A few key points I took away from Journalism in the Age of Data:
- Half our brain is hardwired for visual interpretation.
- Visualization needs to come more from an art/writing perspective as opposed to programming focus.
- User frustration is a primary factor in bad visualization. Simply showing data and letting people draw their own conclusions it not necessarily serving the purpose. Needs perspective.
- Computers can only do so much, a human brain is still necessary to frame and analyze data in the proper context and narrative.
- “Transparency is the new black” – in the growing trend of governments and organizations to share raw data it is becoming increasingly more important to find a way to shift through this information and present it concisely and compellingly.
The final chapters of Journalism in the Data Age, mention several data visualization sites (Swivel, Protovis, Google Charts, WebGenie, Many Eyes) that would allow for the easy creation of visual graphics. I wanted to explore these sites more, but was disappointed to find that many we’re no longer in existence, just a year later. Swivel had shut down completely. Protovis is “no longer under active development,” and I couldn’t even find WebGenie data visualization in a quick Google search.
Google Charts and Many Eyes are still up and running, but I think it is interesting that the other tools have so quickly disappeared. I think this reflects one of the points that was made in the video: data still needs a human brain to make the visuals relevant. It is not enough to throw up information in a graphic, it must have context and narrative. These sites create a visual but it still needs an astute individual to best present the information dynamically.
Click on the graph to go to the “interactive” version.
I made this pie chart above using Google Charts. In a matter of minutes I was able to take a dry list of Census information about the demographic breakdown of North Carolina, and present it in a visual matter. However, this visual is one dimensional. It doesn’t give context (compared to other states, by NC county/region). It doesn’t tell a story (how have the demographics changed over time). There are limits to what a template can do, and this is why it is not enough to simply plug numbers into some lines of code. We have to be able to step back, assess, edit and compile the data. We have to think of what the overarching story is. What does it mean at the Macro and Micro levels? Visualization and interaction are not enough. Data needs perspective, interpretation and digestion – and that can be accomplished by presenting the date inside a narrative or by letting the user uncover the information for themselves. However, the visuals still need a structure and human thought to give it the most impact. In short, successful data visualizations need us.