We proved smoking kills without a single experiment
Epidemiology studies whole populations to find disease patterns that experiments can't ethically test. Four study types give different strengths of evidence — understanding what each can prove is the key to judging real research. Drag the slider to explore how study strength changes what we can claim.
Study Design Strength
Descriptive
Weaker designs describe patterns; stronger designs test causation
Population patterns only
Study Type
DescriptiveCase-ControlCohortIntervention
Evidence Strength
Weak
Can Prove Cause?
No
Incidence measures new cases appearing over time — the rate at which disease strikes. It's what changes when you intervene: smoking bans drop lung cancer incidence years later. Prevalence is a snapshot — total existing cases right now. High prevalence might mean high incidence or just people living longer with the disease. When folate fortification began in 1998, neural tube defect incidence fell immediately, but prevalence took years to drop as older cases aged out of the data.
Know This
Correlation never proves causation, but epidemiology builds the case through multiple studies — when cohort strength meets dose-response and biological plausibility, we act on the pattern without waiting for intervention trials.