A Linear-Linear Growth Model With Individual Change Points: The Maximum Likelihood and Bayesian Solutions

• In Event: Poster Session 15
  In Poster Session: Division D Poster Session

Abstract

The latent growth curve model with piecewise function is a useful analytics tool to investigate the growth trajectory consisted of distinct phrases of development in observed variables. However, most of the existing latent growth curve models assume the same change point for all subjects, therefore do not provide an effective approach to identify individual change points. In this study, we propose a simple linear-linear growth model for individual change point detection, along with two estimating algorithms including a maximum likelihood estimation and a Bayesian method. The new approach is applied to an Early Childhood Longitudinal Study, Kindergarten Class of 1998-99 (ECLS-K). The results demonstrate the effectiveness of our approach in identifying individual change points using longitudinal data.

Authors

• Ping Zhang, Wal-mart

• Qingyang Zhang, University of Arkansas at Fayetteville

• Wen-Juo Lo, University of Arkansas