Simulation-based Tests for Regression Models with IID Errors: Some Standard Cases

Abstract

Many applied studies involve the estimation and analysis of a linear regression model with IID errors. Several tests for this model were described in Chapter 1 and comments were made about the possible dangers of using asymptotic theory as the foundation for inference in empirical investigations. The purpose of this chapter is to show how simulation methods, which were discussed in Chapter 2, can be used to provide an improved basis for testing. The general structure of this chapter follows that of Chapter 2 in so far as exact Monte Carlo techniques are illustrated before the more generally applicable (but only asymptotically valid) nonparametric bootstrap methods are considered. When appropriate, the specific examples are accompanied by some comments on general issues relevant to applied work.KeywordsBootstrap TestMonte Carlo TestRejection ProbabilityAsymptotic TestBootstrap SchemeThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.