Error Covariance Structure in a Latent Quadratic Effects Model

• In Event: Modeling Latent Variable Interactions and Multiple-Group Models

Abstract

This research evaluates the effects of non-normal indicator error terms on estimation of latent quadratic effects. An unconstrained modeling approach was used to observed the extent to which the parameter estimates and standard errors change when an error covariance structure is not specified for the indicators of the linear effects and quadratic effects. Varying levels of non-normality and different SEM estimation procedures were simulated. Our results suggest the importance of an error covariance structure when errors are non-normal.

Authors

• Kristina Cassiday, University of Maryland - College Park

• Daniel Y Lee, University of Maryland - College Park

• Gregory R. Hancock, University of Maryland