Mattingly "AI & Prompt Design: The Basics of Prompt Design"
6.1 big picture
1. Big picture
• Design that captures the experience of a cohort
(whether an actual cohort study or a more
theoretical “study base”) by collecting data from
a subset of the cohort
– Incident cases of disease are identified
– Controls (non-diseased) are selected from the study
base that gave rise to the cases (in one of several
ways depending on the design)
• Best understood as a sample of the denominator for a rate or
risk
– Exposure information is gathered on the cases and
controls (using interviews, records, biological tests
etc.)
3. Big picture
• When disease is rare, cohort studies are
impractical and case-control are the best
alternative
• When exposure assessment is expensive, case-
control provides a way to minimize costs by
allowing exposure assessment on only a sample
of a cohort