台灣高鐵 (High Speed Rail) 在2007年加入台灣西部運輸市場,但由於高鐵為一新型運具且票價較其他運具高,因此造成民眾搭乘意願低落,為此高鐵公司利用藍橘優惠(15% off and 35% off)以調整票價促進國人搭乘意願。高鐵與一般運具最大不同點為能夠依據不同類型的旅客,而給予不同的服務,但就各種車廂的運量資料顯示,商務車廂票數卻僅占總票數的2.6%~3%,因此本研究將針對高鐵商務車廂為主題進行相關討論。在研究中探討民眾對於未來搭乘商務車廂意願,以及現行商務車廂與商務車廂加入新服務時的願付價格 (willingness-to-pay),在探討搭乘意願方面,建立Logistic Regression Model、Poisson Model探討搭乘意願選擇模式,此外,並建立Hurdle Model探討有意願搭乘與無意願搭乘與有意願但尚不搭乘的部分對其作比較,以歸納影響旅客對於商務車廂搭乘意願的顯著因素;而在願付價格方面,由於會有少部分民眾對於搭乘商務車廂皆無意願之樣本,為解決願付價格為0的問題,在情境中利用Spike Model進行校估並與一般常用的Logit Model作比較,並分別針對服務不變下與增加新服務之商務車廂願付價格進行探討,希望能夠藉此了解旅客對於商務車廂願付價格之變化情形。
High Speed Rail (HSR) has been merging to the transportation pattern in western Taiwan since 2007. However, the demand of passengers still seems low due to the newly developed transit system and higher cost compared with other transport. In order to enhance usage of transit, High Speed Rail Corp. has presented special promotions such as 15% and 35% discounts. Besides, the major difference between High Speed Rail and other public transport systems is providing different services based on different type of passengers however, statistic results indicate that only 2.6% to 3% of total passengers are taking business cabin. Hence this study investigates a detail inside into the subject of the service of business cabin from HSR. This study focuses on the willingness of passenger traveling by business cabin and the willingness-to-pay on the business cabin and additional services. Hurdle model is used to discuss about the willingness of passenger traveling by business cabin, we are not only developing two perspectives about willing and unwilling on the traveling by business cabin but also define a perspective regarding the people who desire but unable to travel by business cabin. Besides, Logistic Regression Model, and Poisson Model can be applying to investigate a choice model which analyzes the potential effects on passengers’ preference for business cabin. Bias can be happening due to people may not be willing to take business cabin which leads to “zero” on the acceptable price can be included during research. In order to understand passengers’ preference accurately under different scenarios, Spike model can be sorted zero WTP and comparing with other popular models such as the logit model.