Xiao-Hua Andrew Zhou
Advanced-level topics in biostatistics offered by regular and visiting faculty. Prerequisite: permission of instructor. Offered: jointly with STAT 578; AWSpS.
In this course we will discuss various statistical methods for the analysis of missing data. Methods considered will include likelihood-based, weighted GEE, multiple imputation, and Bayesian approaches. Computational tools include the EM algorithm and extensions and the Gibbs' sampler. We will also cover both ignorable and non-ignorable missing-data mechanisms. Study designs considered will include both cross-sectional and longitudinal ones. Finally, we will discuss the application of the general methodology to several active areas of research, including causal inferences for randomized trials with non-compliance and ROC curve methodology.
Student learning goals
Understand the theorems behind the current methods for the analysis of data with missing values
Are able to apply the learned methods to a real project
Learn some new research areas with missing data problems
General method of instruction
Lectures and discussions
Class assignments and grading
Grade as Pass or Fail