Longitudinal Analysis

In longitudinal studies, subjects are followed for a period of time and outcomes along with other characteristics for each subject are measured at multiple pre-specified time points. Unlike data collected from cross-sectional studies where outcomes and other characteristics are measured at only one time point for each subject, data collected from longitudinal studies allow investigators to study changes in response over time. However, designing and analyzing longitudinal studies require careful considerations. Two major issues that arise in longitudinal studies are the correlation between measurements from the same subject and potential missing data due to loss of follow-up. In this talk, we will introduce common statistical analysis methods that account for the correlation structure of longitudinal data. We will also introduce missing data mechanism and approaches to address this important issue.