Course Overview

This course is a thorough introduction to generalized linear models, using the open-source computational software R and BUGS. Course objectives include the following:

• Students will be able to read and understand published articles using GLM.
• Students will become aware of the varied (appropriate) applications of GLM in the behavioral sciences.
• Students will be able to analyze their own data using GLM and the R software package.
• Students will understand the conceptual foundations for appropriately analyzing data using GLM.

The instructor is available by appointment, and can be contacted by email (d.cross@tcu.edu). The textbooks for the course are:

• Agresti, A. (2002). Categorical Data Analysis (2nd ed.). Wiley-Interscience.
• Dalgaard, P. (2002). Introductory Statistics with R. Springer.
• Gelman, A., & Hill, J. (2007). Data Analysis Using Regression and Multilevel/Hierarchical Models. Cambridge University Press.
• Singer, J. D., & Willett, J. B. (2003). Applied Longitudinal Data Analysis: Modeling Change and Event Occurence. Oxford University Press.

The course is typically offered during the spring term, and students are expected to turn in one assignment every two weeks (eight assignments in all). In addition, students are expected to give two presentations to the class: One presentation on a published study using generalized linear models, and one presentation on a data analysis project conducted by the student. Each assignment is worth 10 points, and students must earn 85 points for an A, and 70 points for a B.

Here are some helpful links:

• The R Project
• Resources to help you learn and use R (UCLA)
• R for SAS and SPSS Users
• R: A Primer (pdf)
• Log-Linear, Logit, and Probit Models (G. David Garson)
• Generalized Linear Models (Germán Rodríquez)
• R/S-PLUS example: loglinear models for 2-way and 3-way tables (Kevin Quinn)
• extended Rasch modeling (eRm)
• Multilevel Modeling in R (2.2) (pdf)
• The BUGS Project
• OpenBUGS: the Homepage
• JAGS for Mac Os X
• The 10/20/30 Rule of PowerPoint

Students should have taken at least two courses in quantitative methods prior to enrolling in this class. The class meets once each week, and will cover the following topics:

1. Categorical Data and Contingency Tables
• Agresti: Chs. 1–3
• Dalgaard: Chs. 1–3 & 8
• Gelman & Hill: Cs. 1 & 2
2. GLM and Logistic Regression
• Agresti: Chs. 4–7
• Dalgaard: Chs. 13–15
• Gelman & Hill: Chs. 5 & 6
3. Loglinear Models for Contingency Tables
• Agresti: Chs. 8–10
4. Multilevel Models for Categorical Responses
• Agresti: Chs. 11–15
• Gelman & Hill: Chs. 14-18, 24
• Singer & Willett: Part II

Class Photos

• Spring 2010

Home | Research | Lab | Publications | Classes | Links