Biostatistics Seminar: Part 2 of 2: Feedback Modeling of Correlated Binary and Continuous Outcomes with Multiple Membership Structure

The Biostatistics Consultation Core continues its Biostatistics Seminar Series with a two-part seminar addressing key challenges in modeling correlated longitudinal data with time-varying covariates. The seminar introduces a Bayesian analytical framework and illustrates its application using NHANES data.

Part 2:
This presentation introduces a Bayesian framework for analyzing longitudinal data with correlated binary and continuous outcomes, time-dependent covariates, and overlapping clustering. The model captures feedback between outcomes and predictors over time, estimated via MCMC. Applications to NHANES data illustrate how prior health outcomes influence treatment intensity and adherence dynamics.

Find Part 1 here: Biostatistics Seminar: Part 1 of 2: Introduction to Bayesian Data Analysis and Multi-Membership Multi-Level Modeling.