Political Analysis Advance Access published online on November 20, 2009
Political Analysis, doi:10.1093/pan/mpp033
| ||||||||||||||||||||||||||||||||||||||||||||||||
The Effect of Misclassifications in Probit Models: Monte Carlo Simulations and Applications
Département de science politique, Faculté des sciences économiques et sociales, Université de Genève, 40, Bd du Pont d'Arve, 1211 Genève 4, Switzerland e-mail: simon.hug{at}unige.ch (corresponding author)
The increased use of models with limited-dependent variables has allowed researchers to test important relationships in political science. Often, however, researchers employing such models fail to acknowledge that the violation of some basic assumptions has in part difference consequences in nonlinear models than in linear ones. In this paper, I demonstrate this for binary probit models in which the dependent variable is systematically miscoded. Contrary to the linear model, such misclassifications affect not only the estimate of the intercept but also those of the other coefficients. In a Monte Carlo simulation, I demonstrate that a model proposed by Hausman, Abrevaya, and Scott-Morton (1998, Misclassification of the dependent variable in a discrete-response setting. Journal of Econometrics 87:239–69) allows for correcting these biases in binary probit models. Empirical examples based on reanalyses of models explaining the occurrence of rebellions and civil wars demonstrate the problem that comes from neglecting these misclassifications.
Author’s note: This paper draws in part on work carried out with Thomas Christin, whom I wish to express my gratitude for extremely helpful research assistance. Thanks are also due to James Fearon and Patrick Regan for making available data used in this paper and to the anonymous reviewers and Dominic Senn for helpful comments on an earlier version of this paper.