An attitudinal modal choice model

Abstract

Explanatory variables in this model of urban residents' mode choice behavior for the home-to-work trip are attitudinal ratings by survey respondents of descriptive attributes of modes perceived to be available to them. The model methodology is presented in three parts: individuals are clustered with respect to homogeneity of their choice constraints; the covariance structure of attitudinal variables is characterized through development of latent factors describing respondents' perceptions of choice alternatives; and logit probabilistic choice models are estimated in terms of subsets of descriptive attribute ratings chosen to represent latent perception factors. Empirical results from application of the model in the Ottawa, Ontario/Hull, Quebec Canadian National Capital Region are extremely encouraging. Goodness-of-fit summaries indicate model performance better than that obtained using conventional logit models based upon perceived time and cost data. Moreover, it is proposed that the attitudinal models provide valuable insights concerning the importance of factors underlying the choice decision structure of the sample population.