An Evaluation of Second-Order Latent Growth Models for Nonlinear Growth Trajectories

• In Event: Methods for Multilevel, Categorical, and Multidimensional Data

Abstract

With longitudinal data on the rise, the use of second-order latent growth models (SOLGMs) is building prominence. The current paper investigates growth parameter recovery in SOLGMs that follow a variety of functional form specifications. In particular, linear, quadratic, and piecewise growth trajectories were used to specify the data generation and estimation for SOLGMs. A simulation study reveals that SOLGMs exhibit parameter bias for variance components under a quadratic and piecewise model, even when these are correctly specified. When growth trajectories are misspecified, parameter bias is also introduced for SOLGMs with linear growth trajectories.

Authors

• Daniel Peche Gonzalez, The University of Texas at Austin

• Tiffany Ann Whittaker, The University of Texas at Austin