Mixed Models with R
Introduction
Overview
Goals
Prerequisites
Workshop
Key packages
Mixed Models
Terminology
Kinds of Clustering
Random Intercepts Model
Example: Student GPA
The Standard Regression Model
The Mixed Model
Initial depiction
As a multi-level model
Application
Initial visualization
Standard regression
Regression by cluster
Running a mixed model
Cluster Level Covariates
Summary of Mixed Model Basics
Exercises for Starting Out
Sleep
Adding the cluster-level covariate
Simulating a mixed model
More Random Effects
Application
Comparison to many regressions
Visualization of effects
Summary of Random Slopes
Exercises for Random Slopes
Common Extensions
Additional Grouping Structure
Cross-classified models
Hierarchical structure
Crossed vs. nested
Residual Structure
Heterogeneous variance
Autocorrelation
Generalized Linear Mixed Models
Exercises for Extensions
Sociometric data
Patents
Issues
Variance Accounted For
Common Alternatives to Mixed Models
Growth curve models
Sample Sizes
Small number of clusters
Small number of observations within clusters
Balanced/Missing values
Big data
Model Comparison
Convergence
Bayesian Approaches
Priors
Fixed effects
Variance components
Demonstration
Example Models
Beyond the Model
Going Further
Other Distributions
Other Contexts
Nonlinear Mixed Effects Models
Connections
Summary
Supplemental
A Comparison to Latent Growth Curve Models
Random effects as latent variables
Random effects in SEM
Running a growth curve model
Random intercepts
Random intercepts and slopes
Random effects with heterogeneous variances
Other covariates
Some differences between mixed models and growth curves
Recommended packages that can do growth curve models
Summary of LGC
Correlation Structure Revisited
Summary of residual correlation structure
Appendix
Data
Programming languages
R
Python
Julia
Proprietary
Reference texts and other stuff
Mixed Models with R
Mixed Models with R
Getting started with random effects
Michael Clark
m-clark.github.io