Overview and Learning Objectives

FW8051 Statistics for Ecologists

Learning Objective

Develop and demonstrate model literacy: be able to fit, evaluate, and describe a variety of statistical models

Fit: consider frequentist and Bayesian approaches in R/JAGS
Evaluate: understand assumptions and how to evaluate them
Describe models using equations and text
Interpret model parameters and link estimates of parameters in computer output to model equations

Will consider models for:

Normal (i.e., Gaussian) data that are independent and have constant variance (linear regression)
Non-linear predictor-response relationships
Data with non-constant variance (generalized least squares)
Count and binary data (generalized linear models)
Correlated data (mixed models, generalized estimating equations, cluster-level bootstrap).

Learning Objective

Be able to choose an appropriate method depending on characteristics of the data.

Learning Objective

Gain an appreciation for challenges associated with selecting among competing models and performing multi-model inference.

We will briefly cover some popular model-building strategies
We will consider directed acyclical graphs and their implications for choosing appropriate models

Learning Objective

Conduct research using a workflow that maximizes ‘reproducibility’ of your work.

Homework and in-class assignments will give you experience with tools in R/Rstudio for reproducible research

Homework Assignments

I use a combination of self- and peer-evaluations:

To make this work, it is important that you turn in assignments on time, so please plan ahead if you will be away or have a busy stretch.

Is so important (and challenging these days).

Communicate with me as early as possible if you are struggling to focus, keep up, understand the material, or meet expected deadlines.

Submit your answers in the form of a reproducible (pdf, or word) document created using R/Rstudio.

Self assessment: view the answer key and then note: a) any areas where your understanding was lacking prior to viewing the answer key; and b) whether or not any aspects of the problem (or solution) remain unclear after having viewed the answer key.

Peer assessment: review your peer’s solution for completeness and accuracy and provide comments to help the student improve their understanding. If your peer clearly understood and fully completed the original assignment, please note this in your peer review.

And, an overall rating on a 1-5 scale. I will adjust final scores!

Comments from previous year’s course evaluations

I like the self and peer assessments. It helps to be forced to review answers and process, otherwise I would just move to the next 10 things I have to work on.
Homework grading format is awesome: by evaluating ourselves and our peer before instructor sees our work, we are able to better understand mistakes that we made and /or others made and subsequently offer suggestions in a “learning by teaching” fashion.
I found the peer assessments to be highly valuable. The code we were exposed to throughout the course plus looking at more variations within the peer assessments has brought my level of understanding up significantly.

S/N students will be required to complete all homework assignments and self and peer reviews, but will not have to take the exams or quizzes.

9 Grad Programs:

Characteristics:

Treat this class as a sport, where we are all on the same team!