FW 536 · Oregon State University · accessible edition
Statistical Modeling for Ecology and Conservation
A five-day workshop covering probability, distributions, GLMs, mixed models, model selection, maximum likelihood, and Bayesian hierarchical modeling in R, JAGS, and Nimble. This is the screen-reader-optimized version of the course site.
New here?
Read the accessibility guide first — it explains how every page is laid out, how to navigate the math, and how assignments are submitted.How each day works
Each day has a morning and an afternoon session. Every session includes:
- An accessible lecture-notes page (a text transcript of the slides).
- A plain-language summary of the concepts.
- A practice lab with answers you can reveal (for in-class work).
- One assessed problem set per session — the 20-point graded assignment — with a matching hand-in template.
- A matching R Markdown hand-in template for each problem-set version.
- Any datasets and R scripts used, in the day's
data/folder.
-
Day 1 — Probability & discrete distributions
Plain-language summary Day 1 — the ideas in normal English A short, no-math explanation of what probability, random variables, and distributions actually mean. Read it before the lab to preview the ideas, or afterwards to consolidate what you did. Open the summaryPre-course exam is taken at the start of Day 1 morning: pre-course exam (template: pre_course_exam.Rmd).
Morning — Probability (20 points)
- Lecture notes: Introduction to probability
- Practice lab (answers visible): morning practice
- Graded problem set: problem set (#morning)
- Hand-in template: morning_lab_template.Rmd
Afternoon — Discrete distributions (20 points)
- Lecture notes: Discrete distributions
- Practice lab: afternoon practice
- Graded problem set: problem set (#afternoon)
- Hand-in template: afternoon_lab_template.Rmd
Practice answer key: worked solutions for the Day 1 practice lab.
-
Day 2 — Continuous distributions, linear models & GLMs
Plain-language summary Day 2 — the ideas in normal English A short, no-math explanation of what a GLM is actually doing: why we need a link function, and what a log or logit coefficient means biologically. Read it before the lab to preview the ideas, or afterwards to consolidate what you did. Open the summaryCompanion guide: Log & logit link intuition (16 worked biological examples).
Morning — Continuous distributions (20 points)
- Lecture notes: Continuous distributions
- Practice lab: morning practice
- Graded problem set: problem set (#morning)
- Hand-in template: morning_lab_template.Rmd
Afternoon — Linear models & GLMs (20 points)
- Lecture notes: Generalized linear models
- Practice lab: afternoon practice
- Graded problem set: problem set (#afternoon)
- Hand-in template: afternoon_lab_template.Rmd
Practice answer key: worked solutions for the Day 2 practice lab.
-
Day 3 — Mixed models & model selection
Plain-language summary Day 3 — the ideas in normal English A short, no-math explanation of what a random effect actually does, why partial pooling beats both extremes, and what model selection can and cannot tell you. Read it before the lab to preview the ideas, or afterwards to consolidate what you did. Open the summaryMorning — Mixed models (20 points)
- Lecture notes: Mixed effects models
- Practice lab: morning practice
- Graded problem set: problem set (#morning)
- Hand-in template: morning_lab_template.Rmd
Afternoon — Model selection (20 points)
- Lecture notes: Hypothesis tests & model selection
- Practice lab: afternoon practice
- Graded problem set: problem set (#afternoon)
- Hand-in template: afternoon_lab_template.Rmd
Practice answer key: worked solutions for the Day 3 practice lab.
-
Day 4 — Maximum likelihood & Bayes I
Plain-language summary Day 4 — the ideas in normal English A short, no-math explanation of what a likelihood really is, what maximum likelihood is actually doing, and how Bayes turns a prior plus data into a posterior. Read it before the lab to preview the ideas, or afterwards to consolidate what you did. Open the summaryTopic split
The morning covers maximum likelihood only (problems 1–3). Bayesian methods are introduced in the afternoon (problems 4–10).Morning — Maximum likelihood (20 points)
- Lecture notes: Maximum likelihood
- Practice lab: morning practice
- Graded problem set: problem set (#morning)
- Hand-in template: morning_lab_template.Rmd
Afternoon — Bayes I (20 points)
- Lecture notes: Bayesian statistics I
- Practice lab: afternoon practice
- Graded problem set: problem set (#afternoon)
- Hand-in template: afternoon_lab_template.Rmd
- Nimble script: logistic growth
Practice answer key: worked solutions for the Day 4 practice lab.
-
Day 5 — Bayesian hierarchical modeling
Plain-language summary Day 5 — the ideas in normal English A short, no-math explanation of hierarchical models, latent states, and derived quantities — what each is for, and why ecologists reach for them. Read it before the lab to preview the ideas, or afterwards to consolidate what you did. Open the summaryMorning — Bayes II (20 points)
- Lecture notes: Bayesian statistics II
- Practice lab: morning practice
- Graded problem set: problem set (#morning)
- Hand-in template: morning_lab_template.Rmd
- Nimble scripts: lizards, spawner-recruit, dynamic occupancy, N-mixture, multilevel
Afternoon — Bayesian modeling & catch-up
- Open lab time: keep fitting Bayesian models — finish the Day 5 problem set and catch up on any earlier problem sets, with the instructor available to help.
Practice answer key: worked solutions for the Day 5 practice lab.
Notes for this accessible edition
- The original PowerPoint files are still in each day's folder, but the lecture-notes pages above are the recommended, screen-reader-friendly way to read the lectures.
- Some lecture figures still need text descriptions; if you reach one you need, ask the instructor.
- Where a problem says to plot and inspect, you may report and interpret the numeric results instead. See the accessibility guide for details.