NEWS: There is a companion class to this one!  You will need to use real statistical software to complete the assignments in FW5411 - Applied Regression Analysis.  I suggest you use R and if you do, you can sign up for FW5412 get extra credit for learning R.  Go here for more information.

FW 5411 - Applied Regression Analysis Regression as a tool for the analysis of forest and environmental science data. Topics include multiple linear, curvilinear and non-linear regression, hierarchial and grouped data and mixed-effects models. Emphasis is placed on application of tools to real-world data. Credits: 3.0 Lec-Rec-Lab: (3-0-0). Restrictions: Must be enrolled in one of the following Level(s): Graduate

Course Syllabus

You can download the 2006 syllabus here.

Textbooks

Weisberg, S. 2005. Applied Linear Regression (3rd Ed.).  John Wiley & Sons, Inc., Hoboken, NJ. 310 p.

Grading

Your course grade has four parts: homework (40%), a midterm exam (20%), a final exam (20%)
and a project (20%).

Projects

For the project you will identify three papers from your field that rely upon or use regression as a core analytical tool.  Together you and I will choose one, and you will prepare a written critical review.  Later, you will present it to the class.

Project for 2006: for information about it and a schedule of presentations for it

Assignments and Readings

Readings and homework assignments will be given periodically and will be drawn from questions in each chapter of Weisberg's text.  Please see the sidebar at right for details.

The following is course material from Spring 2005.  The course is being revised to improve it for Fall 2006.  A new textbook is being selected, and assignments will be somewhat more straightforward an more frequent.

You can download the 2005 syllabus here.

Assignments

Assignment One
Assignment Two and the data
Assignment Three and the data
Assignment Four
Assignment Five and the data

Lecture notes, examples and handouts:  these are available to class members only.

• Lecture 02, January 12, 2005
• Lecture 03, January 14, 2005
• A brief tutorial on R, the R script that generates the results and some data used by the script
• An annotated example of simple regression using R
• An R script for a multiple regression example using the wine data from Montgomery et al. (2001)
• An example for multicollinearity, using R, with associated data
• You can find the Catalog of Curves for Curve Fitting here
• An example of weighted regression, also using R, with associated data
• An example suitable for indicator variables, with script in R and associated data
• A handout on subset selection using all possible combinations, and an R script to accompany it.
• An excerpt from Schabenberger and Pierce (2002) discussing non-linear regression.
• Click here to download Rykiel, E.J. Jr. 1996. Testing ecological models: the meaning of validation. Ecol. Model. 90: 229-244. This paper is copyrighted and available to class members only.
• Click here to download Robinson, A.P. and R.E. Froese. 2004. Model validation using equivalence tests. Ecol. Model. 176: 349-358. This paper is copyrighted and available to class members only.