36th Annual Craig Lectures April 26 and 27, 2006 Alan Agresti Department of Statistics University of Florida Alan Agresti is Distinguished Professor of Statistics at the University of Florida in Gainesville. He earned a bachelor's degree from the University of Rochester and a doctorate in Statistics from the University of Wisconsin (where he was Stephen Stigler's first PhD student). He has published over 100 articles on statistical methodology, mainly on a variety of topics dealing with categorical data analysis. His current research interests include small-sample confidence intervals for contingency tables and the analysis of clustered categorical measurement data. Agresti is best known for being author of the text "Categorical Data Analysis" (2nd ed. 2002, Wiley). Among other books he has authored are "Statistical Methods for the Social Sciences" (3rd ed. 1997, Prentice Hall), and "Statistics: The Art and Science of Learning from Data (2006, Prentice Hall, with Chris Franklin). Agresti has received many awards for his research and his teaching, including Fellow of the American Statistical Association (ASA) and an honorary doctorate from De Montfort University in the U.K. In 2002 he won the award for Excellence in Continuing Education from ASA. In 2003 he was named Statistician of the Year by the Chicago chapter of ASA. In 2004 he was the first honoree of the Herman Callaert Leadership Award in Biostatistical Education and Dissemination awarded by the University of Limburgs, Belgium. He has presented invited lectures and short courses for universities and companies in about 25 countries.

Lecture #1 Wednesday, April 26 Refreshments at 3:30pm in 241 SH Lecture at 4:00pm in 40 SH "Reducing Conservatism of Exact Small-Sample Inference for Discrete Data" 'Exact,' small-sample methods for categorical data are exact in terms of using probability distributions that do not depend on unknown parameters. However, they are conservative inferentially, having actual error probabilities for inference that are bounded above by the nominal level. We examine the conservatism for confidence intervals and survey ways of reducing it, illustrating for estimating and comparing binomial proportions. Fuzzy inference is an adaptation of randomized inference that achieves the error probability exactly. In practice, many would find this approach unsuitable. However, it motivates inferences based on the mid-P value that are less conservative than standard exact methods yet usually approximate well desired error probabilities. We also summarize simple ways of adjusting standard large-sample confidence intervals to improve dramatically their small-sample performance. Lecture #2 Thursday, April 27 Craig Cake and Punch at 3:00pm in 241 SH Student Awards Presentation at 4:00pm in 140 SH Lecture at 4:10pm in 140 SH "A Twentieth Century Tour of Categorical Data Analysis" Since Karl Pearson's introduction of the chi-squared goodness-of-fit test in 1900, the development of methodology for categorical data analysis (CDA) has seen significant contributions from many historically important figures in statistics who are not necessarily readily identified with CDA. These include R. A. Fisher, Jerzy Neyman, G. Udny Yule, and William Cochran. This seminar presents a non-technical chronological survey of the development of the basic methods of CDA in the twentieth century. It focuses on the contributions of these and other statisticians who introduced methods such as the odds ratio, tests and confidence intervals about proportions, logistic regression, probit models, loglinear models, and Bayesian approaches for CDA.

Copyright c. 2007 The University of Iowa. All Rights Reserved. Photography c. UI Photo Service and University Relations, University of Iowa This page was last modified on Friday, 18-May-2007 15:32:24 CDT. /ts