A Method for Breaking Down and Measuring the Effects of Correlative Explanatory Variables: a Application to the Effects of Urban Sprawl, Car Ownership and Transport Supply on Change in the Market Share of Public Transport



A Method for Breaking Down and Measuring the Effects of Correlative Explanatory Variables: a Application to the Effects of Urban Sprawl, Car Ownership and Transport Supply on Change in the Market Share of Public Transport

Authors

BONNEL P and CABANNE I, Laboratoire d'Economie des Transports, France

Description

Most analysis of travel behaviour attempts to measure the individual impacts of a variety of explanatory factors on the investigated phenomenon. The same is true of modelling, which aims to identify the contribution of a number of factors in order to simu

Abstract

Most analysis of travel behaviour attempts to measure the individual impacts of a variety of explanatory factors on the investigated phenomenon. The same is true of modelling, which aims to identify the contribution of a number of factors in order to simulate how a change in these would affect the studied situation. However, both these types of studies encounter methodological problems. The different explanatory variables are frequently correlated, as are the ways they change. This raises difficulties for assessing and interpreting the respective impacts of the different factors.

To overcome this difficulty this paper proposes a methodology for separating the impacts of the different variables. This methodology involves decorrelation of the fundamental variables and their change and a systematic study of the combined effects of the various factors. We shall then apply this methodology to a situation whose determinants are highly correlated in order to demonstrate the value of the technique.

We shall apply the technique to car/public transport modal split. The explanatory Variables relate to locations, car ownership and transport supply. These are important input variables in most urban models. The Lyon conurbation in France has been selected, for which we have the results of three household travel surveys conducted between 1976 and 1995.

Publisher

Association for European Transport