Instructor: Jack Weiss, Curriculum in Ecology, 218 Miller Hall, 962-5930, jack_weiss@unc.edu
Meetings: 5 lectures hours per week—MWF 10-11 am, 224 Phillips Hall and Tues 9-11 am, 214 Sonja Haynes Stone Center
Website: http://www.unc.edu/courses/2006spring/ecol/145/001/
Office Hours: I'm available at any time on M, T, W, and F. Miller Hall is on the corner of McCauley and Pittsboro, next to the Carolina Inn
Evaluation: Weekly homework, midterm, and final
Registration Details: Seats are reserved separately for ECOL 145 (15 seats), BIOL 145 (5 seats), and ENST 145 (5 seats). If the course is closed for one of these listings, try registering for one of the other two.
ECOL 145 is intended to be an intense introduction to the analysis of ecological data. Its target audience consists of highly motivated graduate students and upper level undergraduates in biologically-related disciplines who ideally have data of their own to analyze. This is a serious, hands-on course not suitable for dilettantes or those who wish to merely audit and observe. We focus on the use of two modern statistical packages, R and WinBUGS, and use them to tackle real data sets with all their foibles. The closer you are to carrying out your own research and analyzing your own data the more useful this course should turn out to be.
The perspective of the course is that probability models are best thought of as data-generating mechanisms and in keeping with this viewpoint we use likelihood-based methods to directly model ecological data. Data sets are from the published literature, from my own consulting projects, or are supplied by students who are enrolled in the course. If you have data you need to get analyzed you are welcome to submit it to me for use in class exercises.
The prerequisites are modest, a one semester undergraduate course in statistics and some exposure to the concepts of calculus. If you have never studied statistics, this is probably not the course for you. Having said this I think it is quite possible to obtain the requisite background through self-study before the course begins. I expect you to be familiar with the following ideas:
Just about any elementary statistics text covers this material. The book Introduction to the Practice of Statistics by David S. Moore and George P. McCabe, the text that is used in undergraduate statistics courses at UNC, is an adequate choice as is any other text written at this level. Personal favorites of mine include Statistics with Applications to the Biological and Health Sciences by M. Anthony Schork and Richard D. Remington (any edition) and Biometry by Robert R. Sokal and F. James Rohlf.
No familiarity with statistical software is assumed. We will be using R, a free implementation of the S language, available for free download at http://cran.r-project.org. R runs on all major operating systems, including Windows, Unix, and Mac OS X. R is a state-of-the-art modern statistical package actively supported by the scientific community. It is not easy to use but it has become the de facto standard for scientific research. Although I've used many statistical packages in my career and even spent a number of years working for the SAS Institute, I now do all my work using R. Our lecture sessions in Saunders Hall will include hands-on instruction in the use of R. In addition there are many online resources and published textbooks in the use of R. Links to these will appear on the course web site.
The textbook for this course is
Crawley, Michael J. 2002. Statistical Computing: An Introduction to Data Analysis Using S-Plus. Wiley, New York.
I have made purchase of this text OPTIONAL, because I think it is prohibitively expensive (lists at $140). Because I listed it as optional, the UNC bookstore does not stock it and if you want it you will have to special order it. Having said this there are some good deals available on the web. Amazon.com discounts the book by almost $50 and lists other online sellers offering new copies of the book at almost half price. www.addall.com and www.bookfinder.com are examples of search engines that can be used to turn up reasonably priced new and used copies of the book. The lowest online prices I've seen (on Dec. 10, 2005) for new copies are around $70.00 domestically (including shipping) via sellers at A1Books and Bookbyte.com, and $54.00 (with shipping) from a British bookseller.
There are no exercises in the book so I will be treating it purely as a supplement to what is done in lecture. It also covers only about 70% of what we'll be doing. Crawley is a plant ecologist at Imperial College in London and the examples in the text reflect this. I've suggested this book to a number of students as a way to learn S-Plus and/or R and they invariably have given it rave reviews. Crawley has also written a later book that focuses on R, Statistics: An Introduction Using R, 2005, Wiley, but it is less than half the length of the earlier book and leaves out much of the more advanced material. In Statistical Computing Crawley focuses exclusively on the command line use of S-Plus. It turns out that the differences in syntax between S-Plus and R for what he is presenting are minimal. Both of Crawley's books have associated websites with supplemental material including additional chapters (Michael Crawley's website).
If you're looking for a more thorough reference to R and modern statistical modeling, it would be hard to beat the following two books. We'll be covering much of what's in the second book listed below.
Faraway, Julian J. 2004. Linear Models with R. CRC Press, Boca Raton, FL.
Faraway, Julian J. 2005. Extending the Linear Model with R: Generalized Linear, Mixed Effects, and Nonparametric Regression Models. CRC Press, Boca Raton, FL.
To see what sort of things have been covered in this course in the past, you are welcome to visit the old website. Be forewarned that we will be covering much more than what you'll find there (and omitting some as well). When I last taught ECOL 145 I anticipated that it would become the first installment of a two semester sequence. Since this turned out not to be the case, I've now combined the two courses into one and have selected the most important topics from each for the current incarnation.
As noted in the overview above, this is a course in statistical modeling as it is currently practiced in ecology, or at least the way I think it should be practiced! I plan to the cover the topics listed below.
While the course will begin with statistical distributions and likelihood theory and end with Bayesian analysis, the order of the remaining topics is flexible.
| Jack Weiss Phone: (919) 962-5930 E-Mail: jack_weiss@unc.edu Address: Curriculum in Ecology, Box 3275, University of North Carolina, Chapel Hill, 27516 Copyright © 2006 Last Revised--Jan 20, 2006 URL: http://www.unc.edu/courses/2006spring/ecol/145/001/descrip.html |