’s attention to where we want them to focus: in other words, to highlight the important stuff. Let’s continue to explore this strategy. Critical here is to only highlight a fraction of the overall visual, since highlighting effects are diluted as the percentage that are highlighted increases. In Universal Principles of Design (Lidwell, Holden, and Butler, 2003), it is recommended that at most 10% of the visual design be highlighted. They offer the following guidelines:

  Bold, italics, and underlining: Use for titles, labels, captions, and short word sequences to differentiate elements. Bolding is generally preferred over italics and underlining because it adds minimal noise to the design while clearly highlighting chosen elements. Italics add minimal noise, but also don’t stand out as much and are less legible. Underlining adds noise and compromises legibility, so should be used sparingly (if at all).

  CASE and typeface: Uppercase text in short word sequences is easily scanned, which can work well when applied to titles, labels, and keywords. Avoid using different fonts as a highlighting technique, as it’s difficult to attain a noticeable difference without disrupting aesthetics.

  Color is an effective highlighting technique when used sparingly and generally in concert with other highlighting techniques (for example, bold).

  Inversing elements is effective at attracting attention, but can add considerable noise to a design so should be used sparingly.

  Size is another way to attract attention and signal importance.

  I’ve omitted “blinking or flashing” from the list above, which Lidwell et al. include with instructions to use only to indicate highly critical information that requires immediate response. I do not recommend using blinking or flashing when communicating with data for explanatory purposes (it tends to be more annoying than helpful).

  Note that preattentive attributes can be layered, so if you have something really important, you can signal this and draw attention by making it large, colored, and bold.

  Let’s look at a specific example using highlighting effectively in data visualization. A graph similar to Figure 5.2 was included in a February 2014 Pew Research Center article titled “New Census Data Show More Americans Are Tying the Knot, but Mostly It’s the College-Educated.”

  Figure 5.2 Pew Research Center original graph

  Based on the article that accompanied it, Figure 5.2 is meant to demonstrate that the 2011 to 2012 increase observed in total new marriages was driven primarily by an increase in those having a bachelor’s degree or more (there doesn’t actually appear to be an increase based on the “All” trend shown, but let’s ignore this). The design of Figure 5.2 does little to draw this clearly to our attention, however. Rather, my attention is drawn to the 2012 bars within the various groups because they are rendered in a darker color than the rest.

  Changing the use of color in this visual can completely redirect our focus. See Figure 5.3.

  Figure 5.3 Highlight the important stuff

  In Figure 5.3, the color orange has been used to highlight the data points for those having a bachelor’s degree or more. By making everything else grey, the highlighting provides a clear signal of where we should focus our attention. We’ll come back to this example momentarily.

  Eliminate distractions

  While we highlight the important pieces, we also want to eliminate distractions. In his book Airman’s Odyssey, Antoine de Saint-Exupery famously said, “You know you’ve achieved perfection, not when you have nothing more to add, but when you have nothing to take away” (Saint-Exupery, 1943). When it comes to the perfection of design with data visualization, the decision of what to cut or de-emphasize can be even more important than what to include or highlight.

  To identify distractions, think about both clutter and context. We’ve discussed clutter previously: these are elements that take up space but don’t add information to our visuals. Context is what needs to be present for your audience in order for what you want to communicate to make sense. When it comes to context, use the right amount—not too much, not too little. Consider broadly what information is critical and what is not. Identify unnecessary, extraneous, or irrelevant items or information. Determine whether there are things that might be distracting from your main message or point. All of these are candidates for elimination.

  Here are some specific considerations to help you identify potential distractions:

  Not all data are equally important. Use your space and audience’s attention wisely by getting rid of noncritical data or components.

  When detail isn’t needed, summarize. You should be familiar with the detail, but that doesn’t mean your audience needs to be. Consider whether summarizing is appropriate.

  Ask yourself: would eliminating this change anything? No? Take it out! Resist the temptation to keep things because they are cute or because you worked hard to create them; if they don’t support the message, they don’t serve the purpose of communication.

  Push necessary, but non-message-impacting items to the background. Use your knowledge of preattentive attributes to de-emphasize. Light grey works well for this.

  Each step in reduction and de-emphasis causes what remains to stand out more. In cases where you are unsure whether you’ll need the detail that you’re considering cutting, think about whether there is a way to include it without diluting your main message. For example, in a slide presentation, you can push content to the appendix so it’s there if you need it but won’t distract from your main point.

  Let’s look back at the Pew Research example discussed previously. In Figure 5.3, we used color sparingly to highlight the important part of our visual. We can further improve this graph by eliminating distractions, as illustrated in Figure 5.4.

  Figure 5.4 Eliminate distractions

  In Figure 5.4, a number of changes were made to eliminate distractions. The biggest shift was from a bar graph to a line graph. As we’ve discussed, line graphs typically make it easier to see trends over time. This shift also has the effect of visually reducing discrete elements, because the data that was previously five bars has been reduced to a single line with the end points highlighted. When we consider the full data being plotted, we’ve gone from 25 bars to 4 lines. The organization of the data as a line graph allows the use of a single x-axis that can be leveraged across all of the categories. This simplifies the processing of the information (rather than seeing the years in a legend at the left and then having to translate across the various groups of bars).

  The “All” category included in the original graph was removed altogether. This was the aggregate of all of the other categories, so showing it separately was redundant without adding value. This won’t always be the case, but here it didn’t add anything interesting to the story.

  The decimal points in the data labels were eliminated by rounding to the nearest whole digit. The data being plotted is “Number of newly married adults per 1,000,” and I find it strange to discuss the number of adults using decimal places (fractions of a person!). Additionally, the sheer size of the numbers and visible differences between them mean that we don’t need the level of precision or granularity that decimal points provide. It is important to take context into account when making decisions like this.

  The italics in the subtitle were changed to regular font. There was no reason to draw attention to these words. In the original, I found that the spatial separation between the title and subtitle also caused undue attention to be placed on the subtitle, so I removed the spacing in the makeover.

  Finally, the highlighting of the “Bachelor’s degree or more” category introduced in Figure 5.3 was preserved and extended to include the category name in addition to the data labels. As we’ve seen previously, this is a way to tie components together visually for our audience, easing the interpretation.

  Figure 5.5 shows the before-and-after.

  Figure 5.5 Before-and-after

  By highlighting the important stuff and eliminating distractions, we’ve markedly improved this visual.
/>   Create a clear visual hierarchy of information

  As we discussed in Chapter 4, the same preattentive attributes we use to highlight the important stuff can be leveraged to create a hierarchy of information. We can visually pull some items to the forefront and push other elements to the background, indicating to our audience the general order in which they should process the information we are communicating.

  The power of super-categories

  In tables and graphs, it can sometimes be useful to leverage super-categories to organize the data and help provide a construct for your audience to use to interpret it. For example, if you’re looking at a table or graph that shows a value for 20 different demographic breakdowns, it can be useful to organize and clearly label the demographic breakdowns into groups or super-categories like age, race, income level, and education. These super-categories provide a hierarchical organization that simplifies the process of taking in the information.

 

‹ Prev