本研究利用敘述偏好數據建立多屬性效用模式進行運輸替選方案之評選,在模式的構建過程同時探討屬性的可接受門檻與無異門檻問題。研究結果發現利用線性羅機模式所建立之多屬性效用函數解釋能力尚可,大部分屬性均相當顯著。此結果表示利用敘述偏好數據建立多屬性效用模式確實可找出各屬性間的交互損益關係。在考慮同一決策者之決策具一致性,而不同決策者間具異質性之假設下,所建立的可接受門檻或無異門檻模式約可大幅增加模式的解釋能力,由於大部分的屬性都有適用於部分決策者的可接受門檻或無異門檻,故對各屬性間的交互損益關係有相當大的影響。無異門檻模式的解釋能力優於可接受門檻模式。同時考慮兩種門檻的整合模式的解釋能力最佳,所有係數均相當顯著且符號正確,各屬性係數值間之交互損益關係也相當合理。
This research established multi-attribute utility models with stated preference data to evaluate transportation projects. We discussed the attribute thresholds of acceptance and indifference. The results showed that the multi-attribute utility function of linear logit model had relatively good explanatory ability and most attributes were significant. This means that multi-attribute utility model established with stated preference data can indeed find the tradeoff between attributes, Assuming that an individual should behave consistently and different individuals may behave heterogeneously, we found that the threshold models could greatly increase explanatory power Due to the fact that most attributes had threshold of acceptance and/or inc4fference, the tradeoff between attributes were greatly affected. Of the two threshold models, the threshold of indifference model had better explanatory power. The integrated model with both thresholds had best explanatory power. Its coefficients of all attributes were significant and had the correct signs. The tradeoffs between attributes were reasonable.
為了持續優化網站功能與使用者體驗,本網站將Cookies分析技術用於網站營運、分析和個人化服務之目的。
若您繼續瀏覽本網站,即表示您同意本網站使用Cookies。