Random Coefficient Models for Longitudinal Data: The Autism Example
Brady Thomas West, Kathleen Bell Welch, Andrzej T. Gałecki, Brenda W. Gillespie · 2022
This chapter illustrates fitting random coefficient models to data arising from a longitudinal study of the social development of children with autism, whose socialization scores were observed at ages 2, 3, 5, 9, and 13 years. It considers models that allow the child-specific coefficients describing individual time trajectories to vary randomly. Random coefficient models are often used for the analysis of longitudinal data when the researcher is interested in modeling the effects of time and other time-varying covariates at Level 1 of the model on a continuous dependent variable, and also wishes to investigate the amount of between-subject variance in the effects of the covariates across Level 2 units. In the context of growth and development over time, random coefficient models are often referred to as growth curve models. Random coefficient models may also be employed in the analysis of clustered data, when the effects of Level 1 covariates, such as student’s socioeconomic status, tend to vary across clusters.