Specialized statistical methods for categorical data are widely used today for applications in different areas of science such as biomedical and social sciences. This course will cover most important methods for analyzing categorical data. It will include standard descriptive and inferential methods for contingency tables and matched pair data, modeling categorical data using generalized liner models, logistic regression models for binomial and multinomial data. The course will involve the practical application of the ideas and their implementation through statistical software (SAS and/or R).
This course is designed to provide modern statistical methods for analyzing categorical data. The course objectives are:
- Understand inference for proportions and odds ratios
- Understand probability structure and inference for contingency tables
- Perform logistic regression analyses with multiple predictors
- Understand inference for matched pair designs
- Understand when to utilize hypothesis testing versus modeling
Evaluation of Students
Student will be evaluated based on Homework, projects, midterm, and final examinations.
Probability Distributions for Categorical Data
Statistical Inference for Proportion and Discrete Data
Inferential methods for contingency tables
Generalized Linear Models for Binary and Count Data
Logistic Regression Models for Binomial Data
Logistic Regression Models for Multinomial Responses
Log-Linear Models for Poisson Data
Methods for Matched-Pair Data
Textbooks & Software
Agresti, Alan (2007). An Introduction to Categorical Data Analysis. John Wiley & Sons, Inc.: New York.
Statistical software: SAS and/or R.
Submitted by: Mohammad Kafai
Date: October 31, 2016