The New York Times has launched their visualization lab which uses the ManyEyes software from IBM. The site provides some initial datasets and allows users to create their own visualizations. The currently available datasets include several speeches by current US politicians (Obama, McCain, Biden, and Palin) and include:

*NFL Teams Ranked By Rushing Yards
*Number of infant deaths per 100,000 live births in selected countries
*Party Affiliation By Religious Tradition, Percentages
*Obama Acceptance Speech at the DNC
*Sarah Palin Speech at RNC

and 5 others…

In this initial set, I was intrigued by the user-contributed bar chart for Party affiliation by religious tradition (thumbnail above). The fact that the majority of Jehovah’s Witnesses surveyed (60%) either supported a third party or refused to report a party preference jumps out at you when looking at the graph. The majority of all other religious groups reported either strong or moderate preferences for either Democrats or Republicans.

The New York Times generally produces high quality visualizations, some of which I’ve mentioned in this forum before. Many of these visualizations are high quality precisely because they’re tailored to the data at hand – they are “couture” visualizations. ManyEyes, on the other hand, provides “ready-to-wear” visualizations – visualization options designed to fit many datasets, where some techniques fit certain datasets better than others. One of my recent “couture” favorites is this one, entitled Can a President Tame the Business Cycle? from October 18, 2008:

Generating multiple visualizations for the same data allows easy exploration of the data and can help to provide rapid insights. But seeing a custom visualization of an interesting dataset can completely transform one’s understanding of the data.

Hat tip to Ian Mulvany via Friendfeed…