以潛在群體方法模化高速公路電子自動收費系統之選擇行為

The Choice Behaviors of Freeway Electronic Toll Collection : Latent Class Modeling Approach

高筑韻(Chu-Yun Kao)

指導教授 ： 邱裕鈞

國立交通大學/管理學院/交通運輸研究所/碩士(2011年)

https://doi.org/10.6842/NCTU.2011.00438

摘要

國道電子收費系統(Electronic toll collection, ETC)實施至今已逾五年，但是ETC使用率仍不盡理想。為清楚了解汽車駕駛人對使用電子自動收費系統（ETC）之偏好，本研究旨在透過問卷調查建構ETC個體選擇模式。由於每位汽車駕駛人對於ETC的屬性偏好極可能因人而異，為能反應不同群體駕駛人的選擇行為，但又避免事先主觀分群可能造成之偏誤。因此本研究發展出ETC之潛在群體模型，考量樣本異質性，透過分析結果找出潛在區隔特徵及各群影響變數。
基此，本研究以台灣地區22縣市做為ETC使用意願問卷的調查範圍。因應不同收費情境：計次及計程兩種收費機制亦分別調查及建模。計次收費以購機費用、通行費用、一年內滿通行次數退購機費及月租金做為影響變數；計程收費的影響變數為購機費用、通行費折扣、一年內滿通行里程退購機費及帳單作業費。結果顯示，兩模式均一致指出e通機購機費用及ETC通行費折扣是兩項最關鍵之影響因子。此外，各群體之偏好及選擇行為也確有相當差異。在計次收費情境，市場最佳分為五群：群1偏好裝設e通機且使用頻率最高；偏好全民體驗方案的群2，其特徵有相對高的所得及教育程度；偏好租用e通機的群3，年齡相對較低且使用頻率甚高；群4及群5偏好支付現金或通行票券，兩群的使用頻率、所得最低，但群4年齡最低，群5最高。計程收費情境為六群：群1偏好傳統e通機與計次群1特徵相似；群2偏好月底結算繳款，與計次群6特徵相似；偏好e通機外加GPS功能的群3，都有相對高的學歷、所得及頻率，年齡較低；群4及群5偏好全民體驗方案，與計次群2相似；群6則無偏好方案，年齡為最低。由此證實以潛在類別模式建模的必要性。
最後，透過彈性分析，針對不同群體分別研提有效的行銷策略。在計次收費情境中，以群2、群4及群5為行銷對象，並以付費已裝設e通機方案的購機費用及通行費用為行銷策略。計程收費則鎖定群2、群4及群6為目標市場，採GPS-e通機購機費用作為行銷策略。其結果顯示，計次收費情境下執行行銷策略明顯提升e通機的裝機率，但是計程收費卻效果有限。

關鍵字

電子收費系統 ； 個體選擇模式 ； 市場區隔 ； 潛在群體模式

並列摘要

The Electronic Toll Collection System has been used for five years, but the usage rate was still less than ideal to use ETC. To better understand the adoption behaviors of car drivers for Electronic Toll Collection (ETC) system, this study aims to develop an ETC disaggregate choice models based on questionnaire survey. To acknowledge the different preferences existed among car drivers, the latent class logit model is used to classify respondents into different groups without prior knowledge and subjective segmentation. The model consider heterogeneity, then analysis results to identify potential impact of segmentation characteristics and important variables.
To this end,the model based on a large-scale questionnaire survey on car owners of 22 cities/counties in Taiwan. Two toll collection scheme scenarios: per-entry based and distance-based, are surveyed. In per-entry based, there are four variable: purchase price of an e-pass, tolls, achieve times for toll station return purchase price and monthly rent. In distance-based toll collection system, there are also four varuable: purchase price of an e-pass, discounted amount of ETC toll, achieve distances return purchase price and operating fee. The estimation results show that five and six groups of car owners under per-entry based and distance-based toll collection system scenarios are optimally determined, respectively. Both models consistently show that purchase price of an e-pass (i.e. the on board unit of ETC) and discounted amount of ETC toll are identified as two most important factors affecting ETC adoption.
In addition, it is also found that significant differences do exist among different groups. Effective marketing strategies are then proposed for different groups under both toll collection scenarios accordingly. In per-entry based collection system, segment 2, segment 4 and segment 5 for the marketing object, and purchase price of an e-pass and tolls for OBU were as marketing strategy. It effectively improved usage rate. In distace-based collection system, focusing on segment 2, segment 4, segment 5 and segment 6, purchase price of a GPS-OBU was as marketing strategy. However, it was limited.

並列關鍵字

Electronic toll collection ； disaggregate choice model ； market segment ； latent class model

參考文獻

賴思豪，「以整合選擇及潛在變數方法模化高速公路電子自動收費系統之選擇行 為」，國立交通大學交通運輸研究所碩士論文，民國九十九年。

黃上原，「探討影響消費者使用國道電子收費系統意願之因素」，國立成功大學電信管理研究所碩士論文，民國九十五年。

Bhat, C. R., 1997, “An endogenous segmentation mode choice model with an application to intercity travel,” Transportation Science, Vol. 31, pp. 34-48.

Bhatnagar, A. and Ghose, S., 2004,“A latent class segmentation analysis of e-shoppers,” Journal of Business Research, Vol. 57, pp. 758-767.

Bijmolt, T. H. A., Paas, L. J., and Vermunt, J. K. ,2004, “Country and consumer segmentation: multi-level latent class analysis of financial product ownership,” International Journal of Research in Marketing, Vol. 21, pp.323-340.