A comparison of modeling strategies for value-added analysis of education data (Q1855626)

This paper deals with a multilevel model (based on hierarchically structured education data) which is an example of a common situation where classical assumptions of independence of explanatory variables and random effects are no longer valid. This fact causes difficulties in applications of standard estimation techniques. Two alternative estimation strategies are proposed: one using an instrumental variable approach and the other using a Bayesian analysis as available through the BUGS software. The authors conclude that the BUGS analysis has certain advantages over the classical estimation methods.