Latent Growth Models With Floors, Ceilings and Random Knots

• In Event: Growth Curve Modeling Extensions and Factor Analysis Choices

Abstract

Latent growth modeling is a structural equation modeling approach for understanding individuals’ longitudinal growth. Such models typically presume the same functional form holds throughout the time span under examination. When the outcome variable is a rating scale or a score reflecting task acquisition, as is often the case, researchers typically ignore the fact that the outcome has a floor and ceiling and that the appropriate growth function is implicitly piecewise. The current paper will derive and illustrate linear latent growth models to accommodate measured outcome variables that are bounded by known floors and/or ceilings, and illustrate how such models can be fit to data using conventional structural equation modeling software.

Authors

• Gregory R. Hancock, University of Maryland

• Yi Feng, University of Maryland - College Park

• Jeff R. Harring, University of Maryland - College Park

• Hemant Kher, University of Delaware