Online Program

Thursday, February 20
SC2 Design and Analysis of Experiments Using Generalized Linear Mixed Models Thu, Feb 20, 8:00 AM - 5:00 PM
Bayshore I

Outline & Objectives (302852)

1. From Linear Model to GLMM
General Setting
Linear Models and Linear Mixed Model (LM, LMM)
Generalized Linear Model (GLM)
Generalized Linear Mixed Model (GLMM)

2. Marginal or Conditional Models
Defining a Model from Design Properties
Overdispersion and Other Design-Induced Issues
G- and R-side Random Effects
GEE versus GLMM
Distributional Implications

3. Estimation
Restricted) Maximum Likelihood
Quasi-Likelihood/GEE
Pseudo-Likelihood
Laplace and Quadrature
Model-Based and Empirical (“Sandwich”) Estimators

4. Rates and Proportions
Distributions
Binomial Proportions
Binary Data
Multinomial
Beta – Continuous Proportions

5. Counts
Distributions
Poisson or Negative Binomial?
Modeling with Offsets
Zero-inflated Models

6. Within-Subject Correlation
Repeated Measures Background
Review of Methods for Normally-Distributed Data
Extension to Non-Normal Data – Similarities and Differences
Spatial Variation

7. Power, Precision and Sample Size
Background
Power & Sample Size for Continuous, Count, and Binomial Data
Comparing Competing Designs using GLMM tools
Longitudinal & Spatial Data