Bayesian Methods
for Missing Covariates
in Longitudinal Studies

Short Course at the Conference of the
International Biometric Society

Emmanuel Lesaffre, Nicole Erler

Riga, July 10th, 2022

Outline

  1. What, why and how?
  2. Examples of longitudinal data
  3. Frequentist methods for longitudinal data
  4. Missing data processes
  5. Bayesian methods for longitudinal studies
  6. Bayesian Analysis of Incomplete Data with JointAI
  • Practicals
  1. Summary & Extensions

Software

To follow the practicals, please make sure you have a recent version of installed (e.g., version 4.2.1).

In addition, you need to install JAGS , a freely available software used to perform MCMC sampling using the Gibbs sampler. The most recent version is 4.3.1.

Note that for versions ≤ 4.1.3, you need to use JAGS version 4.3.0.

Once both programs are installed, please also install the package JointAI (the development version on GitHub has some features not yet on CRAN):

devtools::install_github("NErler/JointAI")

Materials

.zip file with all materials:

About the Instructors

Prof. emeritus Emmanuel Lesaffre is the author of the Wiley book Bayesian Biostatistics (2012), and leading editor of the Chapman and Hall book Bayesian Methods in Pharmaceutical Research. He has taught courses on Bayesian methods in clinical studies on Master level at the University of Hasselt and KU Leuven both in Belgium, Erasmus Medical Center (Rotterdam, the Netherlands) and that for more than 25 years. He has also taught short course on Bayesian methods at various universities abroad and clinical research institutes.

Dr. Nicole Erler is a PostDoc at the department of Biostatistics at the Erasmus Medical Center (Rotterdam, the Netherlands) and obtained her PhD on the topic of Bayesian Imputation of Missing Covariates. She is teaching courses on Missing Data in Clinical Research and introductory courses to Biostatistics and R at the Netherlands Institute of Health Sciences. Nicole is author and maintainer of the R package JointAI that performs fully Bayesian analysis of incomplete data.