Breaking Defaults

Tuesday, February 19, 2013

By Traci Montgomery

Visualizing data is relatively easy to do these days considering the wide variety of tools at our disposal., a great data viz blog, names Excel, Photoshop, Illustrator, Tableau, Google Public Data, Many Data, and Stat Silk as just a few tools to visualize data.

While some programs require a steeper learning curve to efficiently use, any user with a data set and some basic knowledge of Excel can produce a wide range of visualization types including bar graphs, pie charts, area charts, and scatterplots. Guest author to The Why Axis, Jon Schwabish, takes a look at a Bureau of Labor Statistics (BLS) visualization, done in Excel, for job openings in November 2012. While the visualization passes for use of appropriate chart type, it fails in its details. Because the BLS utilized default settings from Excel, the true story of the data is lost.

Schwabish takes us through minor changes, all done in Excel, to create a visualization that more effectively tells the story of job openings in November 2012. Schwabish explains the things he finds appealing about the visualization: sourcing, a left-aligned title, and values measured in thousands to name a few; however, the default coloring, automatic spacing, and ordering of the bars and industries are a few things to be improved upon.


His first of a series of changes is a quick sort on industry by descending values which helps to give more order to the graph. A change in colors helps to make the most recent data value stand out against the previous months while creating a more cohesive visualization.

First Change

Schwabish goes on to show more suggested changes to better the storytelling of Job Openings in November 2012. The take away here is not exclusive to Excel, but all data visualization programs. Simple design elements including descriptive text, color, font, and order are important to telling a story and, often times, the default settings for telling that story are not optimal.

Read the full blog post here and make sure to examine the transition from one version of the visualization to the next.

Page 1 of 1 pages