Testing Different Panel Models With Standard Error Adjustments on Data With a Small Number of Clusters

• In Event: Application of Longitudinal Models and Statistical Theory

Abstract

The study provides a comprehensive analysis of three panel models with robust estimation adjustment in various conditions and fills the gaps showing the performance of standard error adjustments (i.e., cluster robust standard errors) for longitudinal data with a small number of clusters. The simulation results indicate a good performance of the correlated random effects (CRE) modeling approach when violating the key assumption of the random effects modeling approach. The CR2 adjustment works appropriately to correct standard error bias when panel data with a small number of individuals.

Authors

• Bixi Zhang, Graduate Center - CUNY

• Francis H. Lim Huang, University of Missouri