Beyond Mean: A Review of Location Scale Models With Random Effects

• In Event: New Insights for Designing Multilevel Models: Predicting Random Effects, Using Predictor Aggregates, and Mediation Planning

Abstract

Location scale models with random effects are important in education to explore effects beyond mean while incorporating the higher-level heterogeneity. There are multiple modeling approaches and estimation methods for location scale models. However, the software implementations are not user-friendly for education researchers to be applied. The present proposal summarizes the existing approaches for location scale models with their software implementations for multilevel data. An empirical example was provided to illustrate how different approaches work in R programming with a result comparison.

Authors

• Bixi Zhang, Graduate Center - CUNY

• Francis H. Lim Huang, University of Missouri