Week 1 — Course Introduction, Project Map & Setup

Land acknowledgement

UBC’s Point Grey Campus is located on the traditional, ancestral, and unceded territory of the xʷməθkʷəy̓əm (Musqueam) people.

This land has long been a place of learning, knowledge sharing, culture, history, and community.

By the end of today, you should know…

• how the term-long project works
• how M0–M4, L1–L6, and P1–P5 fit together
• what counts as an eligible project paper
• the course rules for reproducibility, collaboration, and AI use
• what to set up before the first regular class/lab week

The big idea: one paper, all term

You choose one recent, open-data paper and work with it throughout the course.

Replicate → Review → Improve → Defend → Synthesize

• Replicate & interrogate your own paper
• Review another group’s replication
• Improve your own analysis in response to the critique you receive
• Defend what you changed — and what you chose not to change
• Synthesize the final argument in a short reanalysis letter

The regular weekly rhythm

Weeks 2–7

• Tuesday 9:30–12:00 · Zoom — method + epidemiological reasoning
• Thursday 10:00–12:00 · SPPH 143 — fixed-data lab (L1–L6)
• P1–P5 — apply the same idea to your own project

P1–P5 are optional, formative, ungraded, and not separately submitted — no deadline of their own.
They are the building blocks of M1, and the practice you draw on at M2 (critiquing another group’s M1) and M3 (defending your own, alone).

Later in term

From Weeks 8–12 there is no scheduled Thursday lab; Tuesday shifts toward presentations, defense, and instructor-supported project work.

The milestone map

What each milestone asks you to do

• M0 — Feasibility: prove that your chosen paper can support the whole course
• M1 — Replicate + interrogate: rebuild the main result and identify the dominant threat
• M2 — Review: test and critique another group’s M1 professionally and constructively
• Between M2 and M3 — Improve: decide which reviewer points to act on in your own analysis
• M3 — Defend: explain what you changed, what you rejected, and why
• M4 — Synthesize: turn the full argument into a concise reanalysis letter + reproducible archive

How you are graded

The course rewards traceable reasoning, not merely reproducing the paper’s exact number.

Choosing your project: the M0 feasibility gate

A paper qualifies only if your group can point to all six:

1. Reconstructable public microdata
2. A clear primary exposure/predictor and outcome
3. Plausible measured confounders for a defensible adjustment set
4. Enough observations/events for the main analysis and at least one subgroup/interaction analysis
5. The design variables your analysis has to respect — survey weights/strata/PSU, or follow-up time and censoring
6. A genuine incomplete-observation problem — a variable your locked analysis uses that is missing for records which should have carried it

Do not add variables simply to manufacture a missing-data problem.

Reproducibility is the standard

• Every submitted number, figure, and table should be generated by code
• The pipeline must run start to finish on a fresh machine without manual intervention
• Your group works in a shared private GitHub repository during the term
• M1 and M4 include a downloaded reproducible repository zip
• M2 is slides only — you rerun the assigned group’s M1 materials, but submit no new code archive

Code that only works on one person’s laptop is unfinished analysis.

AI is a co-pilot — accountability stays with you

• AI tools are permitted; none is required
• You may use them for appropriate coding, debugging, drafting, and preparation within the rules of each assessment
• You remain responsible for auditing the code and defending the epidemiological reasoning
• M0, M1, M2 and M4 each carry a cover note: contributions, external feedback, AI use
  • final slide or final page of what you submit — plus COVER-NOTE.md in the repo zip at M1 and M4
• M3 preparation is unrestricted and undisclosed — rehearse with an AI if you want; there is no M3 submission to disclose it in
• Presentation: use everything you prepared — slides, your own speaker notes, your own code and output
• Q&A: the answer is yours, spoken — no prepared answer read aloud, no help from anyone in any channel, nothing produced during the session (no AI, no live prompting, no messages)
• Supports approved through your Academic Accommodation Letter are the exception, arranged in advance

“The AI made a mistake” is not a defense for an unaudited result.

Logistics: what happens where

• Today: Thu Sept 10 · 10:00–12:00 · Zoom
• Weeks 2–7: Tue class Zoom · Thu lab in person, SPPH 143
• Weeks 8–12: no scheduled Thursday lab; Tuesday becomes project/presentation time
• Canvas: Zoom links, weekly modules, rubrics, announcements, and official submission instructions
• Learning materials: two open-access textbooks + free software; nothing to purchase
• Communication: administrative email to the instructor; include “SPPH 604” in the subject line
• Accommodation letters: send to the instructor within the first two weeks where applicable

Our demonstration paper (the only paper slide today)

• Kueh et al., BMJ Open 2026 — MASLD & NHANES 2007–2018
• Adults with fatty-liver disease classified into four body-shape groups using BMI × waist-to-height ratio
• Headline: the non-obese/high-waist group shows the poorest survival
• Across the term, we will rebuild the analysis from public data and interrogate the choices behind it

Today is the what and why. The methodological how begins next week and unfolds just in time.

This week: get set up

Before next week’s regular class/lab rhythm:

• Install R + RStudio + Quarto
• Install Git and create/access a GitHub account
• Render one small Quarto document successfully
• Make a first Git commit and push to a private test/project repository
• Form your group
• Begin scanning recent open-data papers for M0

Target for today: leave with the toolchain working — or know exactly what still needs troubleshooting.

Bridge to Week 2

Tuesday Sept 15 · Zoom

• survey data sources, especially NHANES
• how to read a paper’s Methods critically
• feasibility and peer-review principles

Thursday Sept 17 · SPPH 143

• L1: data wrangling on fixed teaching data
• begin shortlisting papers toward M0

Come with: a working toolchain and 1–2 plausible candidate papers.