Offered Every winter semester (next offering: Winter 2025)
Last offered Winter 2024
Prerequisites: (BIOSTAT 522 and EPID 600 (or equivalent)) or PhD standing in Epidemiology
Advisory Prerequisites: BIOSTAT 523 is recommended
Description: This course is an applied epidemiologic data analysis using R statistical software, a free software environment for statistical computing and graphics, and especially covers unique features of environmental exposure data. The course will cover important statistical issues in environmental health data, such as log-transformation and standardization of chemical biomarkers. It will introduce parametric and non-parametric smoothing methods, such as natural splines, penalized splines, and locally weighted polynomial regression (LOESS), and how to implement them using generalized linear models (glm) and generalized additive models (gam). The course will also cover data analysis subject to the dependence issue, such as longitudinal data and complex sampling data (e.g., NHANES) and time-series analysis in air pollution health effects that are widely used in environmental epidemiology. This course will provide an opportunity to analyze actual population data to learn how to model environmental epidemiologic data, and is designed particularly for students who pursue environmental epidemiologic research. The course will consist of lectures and hands-on practices in computer labs, homework assignments and final projects.
