Joint Modeling of Multivariate Longitudinal Profiles: Evaluation of Co-Development of Internalizing and Externalizing Problem Behaviors

• Activity Number: 605
• Type: Contributed
• Date/Time: Wednesday, August 7, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Mental Health Statistics Section
• Author(s): Pingfu Fu*+ and Guang Zeng
• Companies: Case Western Reserve University and Texas A&M University-Corpus Christi
• Keywords: internalizing ; externalizing ; behaviors ; bivariate mixed models ; association ; evolution

Abstract

The random-effects approach to joint modeling of multivariate longitudinal profiles has received a great deal of attention recently. The approach is especially useful when analyzing longitudinal data of two associated markers. In this study, we use a bivariate linear mixed model with random effects and independent measurement errors for both markers to assess how the evolution of internalizing problems is related to the evolution of externalizing problems and vice versa, using a nationally representative sample of 9796 kindergarteners from the ECLS-K. In line with the available literature which suggests the comorbidity of internalizing and externalizing problems, we show that there is a moderate positive association (r=0.32) between the evolutions of the two behavior problems. Additionally, there is a nonlinear and increasing trend of positive association between the two behavior problems over time, with the correlation coefficients ranging from 0.2 at kindergarten entry to 0.28 at fifth grade.