Chart Wars — The power of data visualization
May 19, 2011
The amount of information available to us over the internet grows rapidly. Recent advances in computer technology as well as software design make it easier than ever before for anybody with a computer to visualize all kinds of data. But, the same data can be visualized in very different ways — to serve very different purposes. Alex Lundry, Vice President and Director of Research at TargetPoint, talks about the political power of data visualization and offers a few short lessons about visual literacy.
The more data we have to deal with the more crucial it becomes to interpret them. Since vision is by far our most dominant sense the best way to make data understandable and memorable is to visualize them. Data visualization is a relatively young discipline somewhere between statistics and graphic design. Info graphics can help to display and explain even complex relationships.
It’s really gotten to the point where anybody with a computer can create a data visualization easily enough using either tools on their computer or online. — Alex Lundry
So, just feed your computer with a set of data, the tools will do the rest — et voilà, there is your beautiful info graphic… or so it seems. But, data visualization is more complicated than that. It starts with interpreting the data, with thinking and understanding. In order to create a proper data visualization we need to be aware of the fact that we are using a kind of language who’s grammar needs to be learned first. We have to become visually literate to create as well as read info graphics. Our tools are only as good as our ability to use them.
Alex Lundry demonstrates this using a messy, complicated, and confusing info graphic of the House Democrats’ Healthcare Reform Plan, released by Republican Congressman John Boehner (left) and the elegantly redesigned version based on Boehner’s chart (right) created by graphic designer Robert Palmer. What is more honest, complicating or simplifying the content? Was it inaptitude on the Republican side? Or calculated distortion? However, it makes the plan look far more chaotic than Robert Palmer’s revised version does.
This is political debate fought with visual means — and there are more of these “Chart Wars” to come. So, we better learn to understand that visual language. Lundry shows four common ways of messing with data in info graphics: 1. playing with charts’ origins and axes; 2. omitting crucial data, 3. implying relationships that aren’t there; 4. manipulating the scales. At the end of his presentation, Lundry names Edward Tufte as the “godfather of the dataviz movement”.
Afterthought: With his manifesto, PowerPoint Is Evil, Tufte triggered a heated discussion about the cognitive style of PowerPoint that “routinely disrupts, dominates, and trivializes content”. That seems all right — but for the wrong reasons. In his books, Tufte advocates “Data-Ink Maximization”, i.e. a high information density; his graphics are complex, elegant, and demanding. Given the time necessary, they can be properly digested and understood. Tufte’s principles are aimed at print media. They cannot be applied to slide presentations.
For slide presentations we have to break down the data into pieces of low information density, so that displaying them step by step matches the pace of the presenter as well as the capacity of the listeners to absorb the information. This approach is audience-centered and dynamic rather than data-centered and static. To present data live in a meaningful way we need to: 1. focus on core data; 2. display them in simple, distinguishable shapes (pie graphs are often not appropriate), 3. reveal them in logical steps.