ABSTRACT
This chapter deals with some basic issues of multilevel models (MLM) for longitudinal data analysis. Traditional methods of assessing change and development are often unsatisfactory because of violations of statistical assumptions and because they do not model individual change. Modern longitudinal data analysis methods, including MLM, provide an opportunity to model dynamic fluctuations in individual data across time. The base model is not a longitudinal model, but it does allow one to estimate total between- and within-person variance in the outcome data with which subsequent longitudinal models may be compared. MLM are a powerful set of statistical techniques that overcome some of the problems previously associated with analyzing longitudinal data. To test whether there is an overall difference between treatment conditions, for example, on the dependent variable, one can specify a conditional intercept model.
