[[alternative]]The application of Markov switching model to analyze the variation of value co-creation in service systems

碩士[[abstract]]服務可視為一個資源交換的過程，發生在客戶、服務人員及產品之間，透過服務供應商的價值主張，讓所有角色能參與服務系統中，經由資源交互利用，最終幫助顧客滿足日常所需。在2004年Vargo和Lusch兩名學者提出了服務主導邏輯(Service-dominant logic)，強調價值共同創造(Value co-creation)之重要性。價值共創主要的核心觀念為，服務供應商與顧客能運用資源，滿足雙方的需求。因價值的產生來自於服務系統中角色資源間的互動，而在互動的過程，會因不同角色運用資源而有所差異，若不當利用資源導致價值共同摧毀(Value co-destruction)。因此本研究透過馬可夫轉換模型分析出價值，並且能夠了解到價值的變異程度，針對過去歷史中的資料，推測價值共創及共毀狀態的轉移機率與持續時間，針對這些狀態轉移機率能夠讓服務供應商了解到服務系統產出的價值，如共毀持續時間過長，也能透過服務架構改善服務系統中資源的利用狀況。 本研究定義出來的價值變異衡量模型，能夠為所有服務供應商建置一套標準，了解服務系統中所有角色的關係，並結合Babin等人在1994年提出的實用性價值與享樂性價值的分類，實用性價值被認為是屬於理性的選擇，顧客會依照自己的需求，來衡量服務的效益，而享樂性價值是屬於ㄧ種顧客主觀的認知，顧客會依據曾經有過的服務體驗，來評斷服務流程。透過這兩種價值的分類方式，更能有效地評估服務系統中價值的產生。本研究蒐集了高雄市政府1999客服中心的開放資料，分別分析了應答率、一通電話解決比率及滿意度三種價值，從馬可夫轉換模型分析出的結果得知，應答率與原先客服中心的績效值服務水準做評估分析，利用馬可夫轉換模型更能呈現出狀態的週期變化;而至於第一通電話解決率的分析結果，在2013年及2016年第一季不穩定的狀況，可透過組織資源的改變，進行調整，以利價值產生; 最後則是滿意度的結果，原始數值並沒有穩定成長或是衰弱的狀況，所以原始資料的變動幅度有些微的變異，則馬可夫轉換模型分析出來的結果，狀態的轉變則會較大 本研究有以下三點研究貢獻(1)基於服務主導邏輯之資源觀點，運用價值共創與價值共毀的概念，定義出服務系統中價值的變異程度。(2)透過馬可夫轉換模型，以轉移機率的方式，呈現價值在服務系統中的變異狀況，並運用持續時間預測未來狀態的呈現。(3)價值變異是服務科學領域中重要的議題，可透過本研究所提出的價值變異衡量模型，探討不同服務系統下價值間相關性 本研究也有發現四點管理意涵(1)利用本研究所建構的價值變異衡量模型，能解決理論與實務間的隔閡，將模型應用於實務之服務系統中，更能表達出價值的創造流程，與評估價值變異的狀況。(2)服務供應商對於服務系統產生價值共毀狀態時，能以改善資源的方式，將服務系統以價值共創的狀態呈現。(3)客服中心能夠經由應答率，預測未來客服中心服務水準的績效。(4)客服中心應重視享樂性價值的變異，使得顧客在接觸服務時，有效的運用資源，才能讓顧客獲得愉悅的感受。 本研究有以下三點研究限制包含(1)價值的定義必須要長時間的資料當作依據，才能建立馬可夫轉換模型分析價值的變異狀況。(2)服務系統的價值指標，在不同的產業中，服務供應商會有不同的價值定義。(3)目前本研究只利用高雄市1999客服中心作為研究資料，所以開放資料的蒐集是從2012年開始至今，並無法得知2012年以前的狀況。最後，未來學者可以延伸本研究之方向有以下兩點 (1)後續有不同的學者針對不同的議題有改進馬可夫轉換模型的應用，未來可以利用不同類型或是三種狀態以上的馬可夫轉模型進行操作分析。(2) 建議探討服務供應商與顧客透過不同媒介交流的服務系統，透過不同媒介的服務系統，所分析出的價值共創與共毀狀態，會有不同的結果，當然資源的運用模式也可能會有所差異。 關鍵詞:價值共創、價值共毀、價值變異、服務主導邏輯、馬可夫轉換模型、客服中心[[abstract]]Service can be regarded as a process of resource exchange, between customers, waiter and products. Through the service provider''s value proposition, than roles can successfully enter service systems. According to their resources, the main purpose is the everyday needs of service providers to help customers achieve. In 2004, Vargo and Lusch proposed service-dominant logic, the services is the main core, and value co-creation is the importance thing of service-dominant logic. Value co-creation major implications , expect service providers and customers can use resources to meet the needs of both. Value generated from the interaction between the role of resource in service system. In the interaction process, because of different users use resources some differences. As a result, incorrectly use of resources, than value can be co-destruction. Therefore, this study through the Markov switching model to analyze the value , and we can understand the degree of variation in the value. For the past historical data, we guess co-create value and co-destruction of the state of the transfer probability and duration. For these state transition probability enables service providers to understand the value of the output of the service system. Such as the co-destruction of the state long duration, we can improve the utilization of system resources and services through the service structure. This study defines the value of the measurement model variation, we can build a set of standards, to understanding the relationship between service systems in all roles, and the role will use their resources, after interaction to co-create value. In the final analysis is Markov switching model of the state transition probability values, and enables service providers to understand the value of the duration of each state to improve the allocation of resources, and ultimately produce a stable value in the service system. This study collected open data of the call center of Kaohsiung city government from January/2012 to April/2016. This study also defines critical values of the call center including answer rates, first call resolution, and customer satisfaction (hedonic value). According to Markov switching model to analyze the results of that, answer rate and the performance of the original value of the service level, the customer service center for evaluation analysis, by using Markov switching model showing better state cycle; As for the first call resolution in 2013 and the first quarter of 2016, the unstable situation can be changed through the organizational resources, be adjusted to facilitate value generation; Finally, there is the result of satisfaction, the original value and no stable growth or debilitating condition, so that the fluctuation range of raw materials there is a slight variation, the Markov switching model analysis out of the results, it will change the status of the larger. This study proposes several research contributions as follows 1) This study contributed to define the levels of value variations in service systems by considering value co-creation and co-destruction in terms of the viewpoint of service dominant logic. 2) Value variations can be numerically represented as transition probability to predict the future trend by adopting Markov switching model.3) Researchers can investigate relations among values within different service systems based on the proposed value variation measurement model. There are four managerial implications including 1)The proposed value variation measurement model can be applied into practice that is to decrease the gap between academic theories and practical matters. 2) Service providers can improve the utility of resources in order to create values while they encounter value co-destruction. 3) Call center can accurately predict the performance of service level by using the answer rate.4) Call center have to pay attention to hedonic value variations to satisfy customers by effectively allocating resources. There are three research limitations: 1) it is necessary to collect long-term data to apply into Markov switching model for evaluating the value variations of service systems. 2) There are different values within different service systems and practical fields. 3)The open data of the call center of Kaohsiung city government is only from January/2012. Finally, there are several future directions for researcher to further study: 1) researchers can adopt Markov switching model to other research fields and topics and use different types of Markov switching model to evaluate the value variations. 2) The proposed models and approaches can be applied to different service systems to analyze the levels of value co-creation and co-destruction. Keywords: value co-creation, value co-destruction, value variation, service-dominant logic, Markov switching model, call center[[tableofcontents]]謝辭I 摘要II 目錄VII 表目錄IX 圖目錄X 第壹章 緒論1 1.1研究背景1 1.2研究動機4 1.3研究問題6 1.4研究目的7 1.5研究流程9 第貳章 文獻探討10 2.1服務科學10 2.1.1何謂服務11 2.1.2服務科學之定義12 2.1.3服務系統16 2.1.4小結17 2.2價值共創17 2.2.1服務主導邏輯 18 2.2.2何謂價值20 2.2.3價值共創24 2.2.4價值共毀28 2.2.5小結29 2.3馬可夫轉換模型 30 2.3.1小結34 第參章研究方法36 3.1研究模型36 3.2客服中心與實用性價值及享樂性價值 39 3.2.1應答率40 3.2.2第一通電話解決率41 3.2.3滿意度43 3.3單根檢定44 3.4最適落後期選擇方法47 3.5馬可夫轉換模型 47 3.6小結50 第肆章 模型推導與分析52 4.1資料介紹52 4.2應答率指標52 4.2.1單根檢定結果 53 4.2.2最適落後期數 54 4.2.3馬可夫轉換模型55 4.3第一通電話解決率57 4.3.1單根檢定結果58 4.3.2最適落後期數 59 4.2.3馬可夫轉換模型60 4.4滿意度62 4.4.1單根檢定結果 63 4.4.2最適落後期數64 4.4.3 馬可夫轉換模型64 4.5研究討論67 4.5.1 應答率67 4.5.2 第一通電話解決率70 4.5.3 滿意度71 4.5.4 討論72 4.6 小結 74 第伍章 結論76 5.1研究結論76 5.2 貢獻 78 5.2.1研究貢獻78 5.2.2管理意涵80 5.3研究限制82 5.4未來研究方向83 參考文獻 84 表目錄 =============================================== 表2-1服務科學定義 15 表2-2產品主導邏輯(G-D Logic)與服務主導邏輯(S-D Logic)主要差異 19 表4-1應答率之單根檢定 54 表4-2應答率之最適落後期 55 表4-3馬可夫轉換模型之應答率參數值 57 表4-4第一通電話解決率之單根檢定表 59 表4-5第一通電話解決率之最適落後期 59 表4-6馬可夫轉換模型之第一通電話解決率參數值 62 表4-7滿意度之單根檢定表 64 表4-8滿意度之最適落後期選取 64 表4-9參數的估計值(滿意度) 67 表4-10 服務水準與馬可夫轉換模型應答率狀態轉移比較表 69 圖目錄 ========================================= 圖3-1價值變異衡量模型 38 圖3-2實用性價值(有用) 40 圖3-3實用性價值(有效) 42 圖3-4享樂性價值(愉悅) 43 圖4-1應答率圖(2012:1~2016:4) 53 圖4-2應答率之持續時間(月) 56 圖4-3應答率之狀態轉移機率圖 56 圖4-4第一通電話解決率圖(2012:1~2016:4) 58 圖4-5第一通電話解決率之持續時間(月) 60 圖4-6第一通電話解決率狀態之轉移機率圖 61 圖4-7滿意度圖(2012:1~2016:4) 63 圖4-8滿意度之持續時間(月) 65 圖4-9滿意度之狀態轉移機率圖 66 圖4-10服務水準圖(2012:1~2016:4) 69 圖4-11 2013年2月電話外撥數 71[[note]]學號: 602630526, 學年度: 10