Dynamic analysis is the very effective method for both the improvement of traditional cross section model and the analysis of travel behavior. The study tries to apply dynamic analysis method in travel behavior into the stated preference for route choice and to find the changing structure in time. For these purposes, dynamic model is established by LISREL to validate its significance and to improve the cross section model. As results of applied model, variable parameters on travel time information are statistically significant with high values, which mean that they have greatly an effect on route choice. While roads which penetrate the downtown are regarded as negative to the route choice, one that is a roundabout road is considered as affirmative. The model containing the concept of both serial correlation and stated dependence shows the highest goodness of fit and is reasonable to be applied to dynamic behavior analysis. As the time interval becomes large, parameter values of serial correlation and stated dependence become small. That is, serial correlation and stated dependence has inverse proportion ratio to the time interval.
展开▼
