A method for breaking down and measuring the effects of correlative explanatory variables : an application to the effects of urban sprawl, car ownership and transport supply on change in the market of public transport.

In transport analysis in order to explain the changes of a factor we generally have to deal with several correlative explanatory variables. Quantifying how each variable affects the study object is therefore often difficult. Although this is a familiar problem in the context of modelling it is frequently poorly handled. In a situation like this change in the study object cannot always be expressed in a simple manner as the sum or product of the changes in the explanatory variables. To deal with this question the authors have proposed a method to break down and measure the effect of each explanatory variable. The paper starts with a description of the method and then applies it to the case of changes in the market share of public transport.