過去個體選擇模式考慮知覺與態度等不可觀測變數的方法，會有參數估計值不具一致性與有效性、指標無法直接影響決策、以及指標不適用於預測等問 題。本研究的目的在於整合潛在變數與個體選擇模式，使參數估計值具有一致性與有效性，並增加模式的解釋能力，使政策變數之直接或間接影響更容易了解。研究對象為高雄至臺北路線之城際巴士客運旅客。實證結果顯示加入服務品質滿意度後，選擇模式的解釋能力大幅提升。旅客對各客運公司之便利滿意度與舒適娛樂滿意度會直接影響他們的選擇行為，便利滿意度之影響程度較舒適娛樂滿意度強。旅客的互動滿意度則透過便利滿意度與舒適娛樂滿意度間接影響他們的選擇行為。

In the past, the way that the discrete choice models treated the unobservable variables, such as perception and attitude variables, caused some problems. For example, the estimated parameters were not consistent or efficient, the index could not be used directly to affect the policies, and the index may not be suitable in the prediction. The purpose of this research is to integrate the latent variables with the discrete choice models to make estimated parameters more consistent and efficient, to increase the explanatory power of the choice models, and to make the direct or indirect effects of the policy variables more easily understood. The research subjects were the intercity bus travelers of Taipei-Kaohsiung route in Taiwan. The empirical results showed that the inclusion of satisfaction of service quality variable significantly improved the explanation power of the choice model. The results also showed that the satisfaction of the travelers toward the bus companies’ convenience of service, comfort, and entertainment will directly affect their choice behavior. The effect of convenience satisfaction was stronger than that of comfort and entertainment satisfaction. The passengers’ interaction satisfaction indirectly affected their choice behavior through their convenience, comfort, and entertainment satisfaction.