??Bar Chart: 5 Formatting Do's and Don'ts

Boring Bar Charts? Absolutely not!

Bar charts are one of the most common data visualizations to find in dashboards and one of the simplest tools to communicate data. But did you know that there are general rules for appropriately formatting these visualizations. If you get these wrong you could misdirect your end users' attention to focusing on stylization errors vs. having them focus on the important insights you've prepared.

5 Do's and Dont's for Bar Chart Formatting

??1. Scale: Bar charts showing quantitative data must start at 0. The entire length of the bar represents the value?and thus the entire length of the bar must be shown. Never ever start a bar chart at other than zero, its misleading and confusing.

?? Never adjust your bar charts' y-axis to start at anything but 0

2. ?? Bar Width and Spacing: The width of your bars as well as the spacing between them need to be consistent. That is, a 1:1 ratio between the width of the bars and the spacing between them. (Histograms don’t necessarily follow this rule)

?? Don't squish your bars too close together

3. ??Bar Orientation: Horizontal vs. Vertical. Horizontal orientation is best when the labels for the categories are long as it would be too tight to squeeze them on a vertical axis. Vertical orientation is best when you can abbreviate your labels because your audience is aware of the meaning of those abbreviations.?

?? Don't make users tilt their heads sideways to read the labels on your charts

4. ??Sorting/Ranking: Sort your bars in a logical order, if for example you are visualizing data by day of the week make sure your bars are oriented in the proper order not the default alphabetical order. Ranking your data either high to low or low to high is much more readable than unsorted bars.

?? Don't order or sort your bar charts in a counter-intuitive way

5. Comparing Multiple Data Points: Small Multiples are great for displaying multiple data points that have in common their categorization.?

?? Use Small Multiples instead of Stacked or Clustered Bar Charts when comparing 3 or more categories

In this graph its hard to compare the sizes of the bars that do not start at 0. Instead if we redesigned it as a small multiples chart, as in the chart below it becomes much easier to read.

Here’s a checklist you can use when formatting your bar charts:

• Does the y axis start at 0?
• Are the labels for the bars long? If so, use the horizontal bar orientation.
• Are you visualizing a time series and trying to show a change in value over time? Use a vertical bar orientation.
• Are your bars sorted/ranked?
• Are you trying to visualize multiple categories? Use small multiples instead of Stacked or Clustered Column Charts

In the next addition we will provide some additional data visualization do's and don'ts !