Multilevel Models II: Advanced Topics

Instructor(s):

  • John Poe, University of Kentucky

This course is designed to extend the basic multilevel skills that participants receive from an introductory applied class to more sophisticated and complex models like nonlinear and non-hierarchical mixed effects models.

In the first week we will review the basics of multilevel models and discuss more advanced topics like model tests and diagnostics, dealing with unmodeled or inconveniently structured data, and missing data. In the second week we discuss complex non-hierarchical models and more technical details about generalized linear mixed models. In the third week we will cover models for binary, multinomial choice, ordinal choice, survival, and count outcomes. In the final week we will discuss some advanced topics relating to causal inference, measurement error, and effect heterogeneity. We also spend a significant amount of time on the practical differences between and use of maximum likelihood, simulated/empirical likelihood, and Bayesian approaches to these nonlinear and non-hierarchical multilevel models.

The course will be pitched at around the technical level of other Track III courses, such as Advanced Topics in Maximum Likelihood. Prior exposure to maximum likelihood or basic categorical models as well as some previous background in either multilevel or longitudinal modeling as a prerequisite. Ideally, students will have had courses similar to the Maximum Likelihood Estimation I and Multilevel Models I classes at ICPSR prior to attending. The primary goal of the course will be to allow students to use advanced models and to understand how and when more sophisticated techniques will provide practical benefits.

Fees: Consult the fee structure.

Tags: multilevel models, hierarchical linear models

Course Sections

Section 1

Location: ICPSR -- Ann Arbor, MI

Date(s): July 24 - August 18

Time: 1:00 PM - 3:00 PM

Instructor(s):

  • John Poe, University of Kentucky

Syllabus: