**Official Description**: This is a variable unit class, so after you register, you will need to change the number of units from one to four. Meets twice a week for 80 minutes including about 1/3 lecture & discussion, and 2/3 group work time. There will also be substantial exercises to work on each week outside class.

In this course, we will learn about and use some important tools for analyzing environmental data sets: classical/frequentist mixed models, hierarchical modeling and Bayesian analysis. Throughout the course, we will focus on linear and generalized linear models with random effects. We will spend most of the time talking about how the models work and when to use them, and very little on theory, though there are lots of good, more theoretical resources. All exercises will be done on your own computers using R, OpenBUGS and JAGS (which is almost the same in practice as OpenBUGS). During class time, students will work in groups. For the homework exercises, everyone will do the analyses themselves, though working together is fine.

In putting together the examples and exercises for this class, I will be assuming you already are familiar with R, or are willing to put in a lot of time to get familiar with it at the beginning. I’m not assuming any familiarity with OpenBUGS/JAGS and will start with very simple Bayesian models.

Students should have experience with applied statistics, including linear models (ANOVA, ANCOVA, regression), such as you can get by taking PLS 205, PLS 206 or equivalents. If you have not taken any graduate-level statistics courses please contact the instructor before enrolling.

Things this course should help you do:

- Become more comfortable with data analysis: exploratory data analysis, estimating power and uncertainty, understanding the inferences that models let you make as well as their limitations.
- Learn about mixed effects models, simple hierarchical models and Bayesian analysis, and things to consider when deciding whether to use them.
- Get some practice in using and understanding these methods by applying them to example data sets in class and in homework assignments.
- Learn where to find more information about the methods.

Things the course won’t do:

- Give you a rigorous theoretical understanding of the methods. The stats department has good offerings for that, and there are also many useful books.
- Survey methods in general for environmental statistics or “big data.”

**Instructors**: Andrew Latimer

**Other info and comments**: Instructor plans to offer this course in Fall 2015. Syllabus

**Software Used**: R, OpenBUGS and JAGS (which is almost the same in practice as OpenBUGS)

Useful class for using R to 1) Implement multilevel models, and 2) Run Bayesian models using JAGS. Content will probably be refined going forward.