Abstract
Although it is said that disaggregate mode choice models have good transferability, methods of updating the models have been proposed to improve it. However, temporal transferability of updated models to improve spatial transferability has rarely been analysed and factors affecting it have not been uncovered. This paper has dealt with one of the updating methods, updating alternative-specific constants utilising disaggregate data, and aimed to investigate factors affecting the temporal transferability of the constants. In the analysis, repeated cross-sectional data collected in the Chukyo Metropolitan Area are divided to generate many application areas effectively. The analysis has shown that the factors can relate not only to regional characteristics but also to traffic conditions in the past and irreversibility of travel behaviour.
