ABSTRACT
This chapter defines heteroskedasticity and delineates its consequences. Then three detection techniques are shown with a continuing example. Finally, two separate remedies are applied to the example—weighted least squares and heteroskedasticity-consistent standard errors. As always, the approach is to develop an intuitive conceptual understanding of the econometric problem. In addition, heteroskedasticity is a direct violation of the constant-variance assumption of the classical linear regression model. This violation and the resulting problems were sketched out in Chapter 8. In this chapter heteroskedasticity is more fully elucidated along with detection techniques and remedial measures.
