Measures of Association

Summary

Defines measures of association as the quantitative comparisons between exposed and unexposed groups, building risk ratios, incidence rate ratios, and odds ratios from the two-by-two table and worked Brazilian water-cistern and migraine-rate examples. Shows why the odds ratio is the only valid ratio measure for case-control designs through its cross-product symmetry, then maps each measure to a study design and the rare-disease assumption that makes them numerically converge. Covers attributable risk and attributable fraction, population-level versions for policy decisions, and the hypothesis-testing and confidence-interval machinery that distinguishes strength of association from statistical significance.