osenberg@zoology.ufl.edu
8:30 - 10:25 am (Tuesday)
8:30 - 9:20 am (Thursday)
521 Carr Hall
Instructor: Craig W. Osenberg (627 Bartram Hall)
osenberg@zoology.ufl.edu
Additional Information:
Discussion and critiques of quantitative methods and experimental designs applicable to ecological studies. Emphasis will be placed on the links among hypothesis generation, mathematical and statistical models, experimental and sampling designs, and statistical analysis and interpretation of data. Examples will be drawn from the ecological literature to highlight strengths and weaknesses of different approaches. This is NOT a course on statistics or experimental design, per se. Instead, it will emphasize their application to answering (rather than obscuring) fundamental ecological questions.Format:
The course will meet for approximately 2 hours on Tuesday and 1 hour on Thursday. Each day will provide a mix of lecture and class discussion. Readings from the primary literature will be assigned weekly, and students will participate in the weekly discussions. There will be a series of relatively minor assignments during the term, and a large class project (using meta-analysis) due at the end of the term. The class project will require time during the entire semester. Each student will work on the class project -- depending on enrollment, there may be two projects. The class will write a paper on the results of the project, and each student will also provide a individual critical, written review of the paper submitted by the group. Grades will be assigned based on class participation and quality of work devoted to the class projects.Readings and Prerequisites:
Enrollment:
- Required readings will be taken from the literature and made available as pdfs (contact the instructor for access).
- We will use a specialized software package to conduct meta-analyses relevant to the class project: MetaWin is available from Sinauer. You do not need to purchase it. This software will be available in the Zoology Computer Teaching Facility (611 Carr).
- All students should also have had a course in general ecology (at the undergraduate or graduate level), and familiarity with basic statistics (e.g., anova, ancova, regression).
The course is open to graduate students and advanced undergraduates. The class size is limited. Students interested in taking the course should see Karen Pallone in the Zoology office. She will have you complete an enrollment form; we will let you know when you have been given permission to enroll.Generic Syllabus:
| Week | Topic |
| 1 | Hypothesis testing in ecology. |
| 2 | Biological vs. statistical significance; Estimation vs. null hypothesis tests; Statistical power; Randomization tests; Bayesian statistics. |
| 3 | Meta-Analysis and quantitative synthesis; Quantifying relative importance (limitation, indirect effects). |
| 4 | Group project ideas and presentations.
(two page proposals due) |
| 5 | The interpretation of cause-effect in laboratory and field experiments
-- controls, nesting, confounding factors.
(Sept 24: Select group projects) |
| 6 | Replication, pseudoreplication, and solutions when experiments lack replication; Impact assessment designs. |
| 7 | Work on group projects (literature search, data selection, documentation) |
| 8 | Meta-analysis: computational issues (MetaWin). |
| 9 | Ratios and transformations: effects on estimation and interpretation. |
| 10 | Correcting for "nuisance" variables, scaling, and body size. |
| 11 | Work on group projects |
| 12 | Allometry, regression, and general structural models: effects on estimation;
Predator selectivity, prey mortality rates, and estimation of selection
gradients in evolutionary biology.
(Group papers due in class: Nov 14th) |
| 13 | Null models in ecology; multivariate analyses.
(Review of class paper(s) due in class: Nov 21st) |
| 14 | Multivariate analyses -- analysis and interpretation of complex datasets. |
| 15 | Project presentation and discussion |
osenberg@zoology.ufl.edu