ABSTRACT
In this chapter, we seek to determine the implications of the violations of these assumptions on the properties of the least-squares estimators of the population regression parameters β0 and β1. (In what follows, these estimators will be termed OLS estimators.)
If Assumptions A.1 through A.5 are valid, then the OLS estimators for β0 and β1 are deemed good estimators. In the absence of any compelling evidence to the contrary, these assumptions are assumed to hold. But what if one or more of these assumptions are thought to be invalid?
