A mini-course on using R for statistical analysis

I'm not at Stony Brook anymore, but I'm going to leave these documents up for as long as possible for those who want to use them. If you are considering offering a similar course somewhere else, you might like to know that handouts below are offered under a Creative Commons license that gives you the right to redistribute and modify these works. If you're doing this, I'd love to hear about it, because I would like to know what you found useful and not.

Course material

Day two: Basics of linear models: Regression, ANOVA, and ANCOVA.
Handout, caterpillars.txt, crushing.txt, predation.txt, tetrahymena.txt

Day three: An introduction to mixed-models in R.
Handout, water.txt, lacY.txt.

Day four: R is a programing language: simulations and automation.
Handout, kinetics.txt.

Resources:

• Above all, each other! One point of the course was to start a group of users in the department. We have a departmental listserv to help get people started and deal with issues that arise. Email me from your life.bio account to join.
• The R homepage- the place to get software, as well as a number of good tutorials.
• There are a number of free tutorials on the R website. I can recommend these:
  • An introduction to R- a dense but good resource- all the basics are here.
  • Using R for Data Analysis and Graphics - Introduction, Examples and Commentary by John Maindonald[PDF]
  • Practical Regression and Anova using R by Julian Faraway [PDF]
• There are a number of books on R/S, of which I have three. All have proven useful.
  • Introductory Statistics with R, by Peter Dalgaard. This book tries to teach R and intro stats at the same time. Because you will already be familiar with much of the statistical methodology, you can concentrate on just learning R.
  • Modern Applied Statistics with S by W. N. Venables and B. D. Ripley. This book has everything, usually at a higher level than I am initially comfortable with. When I use it, I often have to learn some theory as well as methodology, which is probably good on the whole...
  • Mixed-Effects Models in S and S-Plus by Jose C. Pinheiro and Douglas M. Bates. Much more specific than the previous two, but I found this book particularly helpful for using the nlme package for mixed-effect models. nlme uses maximum likelihood for its estimation (like proc mixed in SAS), so you need to learn some theory to complement what we learned in biometry.