Schedule (topic flow)
A 15-week pacing plan — the LMS is authoritative for dates and deadlines
This is a pacing plan, not a calendar. It shows the order and arc of topics. Dates, deadlines, exam windows, and project checkpoints are authoritative in Blackboard (the LMS) and are not set here.
The term moves in five parts, from “what does it mean to reason with uncertainty?” to fitting, checking, and communicating Bayesian models.
Part I — Foundations of Bayesian reasoning (Weeks 1–2)
| Week | Theme | The throughline |
|---|---|---|
| 1 | Thinking under uncertainty | Probability as a language for uncertainty; what changes when evidence arrives. |
| 2 | Discrete Bayes & diagnostic reasoning | Conditional probability, base rates, false positives, and updating. |
Part II — Building Bayesian models and the posterior (Weeks 3–7)
| Week | Theme | The throughline |
|---|---|---|
| 3 | Prior, likelihood, posterior | Building a model from assumptions and data; posterior ∝ likelihood × prior. |
| 4 | The Beta-Binomial model | Proportions, conjugacy, posterior simulation, credible intervals. (Lab 4) |
| 5 | Prior sensitivity & summaries | How conclusions change when assumptions change. |
| 6 | Posterior predictive thinking | Prediction, replicated data, and checking fit. |
| 7 | Simulation-first computation | Grid approximation, posterior draws, reproducible Quarto. (Lab 7) |
Part III — Synthesis & midterm (Week 8)
| Week | Theme | The throughline |
|---|---|---|
| 8 | Midterm synthesis | Pulling together updating, simple models, summaries, and interpretation. |
Part IV — Bayesian regression & model checking (Weeks 9–12)
| Week | Theme | The throughline |
|---|---|---|
| 9 | Bayesian regression I | Modeling a numerical outcome; priors on regression parameters. (Lab 9) |
| 10 | Bayesian regression II | Multiple predictors, binary outcomes, prediction. |
| 11 | Model checking & comparison | Posterior predictive checks, predictive performance, model limits. |
| 12 | Bayesian & classical in conversation | Credible vs. confidence intervals, posterior probabilities vs. p-values. |
Part V — Hierarchy, decisions & project (Weeks 13–15)
| Week | Theme | The throughline |
|---|---|---|
| 13 | Hierarchical models | Complete pooling, no pooling, partial pooling, and shrinkage. (Lab 13) |
| 14 | Decisions & communication | Decision thresholds, uncertainty communication, project work. |
| 15 | Project workshop & synthesis | Final report preparation and course synthesis. |
Holiday weeks, the exam window, and project checkpoints adjust this rhythm; the LMS carries the actual calendar. The final exam is cumulative and scheduled in the registrar’s final-exam block.