Discrete Modal Choice Model in suburban districts

In this paper, a kind of Discrete Modal Choice Model is developed as an Multinomial Logit Model. In this model the RP (revealed preference) and VOT of SP (state preference) surveys are used to calibrate the significance of a multinomial Logit model, and a survey and analysis of Shunyi district, one of new urban area of Beijing, which located in north of Beijing, is introduced as a case study. First, cross-tables analysis method was used to select four class variables from seven as the significant. Then a multinomial Logit module of SPSS is given as the model result with and without VOT per trips variable. Further, the accuracy of the model with VOT per trips is better then the other models are founded. From the result, the conclusion is given, that there are advantages to jointly estimating the models with RP and SP surveys.