Analysis of Generalised Linear Mixed Models
Mixed models have become increasingly popular, as they have many practical applications. However, the traditional linear mixed model with normally distributed errors may not always be appropriate for modelling discrete response variables, such as binary data and counts. Typically these types of responses are analysed using generalised linear models such as logistic regression and Poisson regression.
Commonly-used generalised linear models will be extended to deal with multiple error structures, using a variety of scientific examples, including medical applications, such as investigating the presence or absence of adverse events collected in a multi-centre clinical trial.
The emphasis will be on practical understanding, although an outline of the theory will be presented. Practical examples will be used to illustrate the methods, and participants will have the opportunity to fit and interpret models themselves in hands-on computer practicals. The GLIMMIX and NLMIXED procedures of the statistical package SAS will be used, as appropriate, throughout the course.
Who Should Attend?
Statisticians who are already familiar with linear mixed models. It will be assumed that participants are regular SAS users, and have a working knowledge of generalised linear models.
How You Will Benefit
You will learn how to formulate generalised linear models with fixed and random effects for a range of situations, and how to fit and interpret them in the SAS software.
What Do We Cover?
- Review of generalised linear models
- Mixed models for binary and binomial response data: logistic regression
- Count data: Poisson and negative binomial regression with mixed effects
- Ordered categorical response variables: proportional odds model with mixed effects
- Common fitting methods; inferential procedures
- Use of PROC GLIMMIX and PROC NLMIXED and convergence issues
- Interpretation of effects in a generalised linear mixed model and prediction
- Applications such as repeated measurements and multi-centre trials
- Extensions to non-linear models.
This course has practical exercises written for: SAS
Related Courses: Analysis of Mixed Models