Robert E. (Rob) Kass
Everything I've really needed to know (about Statistics), I learned in graduate school. A few basic ideas that appealed to me were: first, scientific inference, the process by which conclusions are drawn from data, may be formalized by applying Bayes' Theorem; second, asymptotic approximations (that is, approximations that assume the sample size is nearly infinite) are often helpful; and third, geometry can help us better understand our mathematical manipulations. In addition, several consulting projects solidified my view that these notions have important consequences for statistical practice.
I am frequently reminded, especially when I play basketball with our graduate students, that it has been quite a while since I was in school. During the intervening years I've listened to, read, and thought about many problems in statistical theory and applications, but my main beliefs and interests have not drifted far from the basics I embraced back then. The papers I've written, at least the ones I take some pride in, have grown from the simple excitement of understanding something better, recognizing the implications and wanting to share the experience. Getting the picture right is rarely easy, but to glimpse the depth of fundamental concepts and witness the beauty that flows from them, it's worth the effort.
Although I spent a lot of time studying mathematics and statistical theory, in college I worked in medical laboratories and, ever since, hoped to return to experimental life sciences. Now, in mid-life, I have found cognitive neuroscience, a marvelous area rich in data analytic challenges, which is a subject of intensive research at Carnegie Mellon and the University of Pittsburgh, especially through our Center for the Neural Basis of Cognition. My scientific collaborations continue to shape my judgment about what is important in our own discipline.
Some Related Publications
Kass, R.E. and Raftery, A.E. (1995) "Bayes factors," Journal of the American Statistical Association, 90, pp. 773-795.
Kass, R.E. and Wasserman, L. (1996) "The selection of prior distributions by formal rules," Journal of the American Statistical Association, 91, pp. 1343-1370.
Kass, R.E. and Vos, P. (1997). Geometrical Foundations of Asymptotic Inference New York: Wiley.
Kass, R.E. and Ventura, V. (2001). "A spike-train probability model," Neural Computation, 13, pp. 1-8.