Logo image
How effective are the reset tests for omitted variables?
Journal article   Peer reviewed

How effective are the reset tests for omitted variables?

Siu Fai Leung and Shihti Yu
Communications in Statistics Part B: Simulation and Computation, Vol.29(4), pp.879-902
2000

Abstract

Dummy variable Linear regression model Model diagnostics Omitted variable RESET Specification error
In regression analysis, RESET has widely been regarded as an effective diagnostic test especially for omitted variables. This paper investigates the limitations of the existing RESET tests in detecting omitted variables. We analyze the sources from which RESET draws its power and point out the circumstances under which RESET will likely be ineffective. We offer some Monte Carlo evidence as well as an empirical application to illustrate the weaknesses of the RESET tests. A more robust RESET type test is proposed. Copyright © 2000 by Marcel Dekker, Inc.

Metrics

1 Record Views

Details

Logo image