BIOS 7346: Advanced Regression Analysis II (Generalized Linear and Longitudinal Models)

Instructor

• Jonathan Schildcrout, PhD, Associate Professor of Biostatistics
• jonathan.schildcrout@vanderbilt.edu
• 11117: 2525 West End Ave Building
• 615-343-5432

Teaching Assistant

• Allison Hainline

Schedule

• Lectures: Tuesday and Thursday 9:00-10:30, Small Classroom, 11th floor, 2525 WEA
• Lab: Friday 9:00-10:00, Small Classroom, 11th floor, 2525 WEA
• Office hours:
  • Jonathan: 10:30am-11:30pm Thursdays and by appointment.
  • Allison: 3:30-4:30pm Wednesdays.

Textbooks

• Required
  • Diggle PJ, Heagerty P, Liang K-Y, Zeger SL. Analysis of Longitudinal Data, 2nd Edition. Oxford University Press, 2002.
• Recommended
  • Verbeke G, Molenberghs G. Linear Mixed Models for Longitudinal Data. Springer Series in Statistics, 20.
  • Molenberghs G, Verbeke G. Models for Discrete Longitudinal Data. Springer Series in Statistics, 2010.

Course Objectives

This course will cover advanced topics in generalized linear models and it will extend general and generalized linear models to the analysis of correlated or longitudinal response data. The vast majority of the course will focus on parametric and semi-parametric regression methods. At the end of the course, students should:

• Understand the theoretical framework for longitudinal data analysis
• Understand statistical properties of longitudinal data analysis methods (e.g., GLMs, mixed effects models, and estimating equations estimators)
• Be able to apply appropriate techniques for inferences in GLM and longitudinal data settings, examine the validity of the approach, and be able to interpret the results.
• Be reasonably familiar with current research topics in longitudinal data methodology and applications.

Course Outline

This course with be comprised of six modules covering mostly distinct topics in generalized linear models and longitudinal data analysis

• Introductory topics in longitudinal data analysis (LDA)
• Linear mixed-effects models
• Marginal models and estimating equation
• Generalized linear mixed-effects models
• Advanced LDA topics

Other information

• Homework needs to be turned in beginning of class on the due date. Often, we will discuss the homework during the Friday discussion section, so it might be helpful to photocopy your homework so you can refer to it.
• Students are encouraged to work together on homework problems, but they must turn in their own write-ups.
• You will often need a laptop for the discussion section on Fridays, but you should not bring them on lecture days

Grading

• Homework: 20%
• Take-home Midterm Exam: 35%
• Take-home Final Exam: 40%
• Class participation: 5%

-- JonathanSchildcrout - 02 Jan 2013