Statistical Methods for Analysis with Missing Data

3 credits

Introduces core concepts in statistical analysis with missing data including missingness mechanisms (missing completely at random, missing at random, and missing not at random). Covers a range of methods such as complete case analysis, weighting, likelihood-based and Bayesian approaches, and multiple imputation. Addresses strategies for addressing non-ignorable missingness. Applications primarily involve data from epidemiologic studies. Prerequisite: either BIOST 513, BIOST 515, BIOST 518, BIOST 570/STAT 570, or SOC 506/CSSS 507. Offered: jointly with EPI 531.