Behavior Genetics

, Volume 33, Issue 2, pp 137–159

Structural Equation Models for Evaluating Dynamic Concepts Within Longitudinal Twin Analyses

  • John J. McArdle
  • Fumiaki Hamagami

DOI: 10.1023/A:1022553901851

Cite this article as:
McArdle, J.J. & Hamagami, F. Behav Genet (2003) 33: 137. doi:10.1023/A:1022553901851


A great deal of prior research using structural equation models has focused on longitudinal analyses and biometric analyses. Some of this research has even considered the simultaneous analysis of both kinds of analytic problems. The key benefits of these kinds of analyses come from the estimation of novel parameters, such as the heritability of changes. This paper discusses some recent extensions of longitudinal multivariate models that can be informative within biometric designs. In the methods section we review a previous latent growth structural equation analysis of the New York Twin (NYT) longitudinal data (from McArdle et al., 1998). In the models section we recast this growth model in terms of latent difference scores, add several new dynamic components, including coupling parameters, and consider biometric components and examine model stability. In the results section we present new univariate and bivariate dynamic estimates and tests of various dynamic hypotheses for the NYT data, and we consider a few ways to interpret the age-related biometric components of these models. In the discussion we consider our limitations and present suggestions for future dynamic-genetic research.

Longitudinal twin analyses dynamic structural equation modeling fluid and crystallized intelligence Wechsler Adult Intelligence Tests New York Twin Study 

Copyright information

© Plenum Publishing Corporation 2003

Authors and Affiliations

  • John J. McArdle
  • Fumiaki Hamagami

There are no affiliations available

Personalised recommendations