ABSTRACT
This chapter identifies five rationales for conducting longitudinal research: direct identification of intraindividual changes; direct identification of interindividual differences in intraindividual changes; examining interrelationships in intraindividual changes; analyses of causes of intraindividual changes; and analyses of causes of interindividual differences in intraindividual changes. It discusses two statistical issues related to the second rationale. Specifically, the first issue is on how to appropriately and powerfully detect interindividual differences in intraindividual changes by variance testing, and the second issue is on the consequences of ignoring interindividual differences in intraindividual changes on statistical inferences about latent factor means or fixed effects. The chapter reviews the development of latent curve modeling and describe a generic latent linear growth curve model. It also discusses the consequences of ignoring interindividual differences in intraindividual changes on point estimation and standard error estimation of latent factor means in linear growth curve modeling.
