how do I know which graph to use?
Illustrations by Catherine Madden
In today’s post, we’ll discuss one strategy for a frequently posed question in data visualization: how do I know which graph to use?
Rarely is there a single “right” way to graph a given set of data. Rather, any dataset can be graphed countless different ways. Furthermore, varying views allow us to more (or less) easily see different things. For a refresher on the use cases for common chart types, check out our “what is…?” series.
In practice, choosing an effective visual means allowing enough time to iterate and select the graph which will create that magical “a-ha” moment of understanding for your audience. Also, recognize that the view you used to arrive at your “a-ha” moment might not be the same view that works best for your audience, depending on their graphical literacy and baseline familiarity with your data.
Let’s look at an example of this iterative process. Imagine you work in marketing for an organization that rents space for events. The following table shows the five types of events and the average yearly revenue per event. Spend a moment taking it in. What can you easily observe about the trends?
When we first aggregate our data, we often summarize it in a table. Tabular data allows for scanning down rows and columns, reading the numbers, and making mental comparisons. While tables may seem like a straightforward way to show our data, they actually take quite a bit of brainpower for our audience to consume. Let’s assume I want to emphasize a pattern or shape in the data—I’m better off using a chart that will enable my audience to more efficiently recognize the patterns.
I set a timer for 5 minutes and allowed myself to brainstorm every possible way I could visualize this data. Then, I narrowed it down to four options shown below: bars, lines by type, lines by year, and a combination of the table + lines by year.
As you scan through the four views above, take note of what each allows you to more (or less) easily see.
Which one would I ultimately choose? It depends. Any of these could work, depending on what I want my audience to take away from the data. You, too, can practice with this example in the community exercise reduce complexity!—and see how others made their decisions.
In summary, rarely is there a single “right” way to graph a given dataset. The next time you’re asking yourself “which chart should I use?”, set a timer for 5-10 minutes and see how many ways you can come up with to visualize your data. Then, evaluate the pros and cons of each. If you’re still not sure, grab a friend or colleague and ask their opinion.
Here are some additional resources to guide you through this process:
Practice iterating with the free community exercise which would you choose?
Listen to Cole’s thoughts in the podcast episode right place right graph
Build your foundational knowledge with our choose an effective visual: chart guide
Get input: subscribe to premium to get personalized feedback during our weekly office hours
Register for a public webinar or workshop and learn how to weave your graphs into compelling data stories.
