I was fortunate enough to have the opportunity to present on behalf of our Visual Budget project to the 2012 International Open Government Data Conference sponsored by Data.gov and The World Bank. In addition to presenting the current status of Visual Budget, I offered up a short summary of our guiding principles in designing interactive data visualizations. It is our manifesto, if you will, about how to create presentations on the web that transform open data into engaging media for a general audience.
The Engaging Data Manifesto:
We begin with an instant master class. Hans Rosling is the world’s biggest data visualization rock-star. General audiences love him. He presents to standing ovations around the world. He found something that all of us are looking for. The magical combination of mathematics and charm, of story and interactivity, of data and engagement.
He is the exemplar for our manifesto. Let’s see him in action.
We've developed the following guiding principles in an effort to capture some of Professor Rosling’s success and bottle it for reuse on the web.
#1 - Math Is Easy; Design Is Hard.
Clear graphics boost understanding. Beautiful graphics go further, engaging the audience and inviting exploration. The visual is the gateway to making statistics relatable and numbers memorable.
My favorite example from history is Florence Nightingale's famous "Coxcomb Chart". During the Crimean war, she meticulously assembled mortality statistics in the military hospitals. She found that more soldiers were dying of treatable illness than combat injuries. So she pressed for rigorous sanitation, and turned the hospitals around. After the war, she drew up this graphic and distributed it to the parliament in London. Soon after, hospital sanitation measures were deployed throughout the British empire. A great graphic can change the world.
And so our first declaration:
- We will create beautiful graphics which instantly and clearly tell the human story of the data.
#2 - Tools Are Awesome; Tools Are Boring.
The potential for interactivity to illuminate is vast. When done right it encourages searching, sorting and investigation; it enables discovery.
Take as an example Professor Hans Rosling’s Gapminder project. Gapminder World not only displays the world health data he uses in his presentations, but it also allows access to over a hundred different dimensions of world health data sets. With his powerful tool, you can compare the health, wealth, population, education, economy and more for every nation in the world. This ease of visualization makes possible interesting and sometimes unexpected analysis, such as Dr. Rosling’s observation that religion is not a significant driver of the number of babies per woman.
So we declare:
- We will build views of data that clearly demonstrate statistically significant trends and inflections.
- Our interactive tools will allow for deep analysis and comparison of related views and data sets.
But choices quickly overwhelm. Tools are useless if the audience isn't interested. A push is sometimes needed, and that's what's next.
#3 - Data tell stories; stories need great storytellers.
Stories are told in sequence. Stories are told by personalities. Stories require drama and structure.
Again, using Gapminder World as an example, I find that using the tool on the website is a lot less interesting to me than Professor Rosling’s presentations. It’s not a problem with the tool, it’s a problem with my attention. I need structure to guide me to my point of interest.
We must present data with context, explaining why the patterns and analysis matter to the audience. We do this by presenting information with a human voice-over, or better yet with video. People will listen when individual personalities give their authoritative opinion anchored directly in the data sets that drive our tools.
- We will find the human element in every data story, and we will present it with the drama and charisma of a storyteller.
- We will structure our presentations with a beginning, a middle, and an end.
- With the audience’s curiosity piqued, we will relinquish control to the user to explore the data using our tools as a sandbox.
#4 - Big numbers break our brains.
The human brain is built for a human world, and it takes shortcuts so we don't get overwhelmed. Unfortunately, it isn't built to understand big numbers. Their scale is too far beyond our experience, and our brains discard them as irrelevant. It is therefore an important duty of data visualizers to provide context for scales that are beyond our ordinary experience. We must present millions, billions, and trillions in terms that feel human such as per-capita figures or percentages of the whole.
We conclude our manifesto by declaring:
- We will always connect large numbers to personal experiences.
- We will present numbers in human terms such as per-capita or percent.
- We will use animation to transition between scales and scopes so we never lose the overall context.
So with manifesto in tow, we built something that we think is really cool. As you can see in this video, Visual Budget implements each of these principles to make the confusing US federal budget clear and engaging.
Armed with the principles of the Engaging Data Manifesto, we hope the community of data visualization designers, technologists, and journalists will work together to liberate the mountains of data from their silos and tell their stories in the human terms that will lead to engagement and understanding.
Our presentation deck from IOGDC 2012 is available on slideshare.
A recording of the presentation at IOGDC will soon be made available at data.gov.