Table of Contents Table of Contents
Previous Page  9 / 104 Next Page
Information
Show Menu
Previous Page 9 / 104 Next Page
Page Background

Bolet´ın de Estad´ıstica e Investigaci´on Operativa

Vol. 32, No. 3, Noviembre 2016, pp. 184-202

Estad´ıstica

Treatment of collinearity through orthogonal regression:

an economic application

Rom´an Salmer´on and Catalina Garc´ıa

Department of Quantitative Methods for Economics and Business

University of Granada (Spain)

!

romansg@ugr.es

,

!

cbgarcia@ugr.es

Jos´e Garc´ıa

Department of Economic and Business

University of Almer´ıa (Spain)

!

jgarcia@ual.es

Claudia Garc´ıa

University of Granada (Spain)

!

claugar@correo.ugr.es

Abstract

The relation between the regressors is usual in the application of multi-

ple linear regression in many different fields. When collinearity exists, the

estimation of the coefficients of the regressors is unstable and the inference

can lead to misleading conclusions. In this paper the successive orthog-

onal regression is applied as an alternative methodology to mitigate the

collinearity. The contribution of this paper is illustrated with an empirical

application to estimate the influence of infrastructures in production for

the 17 autonomous communities of Spain during 2002-2011. The model

presents collinearity that is mitigated by the proposed methodology ob-

taining new conclusions regarding to the interpretation of the parameters.

Keywords:

Infrastructures, productivity, orthogonal regression, collinear-

ity.

AMS Subject classifications:

C13, C51, H54, R11, O47.

c

2016 SEIO