以動態網絡分析探討網路社群口碑與購買態度之關係：以批踢踢nb_shopping板為例

Abstract

近來網路口碑已成為一種重要的行銷手法，許多研究結果顯示，越來越多消費者在進行購買決策時，會選擇參考網路口碑的資訊，並進一步影響購買行為與態度。然而，網路口碑與購買態度之間的關係，因果未明，究竟是網路口碑影響購買態度，或是購買態度影響網路口碑，其效果難以區辨。 本研究透過社會網絡分析的觀點，將網路口碑與購買態度視為互動網絡結構與網絡中的個體行為，並將互動網絡與個體行為視為一個不斷變動的動態過程，經由歷時性的觀察，檢視互動網絡結構與購買態度的動態變化，以區辨兩者間的因果關係。在方法上採用結構—行為共變模型，同時考慮網絡結構對個體行為所造成的影響效果，以及個體行為對網絡結構所造成的選擇效果，在兩者相互影響的情況下，估計各自的效果量。 研究結果顯示，內度數與聲望效果影響力顯著，網絡成員傾向建立向內關係，並與內度數高的成員進行互動。購買態度在影響與選擇上都具有顯著效果，網絡成員傾向與購買態度相近的成員進行互動建立關係，也會受到互動成員中態度相近者的影響而強化購買態度。網絡結構與個體行為均具有其效果，並相互影響。

Indicated by research these days, “online word-of-mouth” has become one of the most important strategies in marketing. More and more consumers choose to refer to online word-of-mouth before they make any decisions, which in turn affect their buying intention and even lead to behavior. However, the causal relationship between online word-of-mouth and buying intention is still unclear. Under the social network analysis paradigm, this research treats online word-of-mouth as the structure of dynamic interacting network and consumer attitude as the individual behavior within this network. Furthermore, this research believes that the interaction between the network and individual behavior is a dynamic change, and examines it with diachronic analysis. Research results reveal that the influence of in-degree effect and popularity effect is significant, indicating that members of network tend to establish inward relationship and interact with members with high in-degrees. Buying intention has significant effect on selection and influence, indicating that members of network tend to interact and establish relationship with the ones with similar consumer attitude. In a word, both the structure of network and individual behavior affects each other.