Second Edition · Oxford University Press · 2028

Epidemiology Matters

An introduction to the methodological foundations of epidemiology — and to the seven steps that make studies matter for the health of populations.

Katherine M. Keyes · Mohammed Abba-Aji · Sandro Galea

L. Graunt Port Snow Farr Harbour Graunt Fields Nightingale Point Hill Crossing THE ISLE OF FARRLANDIA POPULATION 1,000 • EVERY RESIDENT KNOWN N 0 10 leagues reported case resident, well post road

Fig. 1. The Isle of Farrlandia — the book’s teaching population, named for William Farr, founder of modern vital statistics. Every method in the book is tested against a population whose truth you can see.

“The death rate is a fact; anything beyond this is an inference.”
William Farr, 1807–1883 WEEKS SINCE FIRST REPORTED CASE CASES

Fig. 2. Weekly reported cases, Port Snow outbreak. Chapter 4 traces this curve from the first count to the last inference.

Who this is for

One book, one population, three audiences

I.

For Students

Learn epidemiology by doing. Manipulate populations, watch associations form, introduce bias on purpose, and see exactly what it does to your estimates. No formulas required to start — just curiosity.

II.

For Instructors

Replace a lecture with a live demonstration. Preset scenarios show confounding, Simpson’s paradox, and selection bias unfolding in real time, and each interactive exercise is keyed to a chapter of the book.

III.

For Everyone

Epidemiology shapes policy, headlines, and daily life. Understand how we come to know what causes disease — and why it matters who we study, how we study them, and who gets counted.

The framework

Seven steps for an epidemiology of consequence

The book organizes epidemiologic thinking around seven foundational steps. Each builds on the last; together they form a complete route from a question worth asking to an answer worth acting on.

1
Define the population
2
Measure exposures & outcomes
3
Take a sample
4
Estimate associations
5
Evaluate causality
6
Assess interaction
7
Assess external validity

Seven decisions, made deliberately, are what separate a study that is merely published from a study that matters.

Interactive companion

Learn by doing

Each tool maps directly to the book. Use them in class, assign them as exercises, or explore on your own. The print edition includes QR codes linking to each module.

Plate ILive

Causal Thinking Lab

Build a data-generating process. Add confounding, measurement error, and selection bias, and watch them distort causal estimates. Five preset scenarios for classroom use.

Chapters 7–10 · Causation & Bias
E D C
Plate IILive

Population & Prevalence

Adjust disease and exposure in a population of Farrlandians. Watch prevalence, risk ratios, and risk differences update in real time.

Chapters 2–6 · Population & Measures
Plate IIILive

Component Cause Model

Toggle causes for individual Farrlandians. Disease occurs only when a sufficient cause is complete: different people, different pathways, same disease.

Chapters 7 & 11 · Causation & Interaction
Plate IVLive

Study Design Builder

Walk through all seven steps to design your own epidemiologic study of consequence, from defining a population to judging how far your answer travels.

All Chapters · Seven-Step Framework
Plate VLive

Explore Farrlandia

One thousand Farrlandians with full profiles. Hover to meet them. Filter, stratify, and compute associations from the living dataset behind every example.

All Chapters · The Living Population
Plate VIComing 2027

External Validity Explorer

Compare causal effects across populations with different distributions of component causes, and see why the same exposure can matter more in one place than another.

Chapter 12 · External Validity
Plate VIIComing 2027

Screening Simulator

Adjust sensitivity, specificity, and prevalence, and watch predictive values change. Encounter lead-time and length bias through animation.

Chapter 13 · Screening
TP FP FN TN
New in the second edition

Consequence Boxes

Methods are never neutral. In every chapter, a Consequence Box connects one methodological choice to a question of social responsibility — because how we study health decides whose health we see.

Chapter 3 · Measurement

How we define a health indicator determines who counts as a case. Who counts determines what risk factors we find. What we find determines what gets funded.

Consider depression. A clinical interview identifies fewer cases than a screening tool — and those missed are disproportionately people without access to clinical settings. The measurement choice is not neutral.

When you choose a measure, ask: whose health does this make visible, and whose does it obscure?

Launching Fall 2028

The Farrlandia Challenge

An inter-school competition for public health students. An outbreak has struck Farrlandia: investigate it, identify its causes, propose interventions — and compete against teams from schools across the country.

4 Rounds, 4 WeeksEach round follows the seven-step framework: define, investigate, challenge, intervene.
Real ToolsTeams use the platform’s interactive modules to analyze the Farrlandia dataset.
Free to EnterOpen to any team of 3–5 students enrolled in a public health program.
Consequentialist ScoringJudged on rigor, population impact, communication, and creativity.
Get Notified When Registration Opens
The authors

Keyes, Abba-Aji & Galea

Stay Updated

Get word when new tools launch, when the Farrlandia Challenge opens for registration, and when the second edition publishes.