Study designs involve carefully arranged measurements of exposures, outcomes and covariates. These data must then be processed in statistical analyses to identify and estimate the causal effect of the exposure after accounting for errors in measurement and distorted associations from confounding variables.
Classic epidemiologic study designs include cross-sectional, cohort and case-control methods, with many variants of each basic type. Thereafter, the analysis involves the application of statistical modeling to the observed data in such a way that all threats to validity are minimized or eliminated with quantitative analyses of the degree of uncertainty from both sampling and residual biases. This Department applies these methods to a wide array of problems in human health. Basic notions of causal inference, validity and precision can be applied broadly, so that a core sequence of epidemiologic methods relate to many populations regardless of their substantive focus.