showing the insights vs. showing off
For people who love to analyze and visualize data, temptations abound. A specific one that’s on my mind today is common to just about anyone who has developed an interest in data visualization—myself included!
As we develop our skills and enthusiasm for data visualization, we naturally seek out the work of other practitioners in the field. We become more aware of graphs and charts around us in the world. This leads to an increased exposure to unusual, fancy, eye-catching charts and graphs. It’s natural to be drawn to them: they’re designed to grab a viewer’s attention and stand out from the crowd.
Since we’re trying to improve, and we’re learning and getting better as we go, we might aspire to create the same kind of flashy visuals. At the very least, we’re inspired to experiment with them, and share them with our friends and colleagues.
The temptation becomes dangerous when we lose sight of what is important for our specific communications, especially at work. Audience is paramount in business presentations, and the audience doesn’t know—or need to know—that you’re bored with making lines and bars. They don’t want to have to learn how to read a Sankey diagram, a connected scatterplot, a wheat plot, or (as shown above) a sunburst plot.
These visuals might have caught your eye, since they’re colorful and distinctive. The numbers in the middle imply that there’s data behind these rings and wedges. They’re pretty, they might be fun to explore…but they’re hard to compare.
Both visuals use data from 2018. The left-hand graph shows the U.S. population 17 years old or younger as collected by the Census Bureau, while the right-hand graph shows the number of public school teachers in the U.S. as reported by the National Center for Education Statistics.
In each, the outer ring shows the total by state (including Washington D.C.); the second ring aggregates the breakdown to the Census Bureau-designated Division level, and the inner ring is the most aggregated, showing the data at the Region level.
How much insight can you derive from these graphs? Even if we were to label a few wedges with their values, the comparisons would be challenging to make. As they stand, these graphs might better be used as decorations, or as something included in a social media campaign. But for a business audience, we want to make more elegant, simpler visuals.
Yes, that means going back to our old friend, the bar chart. We can pre-calculate a more meaningful metric, and show it in a single visual: the ratio of minors-to-teacher by state. From there, we have a few options for how specifically to show some insights.
We could keep the color-by-Region theme in place:
This view emphasizes that the lowest student-teacher ratios (shown on the left) are in the blue Northeast region, and the highest (on the right) are in the purple West.
If those geographic distinctions weren't critical to show, we could use our color less categorically:
The sparing use of color allows us to focus audience attention on the state with the absolute lowest ratio (Vermont, which is better than 15-to-1) as well as the highest (Arizona, California, and Utah, with ratios worse than 30-to-1).
Using our power pairing of color and words, we might develop a final view something like this:
In a business setting, delivering key messages and making insights clear take priority. Most often, simple and focused graphs will best accomplish this goal. That’s not to say that you shouldn’t experiment with unusual graph types, to broaden your capabilities and gain more aptitude with your tools. However, unsuspecting colleagues may not be as appreciative of your efforts as your fellow dataviz enthusiasts in the SWD community. We humbly suggest that our monthly challenges and exercises are great venues for pushing those boundaries.
If you’re interested in hearing about some other common temptations we all face in our experiences communicating with data, join us on Monday, June 19 at 12:30pm ET for “resist the temptation!” This live event, open to premium members of the SWD community, discusses 12 graphing choices you’ll desperately want to make—but shouldn’t. Hope to see you there!
