Institute of Transportation, MOTC

　　In this study stated preference data were used to examine travel behavior for passengers traveling between Tainan and Taipei. The purpose is to discuss the effects of choice set generation, individual’s heterogeneity, and the inclusion of new alternative in the market on individual traveler’s choice behavior. The results showed that two-stage choice models, whose choice sets were endogenously decided, had very good explanatory power. Travel time and travel cost variables affected both the choice set generation stage and alternative choice stage. Including heterogeneity in the choice set generation stage or alternative choice stage would significantly increase models’ explanatory power. The effect of heterogeneity in the alternative choice stage is greater than that of the choice set generation stage. The model considering heterogeneity in both stages obtained best results. The heterogeneity in different stages would affect each other. The new alternative high-speed rail would affect individual traveler’s choice set and alternative choice.