Estimation under Heteroscedasticity: A Comparative Approach Using Cross-Sectional Data

Abstract

A comparative investigation was done analytically for 4 different Estimation Techniques of a newly-designed Audit Fees model with four exogenous variables. The aim is to explore in depth the effects of the problem of heteroscedasticity in a CLRM of cross-sectional data and to determine an appropriate estimation technique(s) in the presence of such heteroscedasticity. Findings revealed that the estimates are virtually identical for three estimators: OLS, WH and NW, while the performance of the fourth estimator, GLS was found to be outstanding, as it completely eliminates the effect of heteroscedasticity by producing a “BLUE” result.

Key Words: Heteroscedasticity, Audit Fees, Exogenous variables, New-design and BLUE.