Data Visualization Exercises

Exploring exercises and challenges posted to Storytelling With Data

Storytelling With Data is an excellent resource for practicing your data visualization skills, explaining your thought process, and getting feedback. I regularly contribute to their community and participate in their design challenges. Here are some of the latest.

Challenge: Effectively using an area graph

Area graphs can be confusing, and it’s often better to stick with other types of information display (line charts, bars, etc.). In some cases though, the data is well suited for an area graph. Can you research and display data that is best shown as an area graph? Here is my submission. I chose to research and graph U.S. plastic recycling rates:

I’ve graphed the EPA data for recycled amounts of numbered plastics in the U.S. from 2013 - 2017. What is shown here represents about 6-8% of our total plastic usage annually. So, during the years shown, the U.S. used around 30 million tons of plastic, and recycled about 2.4 million tons of it. What actually gets recycled is a murky picture. Since the 1990s (when many people in the U.S. were told to stop sorting their recycling), most plastics have been shipped overseas for processing, mixed, in bulk, and primarily to China. Recycling processors in the U.S. are for the most part only able to process #1s and #2s, and only a few U.S. processors can handle additional resin #s.

What I had really wanted to find was data post-2017. Starting in January of 2018, China enacted National Sword. This policy bans foreign recyclables, and essentially reduces China’s appetite for U.S. plastic recyclables down to <1% of exports. Other countries have stepped in to purchase, but the market is volatile, and the amount being sent overseas has fallen. Hopefully the EPA will publish post-2017 data to shed more light on how this is affecting U.S. plastic recycling rates.

Exercise: Translate unfamiliar data into something accessible

Analyze the following graph and design a new one, making a recommendation on how the information should be displayed, and what might be the most important takeaway.

The following is my update to the sample Model Performance graph:

In this version I’m highlighting what is happening at the left of the graph-the diverging Model/Actual lines.

Recommendations:

Replace acronyms, abbreviations, and anything else that isn’t obvious, to avoid confusing the audience.

Match the title with the main point of the graph, so the viewer understands upfront what the graph is trying to communicate.

Simplify the elements to avoid any confusion-get rid of additional line colors or styles that might confuse the viewer.

Clearly identify the axes. This will allow the viewer to scan and make comparisons more quickly. The second Y axis adds a level of complexity, so it can help to color code it with the bright orange to visually separate it from the first Y axis.

Exercise: simplify table data

Visualize table data in a way that highlights key takeaways from the data.

My update to the event revenue table:

Recommendations:

Removing event color coding.

Adding a color to highlight top trends.

Using the title to explain the purpose of the table.