Two Questions About Mixed Models

Hi all,

I was wondering:

  1. What are some of your favorite resources about mixed effects models?
  2. What are some of your favorite resources about using/programming/interpreting mixed effects models in R?

Just interested in your take on this topic!

Thanks,
Sean

2 Likes

Not a very clever answer because I have not read this resource but it got a lot of attention on Twitter: https://www.juliapilowsky.com/2018/10/19/a-practical-guide-to-mixed-models-in-r/

II think Richard McElreath’s “Statistical Rethinking” is a fantastic framework for them conceptually. The chapter that introduces mixed models is actually the free sample on his website: https://xcelab.net/rm/statistical-rethinking/

The brms package is very good for fitting, it has a nice syntax for mixed effects and gives you full Bayesian output. I also tend to use mgcv::gam a lot as it handles simple mixed models very well and incorporates a lot of other useful features when I don’t need full-blown MCMC estimation. Plug: I’m co-author on a recent paper on using mixed models and GAMs: https://peerj.com/preprints/27320/

This is a fairly broad question but Ben Bolker posts a lot of great material on rpubs: https://rpubs.com/bbolker

See also his paper on the topic that provides a great review including a R packages and etc: Bolker et al 2009 “Generalized linear mixed models: a practical guide for ecology and evolution” Trends in Ecology and Evolution (pdf)

Then there is his text Bolker “Ecological Models and Data in R”

See also

  • Kruschke “Doing Bayesian Data Analysis: A tutorial with R, JAGS, and Stan”
  • Gelman and Hill “Data Analysis Using Regression and Multilevel/Hierarchical Models”