Advanced Regression for Correlated Data

3 credits · Graduate School

This course is the second in a two-course sequence on advanced regression methodology. This course develops the theoretical and methodological foundations of regression models for correlated and longitudinal data. Topics include generalized least squares, likelihood-based mixed-effects models, generalized estimating equations (GEE), transition and marginalized models, missing data mechanisms, and inference under incomplete data. Connections between likelihood-based and estimating equation-based semiparametric inference frameworks are emphasized. The course examines asymptotic theory for correlated data, efficiency considerations, model specification and identifiability, and the theoretical properties of estimators under dependence. Advanced topics may include semiparametric efficiency, robustness under misspecification, and modern extensions relevant to contemporary biomedical research. The course places substantial emphasis on theoretical development alongside methodological and computational components. This course is open to biostatistics PhD students only. This course is co-located with and serves as repeat credit for BIOS 7346. Prerequisite: BIOS 8345. [3]