Course Syllabus

Professor: Dr. Erin Lake

This is an intermediate, applied biostatistics course in both classic and modern regression methods for the analysis of continuous, binary, polytomous, ordinal, count and survival (time to event) response data. The course covers linear, generalized linear, and survival models, as well as their extensions. The course implements the methods in R, SAS and Stata. Parametric, semi-parametric, non-parametric, additive, and regularized approaches are explored. Missing data methods, visualization and graphical techniques, diagnostics, transformations, confounding and effect modification, model building and selection, model assessment and validation, goodness of fit, over-dispersion, model interpretation, and power and sample size calculations for select models, constitute additional topics covered. Discussion focuses on the choice of model, model strengths and weaknesses, and applications. Examples and data are drawn from local hospitals and research institutions around the world. The course conveys the ubiquity and relevance of the regression model throughout most areas of modern science and underscores its foundational presence in machine/deep learning, and artificial intelligence (AI).