Last week, Tyler Dahlberg and I attended the Do Good Data Conference in Chicago, IL. The conference was a two-day extravaganza of workshops and talks geared towards nonprofit data geeks and professionals looking to learn new skills.

One of the sessions I found particularly insightful was hosted by Matthew Scharpnick of Elefint Designs. Matthew offered tips for effective data visualizations and creating infographics. He noted that raw data can deceive and how important context is. One of the case studies he brought up was the on-going debt ceiling debates during in 2012. At the time, there was a hot political debate about which party (and President) was responsible for running up budget deficits and debt. A graph chart of U.S. debt over time is one way to look at this — but there’s more nuance than simply political party in charge or President at the time. There’s even differences depending on whether the data is normalized by GDP, for example.

For example, if you only saw the chart on top left, you might think that U.S. debt is out of control and has never been higher. But if you look at debt as a percentage of GDP, you can see that in fact debt was higher in the 1940s during World Ward II.

So Elefint created an infographic to try to tell multiple stories and show context. Designers chose to symbolize the line by which party controlled the Presidency, House and Senate. Looking at President, it’s natural to infer that in recent times, Republican administrations appear to be more responsible for running up U.S. debt. But when Congress is added, that narrative becomes much more nuanced. Finally, by adding the recession periods to the graph (in gray), it becomes apparent that perhaps debt always increases after a recession, regardless of which party controls the Presidency or Congress.

It’s important to remember that data doesn’t teach us everything and it’s easy to miss the whole picture without the proper context. Matthew also stated the importance of direct labeling and telling a story — but noted that there doesn’t always have to be a linear story to each visualization — each person will see what they want to see. Moreover, he brought up the idea of “Return on Design” and mentioned crafting pieces that can be “productized” and reused.

Tyler and I also presented a session at the conference. We outlined the bike theft and pedestrian crash analysis we worked on at Azavea — offering numerous spatial analysis tips along the way. The workshop portion was devoted to a hands-on demo of QGIS mapping, including the QGIS2Leaf plugin. We also demonstrated web mapping in CartoDB, including setting up infowindows with Google Streetview images and the Torque library for animated heatmaps. We’ve posted the presentation and workshop instructions online and we hope to see everyone next year!