As I do with any project, I began my work on the GPI visualization with the idea in mind to go broad and deep with the data, including cross-source analysis. And this time I also wanted to challenge myself (just for dashboard practice) to:
- get the entire message on a single (not long form) screen
- include all of the most important metrics regardless of whether there was a story
- use annotation, navigation and/or show/hide for story points
- design it to be fully-refreshable (again just for practice)
My first analysis of the data was via a long-form heat map with every country on rows and every indicator on columns and score on color sorted by overall score. The dark red in a sea of blue would eventually lead to the “Otherwise Peaceful Countries” table at the very bottom of the final visualization.
The plan was to next connect all the data sources to each other to identify patterns or relationships, but the reality set in that this was a (my first) quick turnaround project and I simply wasn’t going to be able to be as aggressive as I usually am.
This led to a bit of a crisis because I no longer had a plan and had never had to restrict myself from my usual process of digging deep and daydreaming about projects for days on end. At least once, I considered abandoning the project.
But then it hit me. This project is not for consumption by the IEP. This project is to explain the importance of and dynamics around the GPI to a broad audience. And there is a lot to explain if viewers are going to be able to identify with the problem.
Deciding to focus on the full decade and realizing that I needed to more fully educate myself before I could adequately explain things to others, I immediately scanned the entire 2017 GPI report and also found the launch video…
…which I watched in full and highly recommend. Now I was ready!
So there is the main GPI index and that is divided into domains and each of those have indicators. Perfect! So, as I typically do, I resolved to lay it out from broad to specific ; from the overall GPI to individual indicators. But how?
Since I wanted to show both overall dynamics and individual countries, a scatter plot was the first thing that came to mind. But to be sure, I thoroughly rummaged through my memory of good distribution visualizations and the various chart lists and favorite visualizations that I keep. I also googled alternatives for scatter plots and histograms, but in the end I decided nothing that I saw would work better than a scatter plot. And whenever a standard chart type is available that does the job, I choose it every time over more “innovative” chart types because people are more likely to be familiar with them. A test of the scatter plots with the real data confirmed that they would work beautifully.
The line chart showing change over the decade for the three domains is a very standard chart type for this purpose and worked perfectly.
The next challenge was how to describe movement in all the indicators for all the countries and, almost exactly as I was going to begin considering what chart type to use for this, a swarm of butterfly charts from #MakeoverMonday submissions relating to White House salaries began flooding my twitter feed. Perfect! If it ain’t broke, don’t fix it. All I had to do was add color coding for whether more countries became more peaceful or more countries became less peaceful. Done!
It was now time to execute and I decided to practice some of the concepts from Cole Knaflic’s book, Storytelling with Data, which suited the need to educate the audience perfectly with explanations throughout but “pushed to the background” so as to be there when needed but not overly complicate or overwhelm the visualization.
And using a table for “Otherwise Peaceful Countries” was an obvious choice and also functions somewhat as an upside down histogram in that you can easily see which problems the most countries suffer from.
That’s about it, or at least what I can remember as it was quite a whirlwind. Thanks for reading and a special thanks to the IEP, Vision of Humanity, and Olga of #DataForACause for the important work that they do.