George presented an analysis of papers in previous InfoVis proceedings: 40% of them had no evaluation of the technology described, 5% percent planned to do some evaluation in the future, less than 5% had some subjective evaluation, and less than 5% included formal evaluation information. (Note numbers from my notes don’t add to 100%). He called for researchers to consider usability and publish evaluations. We need to have solid cost-benefit analysis before people will invest lots of time and money in InfoVis. Intuition may be great for design concepts, but it is not always right. He gave an example from his own published work that included peripheral areas in a 3D display. The idea was that people use peripheral vision to maintain spatial awareness without focusing on the objects. The tests had people navigate down a hallway and avoid bumping into corners, etc. However, the results were suprising – people using the system which mapped peripheral vision onto the screen actually did a bit worse than those without it. (I noted that this may be due to the remapping of peripheral vision to the central focus area. There is a disconnect with normal pre-attentive processing. It might do better if there were actually peripheral displays. George was not receptive to this since Windows does not natively support multiple monitors. He needs a Macintosh!)