This talk is about designing visual interfaces to facilitate analytic reasoning. There are many forms of analytic reasoning to support and many perceptual/cognitive challenges to address. Here the emphasis here on visual interfaces for statistical reasoning. These often follow a “partition, compare and decide” paradigm that contrasts sharply popular “search to find it” interfaces. Statistical paradigms may never be popular, but they can be appropriate in contexts involving uncertainty and the visualization setting can make the methods more accessible.

The first example concerns action prioritized visual analytics now being distributed over the web by the National Cancer Institute for use by health planners. Usability assessment indicates that health planners can put the graphics to good use even though the majority has not taken a statistics class. Subsequent examples in heath, environment and educations areas address discovery and hypothesis generation for maps and comparisons involving uncertainty.

The whole field of visual analytic design is poised to become a vibrant research area. Visual analytics are being increasingly used in business and science, and the Department of Homeland Security through its National Visualization and Analytics Center has established a five-year research and development agenda