An Analytical Framework for Research Questions With Variability as the Outcome Variable

• In Event: Methodological Challenges in Multilevel Modeling

Abstract

In some research scenarios, examining intra-individual stability as a personal level characteristic, or intra-group cohesiveness as a group/team-level characteristic, may be of key interest. For these types of research questions, a special modeling framework is required, one that can model variance of the lower-level unit as a random coefficient at the higher level. In the current study, a random variance modeling framework based on multilevel structural equation modeling is proposed. The conceptual development of the modeling framework will be presented, followed by a brief discussion of the estimation methods. Illustrative examples using Bayesian estimation will be provided. The proposed modeling framework is an innovative and powerful tool that has broad implications and many possible extensions.

Authors

• Yi Feng, University of Maryland - College Park

• Gregory R. Hancock, University of Maryland