Quantitative Ecology & Evolution (Multivariate Statistics) 215:575;
Sept. 7 INTRODUCTION ; DATA Chapter 1
14 DESCRIPTIVE STATISTICS;
CLUSTERING VS. ORDINATION Chapter 1
21 BASIC CLUSTERING
ALTERNATE CLUSTERING METHODS Chapter 3
28 No Class – PJM away
Oct. 5 No Class – PJM away
12 CHOOSING CLUSTERING METHODS Chapter 3
CLUSTERING WITH SAS
19 MATRIX ALGEBRA Handout
LINEAR TRANSFORMATIONS Chapter 2
26 PCA Chapter 2
PCA WITH SAS
Nov. 2 DIVISIVE CLASSIFICATION Chapter 5
9 M.V. HYPOTHESIS TESTS Chapter 4 & Handout
16 MANOVA Chapter 4 & Handout
23 No Class – Thanksgiving Break
30 MANOVA & DISCRIMINANT ANALYSIS Chapter 4 & Handout
Dec. 7 REPEATED MEASURES ANALYSIS Handout
Project I is due on 26 Oct. Project II is due on Dec. 7 See reverse for details
McGarigal, K., S. Cushman,
and S. Stafford. 2000. Multivariate Statistics for Wildlife and Ecology Research.
P. J. Morin; e-mail: pjmorin "at" rci "dot" rutgers "dot" edu , office - ENR 148, phone 2-3214
Quantitative Ecology and Evolution (MULTIVARIATE STATISTICS) - 215:575
1. The goal of your FIRST project is to gain familiarity with the scientific literature involving one or more methods of multivariate statistical data analysis. The project involves writing a short essay (5-10 pages max. less references, typed, double-spaced) reviewing the use of one or more types of multivariate analysis in ecology or evolution. The essay should focus on the specific attributes of a given class of techniques, and should emphasize the ecological or evolutionary interpretation of the technique. Potential topics include cluster analysis, ordination, discriminant analysis, or any other topics listed on the syllabus.
2. The project will involve reviewing some of the original literature on a given topic. References listed in the bibliography of your text should offer a useful entry point into the literature. This is not meant to be an onerous time-consuming exercise. Use it to your advantage to learn about an approach that you plan to use in your own research.
1. The goal of your SECOND project is to gain familiarity with one or more methods of MULTIVARIATE data analysis, USING SAS. Ideally, the analysis should be of your own data. If you lack data, we can provide you with some. The analysis should describe some aspect of the data set that is not obvious from simple inspection. Possibilities include a cluster analysis of community samples describing species abundance patterns, analysis of spatial patterns, ordinations of small data sets, discriminant analysis of differences between sets of samples, etc.
2. Your analysis should be described in a short (10 pages max., typed, double-spaced) report divided into the following sections:
A. INTRODUCTION : briefly describe the phenomenon that you are analyzing, and the questions that you attempted to answer.
B. MATERIALS AND METHODS : describe the data, how the data were collected, the analytical techniques used, and why those techniques were selected.
C. RESULTS : concisely describe the outcome of the analysis. One table or figure can and should replace many pages of prose.
D. DISCUSSION : interpret the biological significance of your results.
E. LITERATURE CITED : list any references cited in the text of your report. See a recent issue of the journal ECOLOGY
F. APPENDIX : append copies of your raw data, computer output, and/or any additional calculations.