2019 Hogg and Craig Lecturer is David A. Harville

Professor Emeritus, Department of Statistics, Iowa State University
Date: 
Thursday, April 25, 2019 to Friday, April 26, 2019

Dr. David A. Harville from Iowa State University will be our 47th Hogg and Craig Lecturer.  

Early in the 1969-70 academic year, Professor Allen T. Craig announced his retirement. He gave a retirement talk in January 1970. Under the leadership of Craig’s student and co-author, Professor Robert V. Hogg, the department decided to establish a lecture series to honor Professor Craig. His January 1970 talk was the first in this series. When Professor Hogg passed away at the age of 90 in 2014, the department decided to incorporate his name into the lecture series.

https://stat.uiowa.edu/hogg-and-craig-lectures

David A. Harville served for 10 years as a mathematical statistician in the Applied Mathematics Research Laboratory of the Aerospace Research Laboratories (at Wright-Patterson AFB, Ohio), for 20 years as a full professor in Iowa State University’s Department of Statistics (where he now has emeritus status), and for 7 years as a research staff member of the Mathematical Sciences Department of IBM’s Thomas J. Watson Research Center. He is the author of more than 80 research articles and of 3 books: Matrix Algebra From a Statistician’s Perspective, Matrix Algebra: Exercises and Solutions, and Linear Models and the Relevant Distributions and Matrix Algebra. He has served as an associate editor of Biometrics and of the Journal of the American Statistical Association. He is a Fellow of both the American Statistical Association and the Institute of Mathematical Statistics and is an elected member of the International Statistical Institute.

Schedule:

Thursday, April 25

2:30 p.m. Refreshments and Student Awards in 302 Schaeffer Hall (SH)

3:30 p.m. Lecture #1 in LR2 Van Allen Hall (VAN)

Ranking/Rating Basketball or Football Teams: the NCAA Way and the “Right” Way

Systems for ranking or rating high school or college basketball or football teams take many forms, ranging from polls to so-called computer rankings. Such systems are sometimes used in a way that affects the teams being ranked or rated, as in the case of the NCAA’s use of the RPI (Rating Percentage Index) or more recently the NET (NCAA Evaluation Tool) in the selection and seeding of college basketball teams for “March Madness.” Then, ideally, the ranking/rating system should have certain attributes, including accuracy, appropriateness, impartiality, unobtrusiveness, nondisruptiveness, verifiability, and comprehensibility. The RPI and the NET lack some of these attributes. A system possessing all of the attributes, except for unobtrusiveness, can be achieved by applying least squares to a statistical model in which the expected difference in score in each game is modeled as a difference in team effects plus or minus a home court/field advantage. The potential obtrusiveness of this approach can be largely eliminated by introducing modifications to reward “winning per se” and to limit or do away with any incentive for “running up the score” or for extending a game into overtime.

Friday, April 26

3:00 p.m. Reception in 241 Schaeffer Hall (SH)

3:30 p.m. Lecture #2 in W151 Pappajohn Business Building (PBB)

Model-Based Prediction in General and in the Special Case of Ordinal Data

Prediction problems are ubiquitous. In a model-based approach to predictive inference, the values of random variables that are presently observable are used to make inferences about the values of random variables that will become observable in the future, and the joint distribution of the random variables or various of its characteristics are assumed to be known up to the value of a vector of unknown parameters. Such an approach has proved to be highly effective in a wide variety of important applications. In cases where the model is taken to be a linear model and the form of the joint distribution to be multivariate normal, the implementation of a model-based approach is relatively tractable. And the results obtained for such cases can be extended to cases where the variables are ordinal in nature by relating the joint distribution of those variables to that of “latent variables.” The performance of a prediction procedure in “repeated application” may be important and can be evaluated from a theoretical (model-based) perspective and/or from empirical evidence. For purposes of illustration, the weekly prediction of the outcomes of college football games will be considered.