Monday, March 16, 2009

A roadmap to generalized linear mixed models

In a recent paper in TREE, Ben Bolker (from the University of Florida) and colleagues describe the use of generalized linear mixed models for ecology and evolution. GLMMs are used more and more in evolution and ecology given how powerful they are, basically because they allow the use of random and fix effects and can analyze non-normal data better than other models. The authors made a really good job at explaining what to use when. Despite the fact that you need more than basic knowledge of stats to fully understand this guide, I think that people should take a look at it before starting to plan their projects, since it outlines really well all the possible alternatives (and challenges) that one can have when analyzing data. This article also describes what is available in each software package; this is really useful since is not obvious with program in SAS or R you need to use when dealing with some specific GLMMs.

Bolker, B., Brooks, M., Clark, C., Geange, S., Poulsen, J., Stevens, M., & White, J. (2009). Generalized linear mixed models: a practical guide for ecology and evolution Trends in Ecology & Evolution, 24 (3), 127-135 DOI: 10.1016/j.tree.2008.10.008

6 comments:

Anonymous said...

I'm so glad you pointed this article out - I've been looking for something like this for awhile!

Martin Nuñez said...

Great that it helps! I think that it is a useful article in a very important topic.

Marc Cadotte said...

Another really useful paper is Johnson's and Omland (2004), Model Selection in Ecology and Evolution TREE 19: 101.

http://dx.doi.org/10.1016/j.tree.2003.10.013

trav said...

The latest SAS guide for PROC MIXED (SAS for Mixed Models, Littell et al 2006) includes a lot of great information for generalized linear mixed model as used in PROC GLIMMIX.

[Sorry, I know I should be moving to R and am doing it slowly.]
travis

Anonymous said...

Thanks Travis! I didn't know about that. I'm also moving slowly towards R. I know that R is the future, but for now I’m better at SAS...

Sadie said...

Thanks muchly for this bloglet!