偏遠地區居民運具選擇行為之影響因素分析

Abstract

偏遠地區公共運輸偏低的乘載率及不符當地居民需求的服務型態，一直是一項亟待克服的難題。因此，了解偏遠地區居民的運具選擇行為及偏好，無疑是改善偏遠地區公共運輸服務的重要步驟。惟傳統運具選擇行為的相關研究，過度重視量化的變數，而忽略質化的服務品質變數，致許多公共運輸服務的質化改善無法反應至運具選擇行為中。基此，本研究選擇臺北都會區相對偏遠的三峽鶯歌地區居民作為主要研究對象。在運具選擇方案上，除當地現有之運具外，更以陳述性偏好設計方式，加入需求反應式公共運輸。在變數選擇上除常採用之量化變數外，更利用驗證性因素分析技術，納入運輸服務之安全性、舒適性、可靠性、方便性等四大潛在變數。模式推估結果顯示，外顯變數之票價、步行至站時間、等候時間、車內時間、共乘人數，以及潛在變數的方便性、可靠性、安全性等均對當地居民之運具選擇有顯著影響。基於此一推估結果，相關之公共運輸改善建議也一併提出。

Low patronage rate and improper service types of public transportation in rural area need to be improved urgently. Undoubtedly, without clearly knowing the mode choice behaviours and preferences of local residents, it is impossible to propose effective improvement strategies of public transportation. Meanwhile, traditional mode choice models overemphasize the effect of quantitative variables and ignore that of qualitative ones. Based on this, this study focuses on the mode choice behaviours of Sansia and Yingge districts, which can be considered as the relatively rural area of Taipei Metropolitan by incorporating commonly adopted quantitative variables as well as latent variables of safety, comfort, reliability and convenience, which are estimated by confirmative factor analysis. Estimation results show that the explanatory capability of the model can be enhanced by the introduction of latent variables. In addition, the effects of explicit variables of bus fare, walking distance, waiting time, in-vehicle time, and number of ride sharing persons are significantly tested. Corresponding improvement strategies for local public transportation are then proposed accordingly.