University of Minnesota
Student Topic Assignments & Midterm Exam for Stat 8053
Fall Semester, 2013

Midterm Exam

On Wednesday, October 30 we will have a visitor teaching the class. Nathan Hubble is a statistician with Travellers' Insurance in St. Paul. He will be discussing common practical problems he confronts in the insurance industry. He will also provide a set of data typical of the type that he sees in his work. I will divide you into groups of 3, who can collaborate on an analysis. However, each student must prepare his or her own write up of what you did. In addition, you will prepare 2 or 3 summary slides that you would use if you were making an oral presentation of the results that summarizes your findings. I will provide more instructions later.

You will have a week to work on the problem. Dr. Hubble will return to class on November 6, when the exam is due, and we will jointly discuss your solutions to his problem.

Topic Assignment

The presentation should be similar the way I present material in class, and will consist of presenting the important ideas and then illustrating them with a worked example in R. Here are the topics, dates and presenters:

1. Nonlinear Regression. Date: October 21. Presenters: Lin Chen and Yuting Sun. Nonlinear models with normal errors. Estimation. Inference. Parameterization. Bootstrap. Example. Source material: http://tinyurl.com/carbook, then select ``Web appendix to the text", then select ``Nonlinear Regression".
2. Nonlinear Mixed Models Date: October 23. Presenters: Brandon Whited and Joshua Wiltsie. Source material: Pinheiro and Bates (2000). Mixed-Effects Models in S and S-PLUS, Springer, Chapters 6 and 8.
3. Missing data Date: December 4-6 Presenters: Aaron Molstad, Dootika Vats and Li Zhong. This presentation will take 2 class periods. Fitting incomplete data with the EM algorithm. Reference: Dempster, Laird and Rubin (1977), Maximum likelihood from incomplete data via the EM algorithm, Journal of the Royal Statistical Society. Series B (Methodological), 1-38. http://tinyurl.com/9fpnwvz. Multiple imputation, http://www.multiple-imputation.com
4. Causality with propensity scores Date: December 9. Presenters: Boxiang Wang and Yang Yang. Paul R. Rosenbaum and Donald B. Rubin (1983), The Central Role of the Propensity Score in Observational Studies for Causal Effects, Biometrika, 70, 41-55. http://www.jstor.org/stable/2335942