國道客運旅客選擇行為之研究

Abstract

為藉由相互競爭以提升國道客運整體的服務品質，政府在開放國道路線的經營權後，因經營業者眾多，使得服務水準較過去而言有長足改善。各家客運皆以軟、硬體設施的提升，或離峰時段減價策略來吸引旅客搭乘。各式的行銷手法，無不希望能鞏固業者本身的客源，進而擴大市場占有率。因此，瞭解旅客選擇國道客運公司的行為，以評估行銷策略的成效，乃是客運業者重視的課題。 本研究以問卷方式蒐集搭乘國道客運往返台北-台中與台北-高雄的旅客選擇行為。在顯示性偏好資料部分，要求旅客針對當次及過去曾使用過的客運公司以李克特五點尺度填答服務品質的實際感受。服務品質的問項主要參考SERVQUAL量表及運輸業量表所產生共20題的衡量指標。在敘述性偏好資料方面，本研究運用實驗設計法形成假設情境。各情境中的屬性及其水準值皆以隨機方式產生，探討旅客對服務品質屬性間的權衡關係。鑑於服務品質指標間具有高度的相關性，本研究利用探索性因素分析法由上述的20項指標中萃取出共同因素（構面），並以此建構驗證性因素模式來檢驗資料與理論模式間的適合度。依據最佳之驗證性因素模式計算構面分數，以其作為旅客選擇模式之解釋變數。校估出最佳的顯示性與敘述性偏好模式後，本研究進一步結合兩種資料來構建整合模式，期能有助於模式預測能力的提升。 探索性之服務品質因素模式共萃取出客運公司營運方式、車內硬體設備及駕駛與服務員態度等3個構面。驗證性因素模式的整體配適度、收斂效度與判別效度均相當理想。選擇模式的校估結果顯示，票價、班次、個人所得及服務品質構面為旅客選擇客運公司時的考量要素。整合模式的解釋能力介於顯示性與敘述性模式之間，且尺度參數十分顯著，說明敘述性資料變異程度較大的特點。依據整合模式的校估結果，本研究計算服務品質的價值，即旅客對各項服務品質屬性願意支付的價格。最後建議客運業者在考量改善成本後，可依據旅客較重視的服務品質項目（願意支付價格的高低）來改善目前的服務水準。

The deregulation of freeway bus routes by the government in Taiwan has increased the competition among bus operators. Various operating strategies have been applied and customer service been enhanced by bus operators in order to increase market shares, keep present customers and attract new ones. Some feature in lower ticket price while others provide high quality services with higher charges. This study attempts to understand passengers’ choice behavior for freeway bus operators and identify important attributes affecting the choice. This research designed a revealed preference (RP) and stated preference (SP) survey to interview passengers traveling between Taipei-Kaohsiung and Taipei-Taichung corridors. Twenty items are identified to measure bus service quality. The modeling framework consists of two stages. The first stage employs the exploratory factor analysis (EFA) to extract the underlying common factors and, based on the result of the EFA, the confirmatory factor analysis (CFA) assesses the fitness between model and data. The second stage incorporates underlying latent attributes to estimate discrete choice models with RP and SP data. The discrete choice models include the multinomial logit, the nested logit, and the mixed logit models. The empirical analysis identifies three constructs including operational performance, onboard amenities, and courtesy of personnel. The estimation results of the RP and SP models indicate that fares, frequency, personal income, and service quality constructs are significant factors to influence the choice of bus operators. The coefficients of observed variables, service quality constructs, and fares are used to calculate willingness-to-pay for each item. The willingness-to-pay represents tradeoffs among fare and service quality and provides valuable insights on service design and marketing strategies.