Panel Data Static Model: Testing of Hypotheses

Panel data are used to estimate the unrestricted model, fully restricted model, and error component model. A fixed-effects model considers how heterogeneity from entities affects individual intercepts, while a random-effects model hypothesises that error variance structures are affected by entities. Disturbances in a random-effects model are assumed to be randomly distributed across entities. The key difference between fixed- and random- effects models is that individual effect in a random-effects model should not be correlated with any regressor. A formal test is needed to be carried out to find out which model would be appropriate for a given dataset. This chapter deals with testing of hypotheses to examine these issues.