Five Ways to Fail at Data Visualization

One of the best ways to share insights from a dataset is to visualize it. Graphs and charts can communicate complex ideas where words often fail. Data visualization is an important skill, and its applications span many professions.

But unfortunately, many data visualizations just don’t work. Here are five examples of classic data visualization fails.

1. Bad Scales

Tokyo_historical_population.gif

Notice what happens after 2005. The scale switches from 5- to 1-year increments, giving a completely misleading picture of timescale.

It is vital that scales are accurate and consistent because they dictate how we interpret the context and history of the data being presented. Along with inconsistent increments, other scale sins include using illegible text and not showing the y-axis at zero. Bad scales make charts that are easily misinterpreted, even by experts—which is why they are so often the go-to tactic for those who wish to mislead or turn data into propaganda.

 

2. Not Following Convention

This chart is infamous for being extremely misleading. At quick glance it appears that after the law was enacted the number of gun deaths decreased. But the white section isn’t actually the data—the red section is. The chart is upside down!

This visualization goes against basic conventions—nobody expects charts to show data upside-down. Conventions endure for a reason, and visualizations should always anticipate readers’ expectations.

 

 

3. Bad Math

t3_5a0wp5.gif

Charts with bad math are surprisingly common. The easiest ones to spot are pie charts like the one above. This chart has values that add up to 128, which is an odd total for a percentage. Always check for math mistakes and typos before publishing any visualization.

(As a bonus fail, this chart identifies its source as "MOE +/- 4%" ... which is not a source, but rather the margin of error.)

 

4. Unnecessary 3D

3dbar.png

Three-dimensional visualizations can be extremely useful for studying the relationship of variables, but are too often used frivolously. For example, in the 3-D chart above, it is nearly impossible to distinguish the percentage of each bar. Some bars are hidden behind others.  

Whatever data they are showing, visualizations must be clear and easy to interpret. In this chart, adding a third dimension may have enabled more data points to be shown at once, but it comes at the expense of comprehension—never a good trade-off.

 

5. No Labels

This visualization uses the Olympic rings to compare data from 5 participant continents (the Americas are combined). While it is a beautiful visual, it fails to communicate any real information because it lacks labels. A simple key showing which color ring represents which continent and numbers for comparison would have fixed this flaw.


Now, everyone makes visualization mistakes now and then—the LiveStories team included—so if you see a bad visualization, don’t be too hard on its creator. And the good news is that most visualization fails are not too hard to fix. To explore some of our favorite tips and tricks for visualizing data, check out our blog post on the subject, or download our free e-book on telling stories with data:

Sources

This blog post uses visuals found on WTF Visualizations and Reddit.

Cover photo by Ricardo Viana.