Publications & Resources
Latent Variable Modeling of Longitudinal and Multilevel Data
This report gives an overview of some aspects of latent variable modeling in the context of growth and clustered data. The author emphasizes the benefits that can be gained from multilevel as opposed to conventional modeling techniques that ignore the multilevel data structure. Large-scale educational surveys are used to illustrate key points. Examples are discussed where repeated observations are made on students sampled within classrooms and schools.
Muthén, B. (1996). Latent variable modeling of longitudinal and multilevel data (CSE Report 412). Los Angeles: University of California, Los Angeles, National Center for Research on Evaluation, Standards, and Student Testing (CRESST).