Labs

The reproducible computation strand (R + Quarto)

The labs are where Bayesian ideas become reproducible computation. Each lab is a hands-on Quarto (.qmd) walkthrough you run in local R + VS Code + Quarto: you build something, render it, and verify the output. The graded version of each lab — its prompt, rubric, points, and due date — lives in the LMS; these public pages are the reading/reference companion so you can practice the workflow.

How a lab works

Every lab follows the same loop, the habit we build all term:

1. Goal — what you will have produced (a rendered report, a figure, a posterior summary).
2. Setup — the starting state; the setup page covers install.
3. Steps — a real, runnable sequence with code you can read and adapt.
4. Verify — the “did it actually work?” check (re-render, sanity-check the numbers).
5. AI Use Note — Tool / Purpose / Verification whenever you used an assistant.

The code shown in these labs is runnable as written (seeds and synthetic data are specified). On this build the figures are generated with base-R graphics so they reproduce anywhere R is installed.

The labs

• Lab 4 — Beta-Binomial simulation — simulate and summarize a Beta posterior for a proportion; read a credible interval. (Accompanies Week 4.)
• Lab 7 — Grid approximation in Quarto — approximate a posterior on a grid, draw from it, and build a reproducible report. (Accompanies Week 7.)
• Lab 9 — Bayesian regression — fit and interpret a simple Bayesian regression; read coefficient uncertainty. (Accompanies Week 9.)
• Lab 13 — Partial pooling — compare no-pooling, complete-pooling, and partial-pooling estimates and see shrinkage. (Accompanies Week 13.)

Public vs. graded

These lab pages are public, ungraded walkthroughs. The graded lab deliverables, their rubrics, points, and due dates are authoritative in the LMS (Blackboard).