Application of Advanced Epidemiological Methods · Term 1, 2026W
Repository: https://github.com/ehsanx/SPPH604-2026W — browse or clone every file, including the sources behind these pages.
This page is the student-facing map of the public repository. Use the HTML links to read course material in your browser, PDF links to download or print, and Source links only when you want to inspect the underlying Markdown / Quarto file.
Canvas is used for submissions, grading, rubric grids, Zoom links, AI-use examples, and weekly Modules. This repository contains the course materials themselves.
| Start here | Purpose |
|---|---|
| Course overview | Short orientation to the course, assessment arc, folders, demonstration paper, and AI-use rules |
| Syllabus | Authoritative on deadlines, weights, grading, and policy |
| L0 — Setup | Do before the first class. R, RStudio, TinyTeX, GitHub, and a test knit |
| Practice overview — PDF | How P1–P5 build your project toward M1 |
| Complete file catalogue | Every compiled student-facing document, with its HTML and PDF versions |
| Licence | CC BY 4.0 for instructor-authored material, with stated carve-outs |
| If you want to… | Use |
|---|---|
| Read a lecture, lab, practice increment, or milestone in your browser | HTML |
| Download, print, or annotate the same document | |
| Inspect or edit how a document was written | Source (.md, .qmd,
.Rmd) |
| Find every compiled document in one place | ARTIFACTS.html |
Most students should normally use HTML or PDF. Source files are included for transparency and reproducibility; they are not the preferred reading format.
Orient → Lock → Build → Replicate → Review → Improve & Defend → Reanalyse
| Stage | What happens |
|---|---|
| Week 1 — Orientation | Form project groups, complete setup, understand the course arc, and begin scanning candidate papers |
| M0 — Lock | Establish that one paper, question, dataset, and analysis plan are feasible |
| P1–P5 — Build | Apply each week’s method to your own paper; optional and ungraded, but they become the building blocks of M1 |
| M1 — Replicate + interrogate | Assemble the first complete analysis and identify the most consequential threat to interpretation |
| M2 — Review | Critique another group’s actual M1; reviewers are free to disagree with the group’s own ranking of threats |
| M3 — Improve + defend | Improve the analysis and defend decisions. Changes may be reviewer-prompted, self-initiated, or tested-and-retained |
| M4 — Reanalyse | Carry forward the best-defended analysis and write the final reanalysis letter with the reproducible repository |
The syllabus controls if anything here differs from the formal schedule or assessment policy.
Labs use fixed teaching data to practise the methods that you later apply to your own paper.
| Lab | Topic | Read online | Source | Feeds | |
|---|---|---|---|---|---|
| L0 | Setup | HTML | md | — | |
| L1 | NHANES data wrangling | HTML | md | — | |
| L2 | NHANES–mortality linkage | HTML | md | P1 | |
| L3 | Confounding & collapsibility | HTML | md | P2 | |
| L4 | Effect modification & interaction | HTML | md | P3 | |
| L5 | Complex survey analysis | HTML | md | P4 | |
| L6 | Missing data & multiple imputation | HTML | md | P5 |
These are optional and ungraded. They apply the week’s method to your group’s paper. Nothing is submitted separately, but the work accumulates into M1 and supports later review and defense.
| Increment | Topic | Read online | Source | Worked example | |
|---|---|---|---|---|---|
| P1 | Analytic cohort | HTML | md | Example | |
| P2 | Confounding & Table 1 | HTML | md | Example | |
| P3 | Effect modification | HTML | md | Example | |
| P4 | Design-aware estimate | HTML | md | Example | |
| P5 | Missing-data sensitivity | HTML | md | Example |
Practice overview — PDF · Practice overview — source
| Milestone | What you do | Read online | Source | Worked example | |
|---|---|---|---|---|---|
| M0 | Feasibility gate — lock a viable paper and analysis plan | HTML | md | Example | |
| M1 | Replication + interrogation — assemble the project and make the dominant-threat judgment | HTML | md | Example | |
| M2 | Critique another group’s actual M1 | HTML | md | Example | |
| M3 | Improve + defend — respond to review, make any additional changes you identify, and justify final decisions | HTML | md | Example | |
| M4 | Final reanalysis letter + reproducible repository | HTML | md | Example |
Read the M1 → M2 → M3 → M4 worked examples in order. They form one continuous scientific-review story rather than four independent model answers.
| Week | Topic | Lecture plan | Slide deck |
|---|---|---|---|
| 1 | Intro & reproducibility | HTML · PDF · Source | HTML · PDF · Source |
| 2 | NHANES & reading methods | HTML · PDF · Source | HTML · PDF · Source |
| 3 | Cohort construction & FLI | HTML · PDF · Source | HTML · PDF · Source |
| 4 | Confounding & Table 1 | HTML · PDF · Source | HTML · PDF · Source |
| 5 | Interaction & effect modification | HTML · PDF · Source | HTML · PDF · Source |
| 6 | Complex survey design | HTML · PDF · Source | HTML · PDF · Source |
| 7 | Missing data | HTML · PDF · Source | HTML · PDF · Source |
| Deck | What it covers |
|---|---|
| Decomposing the study aim | Turning the paper’s published aim into a PECOT research question, and the NHANES variables and component files each piece rests on. Source |
| From a Methods section to an analytic cohort | How one published Methods section becomes a dataset — the ambiguities it leaves, the decision each one forces, and what each decision costs in people. PDF · Source |
| The same sequence, audited | The reference version of the deck above: every count carrying the file and line that produced it, closing with a ledger of every decision and its cost. PDF · Source |
Work through this before L1. It names every NHANES variable the demonstration paper uses and the component file each one comes from, so the merge in the lab is not the first time you meet them.
Later weeks shift from lectures to milestone presentations, review, defense, and the closing session. See the syllabus for the formal schedule.
The worked examples and reproduction pipeline use the demonstration paper throughout.
| Resource | What it is |
|---|---|
paper/ |
Demonstration paper, the additional P4 paper, and licence/attribution information |
examples/ |
Worked examples for P1–P5 and M0–M4 |
reproduction/ |
Full R analysis pipeline behind the demonstration-paper results |
| Building the analytic dataset — step by step | The recipe. For each of the eight steps: the paper’s own sentence, the NHANES component and variables that sentence forces you to read, the decision we had to make and why, and the N afterwards |
| NHANES variable map | Every variable the pipeline uses: component file, cycle availability, reserved codes vs measurements, CDC codebook link, and the line of our code that reads it |
| Reproduction walkthrough | How the scripts fit together, the sample funnel, and key NHANES decisions |
| Reproduction report | What reproduced, what did not, and why |
Read them in that order: the recipe for how the cohort was built, the variable map to look a variable up, and the report for what reproduced and what did not.
| Folder / file | Purpose |
|---|---|
syllabus/ |
Formal course syllabus |
labs/ |
L0 setup and L1–L6 lab handouts |
lectures/ |
Weekly lecture plans and slide decks |
practice/ |
P1–P5 project increments |
milestones/ |
M0–M4 assignment briefs |
examples/ |
Worked model answers and exemplar artifacts |
paper/ |
Demonstration/companion papers and attribution |
reproduction/ |
Reproducible analysis code and reports |
tools/ |
Repository rendering utilities; students normally do not need this folder |
| README.html | Course/repository overview |
| ARTIFACTS.html | Complete compiled-artifact catalogue |
README.md, ARTIFACTS.md,
index.md |
Source versions of the root pages |
LICENSE.md |
Licence for instructor-authored material |
THIRD_PARTY_NOTICES.md |
Third-party licence and attribution notices |
Estimated cost of required materials: $0.