CAUSAL INFERENCE, SURVEYS AND MISSING DATA FOR POPULATION HEALTH

Prerequisites & course information

Requisites: B M I/POP HLTH 552 , STAT/F&W ECOL 572 (or HORT 572 prior to Spring 2025), C&E SOC/SOC 361 , or ED PSYCH 761 Course Designation: Grad 50% - Counts toward 50% graduate coursework requirement Repeatable for Credit: No Last Taught: Fall 2026 Learning Outcomes: 1. Identify the design elements (stratification, clustering) and sampling weights (unequal probability sampling, non-response adjustments) of a complex survey design and perform regression analyses of complex survey data to answer epidemiologic and public health research questions using standard statistical software (SAS and R) Audience: Graduate 2. Identify missing data mechanisms and their potential impacts on naive statistical analyses Audience: Graduate 3. Explain the strengths and weaknesses of various strategies for dealing with missing data (weighting, maximum likelihood, Bayesian, multiple imputation) and implement them using standard statistical software (SAS and R) Audience: Graduate 4. Define individual and average causal effects and identify challenges (confounding, selection bias) to their estimation in observational studies Audience: Graduate 5. Identify and implement statistical methods for estimating causal effects (IP weighting, g-formulas, propensity scores, instrumental variables) in settings where their use is appropriate Audience: Graduate Learning Outcomes: 1. Identify the design elements (stratification, clustering) and sampling weights (unequal probability sampling, non-response adjustments) of a complex survey design and perform regression analyses of complex survey data to answer epidemiologic and public health research questions using standard statistical software (SAS and R) Audience: Graduate 2. Identify missing data mechanisms and their potential impacts on naive statistical analyses Audience: Graduate 3. Explain the strengths and weaknesses of various strategies for dealing with missing data (weighting, maximum likelihood, Bayesian, multiple imputation) and implement them using standard statistical software (SAS and R) Audience: Graduate 4. Define individual and average causal effects and identify challenges (confounding, selection bias) to their estimation in observational studies Audience: Graduate 5. Identify and implement statistical methods for estimating causal effects (IP weighting, g-formulas, propensity scores, instrumental variables) in settings where their use is appropriate Audience: Graduate