PUBH 5500

Students will learn to map biostatistical techniques onto the study designs examined in Public Health 5003 without focusing on complicated mathematical formulas. Biostatistics II will build on Biostatistics I by teaching advanced model building and diagnostics for various forms of regression, with a focus on the analysis of incidence-based designs using proportional hazards regression and other techniques. Given the frequency of nonmonotonic associations in public health, methods of working with nonlinear associations (e.g., cubic splines) will be described. Students will learn multiplicative and additive effect modification and use of propensity scores to account for covariates in models. Bayesian statistics will be introduced given the growing importance of these methods in our field. Applications to real data will be emphasized using the R computer package. Prerequisite(s): Public Health 5001 AND Public Health 5003