Answers to panel questions...
I got involved with FODAVA after attending the MSRI workshop a few years ago. This was the first meeting I had attended where mathematicians talked about visualization. In The Grammar of Graphics, I had done a functional analysis of statistical graphics and charts. I was interested in the mathematics underlying visualizations and whether they could be used to formalize areas of visualization that had been historically matters of personal opinion or high-level taxonomies. In FODAVA, I saw an effort to bring together people interested in such formal models for visualization.
I believe every well-formed statistical or mathematical graphic has an underlying model. Understanding this model is a prerequisite to understanding the visualization. Human viewers bring many assumptions and prior knowledge to the viewing of a graph (training in primary and secondary school, etc.). But teaching a computer to understand a graphic requires making explicit the formal model that generates every aspect of the display. This is a mathematical problem -- not an artistic or design problem.
I was trained as a psychometrician and learned advanced statistics from writing a widely-used statistical package. I have been writing code every day for almost 40 years. Therefore, I view myself as part psychologist, part statistician, and part computer scientist. This interdisciplinary perspective is helpful to me in analyzing visualizations. Understanding visualization requires some knowledge of all three fields. I am especially intrigued by generalists like Pat Hanrahan and Colin Ware.
I publish primarily in The American Statistician, The Journal of Computational and Graphical Statistics, and in TVCG. I am branching out by submitting papers to other computer science journals as well. I use the Web for publishing other material, particularly my site at http://www.cs.uic.edu/~wilkinson/.