Multilevel Modeling of Longitudinal and Functional Data
I signed on to write this commentary because I was confused. Thirty-two years of teaching multivariate statistics to psychology graduate students, and of being a consultant on research projects in a range of disciplines, had not left me with a clear sense of what multilevel analysis was all about. Seemingly related terms in statistics journals, such as hierarchical linear model (HLM), empirical Bayes estimation, random coefficient model, variance component, repeated measures, longitudinal data analysis, latent curve analysis, and even the term multilevel itself seem to have similar but not completely consistent implications. Moreover, my own research ...