Data Visualization: It's on...

 For all of you interested in how we display data and other quantitative phenomena (which I hope is nearly everyone reading this), the ongoing throwdown over at FlowingData is worth taking a look at.  In that bloggy way, it's a response to a response to an (older) blog entry about the best data visualization projects of 2008.  But the exchange raises some interesting questions about how statisticians and others think about data visualization, and about the balance (tradeoff?) between graphics for presentation and for analytics.

Comments

[nepali music video]

thanks for sharing this blog.

nepali music video

[Tracy Lightcap]

What I don't understand is the winning entry at Flowing Data. This is pretty much exactly analogous to a display of weather patterns on a high capacity UNIX workstation. There are only two things missing and they would have made the entire enterprise much more useful. First, there is no time referent in the GPS tracings. Of course, one could infer that as traffic picks up as the day goes on, but it would be much more informative if the time were directly attached to the traces, just like in weather satellite images. This would be especially useful for the holding patterns around airports.
Second, there are no grid referents to convey distance traveled or identify air or seaports. The result is that we don't get a clear view of the comparative length of the flights and voyages, a pretty important part of the data. This would all be easy to do with color grids and different colors for different distances.
True, you would have to slow down the presentation and get a LOT deeper into what's going on to do either of these, but if you want people more data literate you have to get them away from thinking that pretty pictures are the end of the story and make them engage the DATA.

[Daniel Katz]

Good for Nathan Yau. In my estimation, his arguments are spot on.

The tone of some of the commenters was pretty much out of line. At a deeper level, I believe those individuals exposed their ignorance about the current state of the world. Hopefully, this push back might be the catalyst for some reflective evaluation on the part of those commenters.

From a personal perspective, we have attempted (perhaps imperfectly) over at Computational Legal Studies Blog to take stock of the amazing developments in Data Visualization. While it is always a work in progress, I believe we should always strive to tinker, to test and to improve everyday.

Best, Dan