On the Dummy Variable Technique and Covariance Analysis in Testing Equality Among Sets of Coefficients in Linear Regressions: An Expository Note

Abstract

Econometric research in the past two decades has vitnessed a considerable use of dummy variables in regression analysis. The analysis of covariance has long been a standard statistical technique to test the equality of coefficients in linear regressions. While students and researchers are generally aware of the close relationship between the two methods, they are often frustrated at choosing one method instead of the other in practice and wonder whether the two methods lead to the same test results. This note shows that the two methods are equivalent from the point of view of hypothesis testing.