Conventions

This page summarizes the visual conventions used throughout the book.

Commands and output

This book mixes terminal commands, AI prompts, and document source. Shell commands appear in syntax-highlighted blocks. A command the reader types is shown after a $ prompt; the prompt itself is not typed. Output appears on the following lines, without a prompt:

$ claude --version
1.5.0

Prompts sent to an AI tool are shown in fenced blocks labeled prompt; the tool’s response, where shown, follows in a fenced block labeled response:

Summarize the methods section of the attached paper in
three sentences, and flag any claim that is not supported
by a citation.
The paper uses a stepped-wedge cluster-randomized design
across six sites... [response abbreviated for illustration]

Inline commands, file names, and tool names are set in monospace. Placeholders that the reader must replace with their own values are written in angle brackets, for example claude "<your prompt here>". The angle brackets are not typed.

Callouts

Three callout types appear:

TipTip

A small practical recommendation.

A short question testing comprehension of the just-read material. Click to expand the answer.

WarningWarning

A pitfall the reader may otherwise hit, most often a place where an AI tool’s output looks correct but is not.

Verification markers

Because this book is about judgment as much as mechanics, we mark claims about AI output explicitly:

  • Verified means the reader (or the author) checked the claim against a primary source or by running the code.
  • Unverified means the claim is exactly as the AI tool produced it and has not yet been checked.

This distinction recurs throughout the book, most heavily on Day 2 (citations) and Day 3 (code).

Cross-references

Within this book, sections are referenced by their Quarto label (@sec-day1, @sec-verification). These resolve to clickable links in HTML and section numbers in PDF.

References to the companion volumes R for Biostatistics, Git and GitHub for Biostatistics, and the Biostatistics Practicum use prose pointers rather than Quarto cross-references, because cross-references do not resolve across separate books. For example: ‘see the version-control chapter of the companion Git and GitHub for Biostatistics volume’.

Chapter structure

Every content chapter follows the same template:

  1. Learning objectives. What the reader will be able to do after the day.
  2. Lecture. The substantive content, with prompts to try and outputs to evaluate in the reader’s own environment as the chapter proceeds. Each major subsection ends with a short Check your understanding callout (see above); attempt it before moving on.
  3. Further reading. The primary sources, studies, and documented incidents behind the chapter’s claims, so that a claim in this book is never the end of the trail. Follow these before taking any claim on this book’s word alone.
  4. Worked example. A small but realistic research scenario that uses the day’s content end to end.
  5. Homework. Problems organized from easier to harder.
  6. Solutions. Worked solutions to all homework problems. Read only after attempting the problem.

Each chapter closes with a short What’s next pointer. The pattern repeats deliberately; by the third chapter the reader knows where to find each component without looking.