Categorical Data Analysis
(CHL 5210, Fall 2011)
http://fisher.utstat.toronto.edu/sun/Teaching/chl5210_index.html
General Information
- Time: Mondays, 10am - 1pm (We start at 10am sharp. Please
arrive on time).
- Location: HSB 790.
HSB: Health Sciences Building, 155 College Street (South on College
and West of University).
Instructor
- Lei Sun (sun@utstat.toronto.edu)
- Laurent Briollais (laurent@mshri.on.ca)
Office Hour
- Mondays noon/12:30-1pm (the last hour or half hour of the class).
- There is no TA for the course.
Prerequisites and Enrollment
- This is a graduate course with the following prerequisites.
- Statistics at the graduate level or consent of instructor.
- Working knowledge of SAS/R or other equivalent software
packages is necessary.
- All participating students must register! This is a
graduate school policy.
- September 26: The final date to enroll the course.
- October 31: The final date to withdraw from the course.
Course Information
- Teaching objectives
This course covers the fundamental statistical methods for
analyzing categorical data, including logistic, poisson, and
log-linear regression models; methods for goodness-of-fit,
2-by-2 tables, and stratified 2-by- 2 tables;
maximum likelihood theory for generalized linear models;
unconditional and conditional likelihood logistic regression;
poisson regression; analysis of multi- dimensional contingency tables
and log-linear models; comparison and contrast of different methods;
model specification - choosing and assessing models; GLM; GEE.
- Format of instruction
Lecutures.
- Evaluation
Student evaluation will be based on homework problem sets (30%),
midterm (30%), term project (30%) and overall participation (10%).
- Text book
Alan Agresti (2002). Categorical Data Analysis.
Second edition. Wiley.
- September 12: session 1 (Lei Sun).
Introduction - key
distributions for categorical variables; a brief summary
of classical statistical inference, hw1
- September 19: session 2 (Lei Sun).
Contigency Tables - hypthesis
testing, CI, comparing
two proportions
- September 26: session 3 (Lei Sun).
Contigency Tables - OR, Small
sample inference, SAS
Tutorial - review hw1 (please hand in hw1 on time) and distributions of various tests
- October 3: session 4 (Lei Sun).
Logistic Regression - single
predictor
Tutorial - type 1 error, power and sample size
- October 10: no class; university closed for Thanksgiving Day.
- October 17: session 5 (Lei Sun).
Logistic Regression - multiple predictors
Tutorial - review hw2 (please hand in hw2 on time)
- October 24: session 6 (Lei Sun).
Logistic Regression and Poisson Regression
Tutorial - review hw3 (please hand in hw3 on time)
- October 31: session 7 (Lei Sun).
OR inference from stratified
samples; Models for matched pairs.
Tutorial - review hw4 (please hand in hw4 on time)
- November 7: midterm exam (10am-noon only; a single page two-sided
notes is allowed; a calculator can be helpful).
- November 14: session 8 (Laurent Briollais).
Logit models for nominal and ordinal responses and introduction to methods
for repeated categorical data.
- November 21: session 9 (Laurent Briollais).
GLM, Quasi-likelihood and GEE for repeated categorical data.
- November 28: session 10 (Laurent Briollais).
Random effects models for repeated categorical data.
- December 5: session 11 (Laurent Briollais).
Transitional models for repeated categorical data and case study.
- December 16: Term project due.