Introduction to Longitudinal Data Analysis Using the Traumatic Brain Injury Model Systems National Database

This presentation introduces latent class analysis (LCA) and latent profile analysis (LPA), person-centered statistical methods used to identify subgroups within a larger population based on shared characteristics. The presentation contrasts these techniques with traditional cluster analysis, highlighting LCA/LPA's advantages, such as incorporating covariates and providing model fit evaluations. A practical walkthrough using traumatic brain injury (TBI) data is described, demonstrating how to apply LCA/LPA, from data preparation to result interpretation. Several research studies utilizing LCA/LPA to analyze TBI are cited as examples. The presentation concludes by emphasizing the importance of considering data quality and the clinical relevance of identified subgroups.

Resources