Theodore Chang Professor Ph.D., University of California, Berkeley Contact Information 119 Kerchof Hall phone: (434) 924-3096 fax: (434) 982-3076 e-mail: tcc8v at virginia dot edu Fellow, Institute of Mathematical Statistics Fellow, American Statistical Association Research Interests Although I pursue several seemingly diverse research paths, the common theme of my work is the use of geometric information in statistics. One problem in which I am vitally interested is the statistics of estimating an unknown rotation of 3-space. Although this problem has many applications, its most scientifically compelling application is in the reconstruction of the past position of tectonic plates. The theory of tectonic plates, previously known as the theory of continental drift, holds that the Earth's surface is composed of tectonic plates which move as rigid bodies. It follows that the motion of each plate can be described by a path of rotations. Since very little theory exists about the statistical properties of estimated rotations, this problem has subproblems which range from the very theoretical to the very applied. Any theory that is developed can be immediately checked with existing geophysical data for which no analysis is currently available. Thus one can realistically hope that one's work will very quickly be used in scientifically important contexts. Another problem in which geometric information is used is in the development of reference priors in Bayesian statistics. Bayesian statistics attempts to combine information from data with prior information, with the latter conceivably being quite vague. Reference priors are used when no prior information is available. Historically, the selection of reference priors has had a strong geometric flavor. Thus, a student who enjoys thinking geometrically and who enjoys developing statistical techniques which solve important scientific problems will enjoy the type of problems which I have.