MINI愛好者，MAXI愛好者：品牌社群成員網路口碑意向前因調查

Fused with the multi-media-based, interactive nature of virtual communities, brand communities in cyberspace should serve as an ideal platform for positive eWOM for brands. However, little research has investigated online brand communities as a focal point of brand related eWOM, and our understanding of the way such eWOM functions in the online social context is lacking. The purpose of the present study, therefore, is to develop a conceptual framework for understanding predictors of eWOM among online brand community members. Divergent from the top-down model of communication with a few highly influentials triggering WOM, interactions in the context of online communities are dynamic and bidirectional as the members take on multiple roles in the process of eWOM (Allsop, Bassett, and Hoskins 2007). They are information givers, obtainers and spreaders. They provide their own opinions and information to others, attain others’ opinions and information and pass those along to other non-members.
網路空間的品牌社區與虛擬社區基於多媒體的互動性質相融合，應該成為品牌積極網路口碑的理想平臺。然而，很少有研究將在線品牌社區作為品牌相關網路口碑的焦點進行調查，而且我們對此類網路口碑在在線社交環境中的運作方式缺乏瞭解。因此，本研究的目的是開發一個概念框架，用於理解在線品牌社區成員中 eWOM 的預測因素。與自上而下的溝通模式不同，少數極具影響力的人觸發口碑，在線社區背景下的互動是動態和雙向的，因為成員在網路口碑的過程中扮演著多種角色（Allsop、Bassett 和 Hoskins 2007）。他們是信息傳遞者、獲取者和傳播者。他們向他人提供自己的意見和信息，獲得他人的意見和資訊，並將其傳遞給其他非會員。

Building upon the logic, the conceptual model presented and empirically tested in this study identifies brand identification, community involvement and trust as key predictors of eWOM intention among members of online brand communities. More specifically, an online brand community member’s identification with the brand and the community as well as trust in fellow members of the community will influence the member’s intention to engage in eWOM activities, both within and outside of the community. Results of this study, added to the growing literature on eWOM, advance our theoretical knowledge of how social groups such as brand communities in cyberspace operate in the eWOM process. Findings of this study also offer strategic insight into the role of online brand
基於這一邏輯，本研究中提出並經過實證檢驗的概念模型將品牌識別、社區參與和信任確定為在線品牌社區成員網路口碑意圖的關鍵預測因素。更具體地說，在線品牌社區成員對品牌和社區的認同以及對社區其他成員的信任，將影響該成員在社區內外參與網路口碑活動的意圖。這項研究的結果，加上關於網路口碑的越來越多的文獻，推進了我們對網路空間品牌社區等社會群體如何在網路口碑過程中運作的理論認識。這項研究的結果還為在線品牌的作用提供了戰略見解
communities in engendering and facilitating eWOM in favor of brands and aid marketers in designing effective viral marketing campaigns supported by brand devotees.
社區生成和促進有利於品牌的 eWOM，並説明行銷人員設計由品牌擁護者支援的有效病毒式營銷活動。

Inspired by literature on eWOM, virtual communities, brand communities and trust, our theoretical work is to identify key antecedents to eWOM activities in the context of online brand communities and to explicate the flow of influence among the variables of interest. In the conceptual model, not only online brand community members’ intention to exchange information with peer community members but also their likelihood of passing information to outsiders is examined. Underlying the conceptualization is the notion that online communities as a platform for social interactions enable truly interactive eWOM where a single person can act as both an information or opinion provider and a seeker at the same time, eWOM among members of a brand community may not stay within the community as they often act as brand evangelists. Brand enthusiasts participating in online brand communities play multiple roles in disseminating positive eWOM for the brand while the degree of engagement in the eWOM process is largely dependent upon their identification with the brand and the community, and trust in fellow community members. The conceptual model is presented in Figure 1.
受關於網路口碑、虛擬社區、品牌社區和信任的文獻的啟發，我們的理論工作是在在線品牌社區的背景下確定網路口碑活動的關鍵前因，並闡明感興趣的變數之間的影響流動。在概念模型中，不僅檢查了在線品牌社區成員與同行社區成員交換資訊的意圖，還檢查了他們將資訊傳遞給外部人員的可能性。概念化的基礎是這樣一個概念，即在線社區作為社交互動的平臺，實現了真正的互動式網路口碑，一個人可以同時充當資訊或意見提供者和尋求者，品牌社區成員中的網路口碑可能不會停留在社區內，因為他們經常充當品牌傳播者。參與在線品牌社區的品牌愛好者在為品牌傳播積極的網路口碑方面發揮著多重作用，而參與網路口碑過程的程度在很大程度上取決於他們對品牌和社區的認同，以及對社區成員的信任。概念模型如圖 1 所示。

While the importance of eWOM has been highlighted in computer-mediated communications in general, research on product-related eWOM behaviors among online community members and subsequent consequences is still in its burgeoning stage. Previous studies have attempted to discover the motivations behind consumers’ information sharing online (Hennig-Thurau el al. 2004; Goldsmith and Horowitz 2006), to identify the characteristics of influentials and their online networks (Smith, Coyle, and
雖然網路口碑在計算機介導的通信中的重要性已經得到強調，但對在線社區成員中與產品相關的網路口碑行為和後續後果的研究仍處於蓬勃發展的階段。以前的研究試圖發現消費者在線資訊共用背後的動機（Hennig-Thurau el al. 2004;Goldsmith 和 Horowitz 2006 年），以確定有影響力的人及其在線網路的特徵（Smith、Coyle 和

Figure 1. Proposed model of eWOM among online brand community members.
圖 1.在線品牌社區成員中提出的 eWOM 模型。

Lightfool 2007) and to assess the impact of eWOM on online community members’ attitudes and behaviors (Bickart and Schindler 2001; Hung and Li 2007). Taken together, these studies have proven online communities in general as an effective platform for generating eWOM.
Lightfool 2007 年），並評估網路口碑對在線社區成員態度和行為的影響（Bickart 和 Schindler 2001 年;Hung 和 Li 2007）。綜上所述，這些研究已經證明，在線社區總體上是產生網路口碑的有效平臺。

What is still lacking is our knowledge of brand-focused eWOM in the context of online brand communities. Online brand communities warrant special attention and careful examination in that they not only enjoy the constructive features of online communities for eWOM but also represent aggregations of committed and passionate consumers of certain brands who often serve as active supporters and advocates for the brands. Members of online brand communities may play a significant role in engendering and facilitating brand-related eWOM.
我們仍然缺乏的是我們在在線品牌社區背景下以品牌為中心的網路口碑知識。在線品牌社區值得特別關注和仔細檢查，因為它們不僅享有網路口碑在線社區的建設性特徵，而且還代表了某些品牌的忠誠和熱情的消費者的集合，他們通常是品牌的積極支援者和宣導者。在線品牌社區的成員可能在產生和促進與品牌相關的網路口碑方面發揮重要作用。

In their pioneering study, Muniz and O’Guim (2001) showed that consumers gather in both face-to-face and computer-mediated environments and form communities through which they are actively involved in the creation and enhancement of meanings of brands as shared objects. In addition to the development of relationship with brands, brand communities allow consumers to form interpersonal ties with fellow community members by sharing knowledge, opinions and experience pertaining to the brands. Brand communities also bring social values to brand consumptions. Brand communities enable and encourage social exchanges with a focus on brands and aid their members in constructing and enhancing their identity through such social engagements. Strong brand communities can lead to positive brand-related outcomes such as ‘socially embedded and entrenched loyalty, brand commitment, and even hyper-loyalty’ (Muniz and O’Guinn 2001, 427).
在他們的開創性研究中，Muniz 和 O'Guim （2001） 表明，消費者聚集在面對面和計算機介導的環境中，並形成社區，通過這些社區，他們積极參與創造和增強品牌作為共用對象的意義。除了發展與品牌的關係外，品牌社區還允許消費者通過分享與品牌相關的知識、意見和經驗，與其他社區成員建立人際關係。品牌社區也為品牌消費帶來社會價值。品牌社區促進和鼓勵以品牌為中心的社會交流，並幫助其成員通過此類社交參與構建和增強他們的身份。強大的品牌社區可以帶來與品牌相關的積極結果，例如「社會嵌入和根深蒂固的忠誠度、品牌承諾，甚至超級忠誠度」（Muniz 和 O'Guinn 2001,427）。

The social dimension of brand-driven interactions marks brand communities, and studies have accordingly focused on the role of social factors in consumers’ engagement in brand communities and have identified two major antecedents to consumers’ intention to continue participating in brand communities, namely brand identification and community identification (Bagozzi and Dholakia 2002, 2006; Algesheimer et al. 2005; Peters and Hollenbeck 2005). Brand identification is the extent to which a consumer sees his or her own self-image as overlapping with the brand’s image, and consumers use the brand as a reference point to distinguish themselves from non-users of the brand (Escalas and Bettman 2005). Similarly, community identification involves a member’s cognitive recognition of shared characteristics of, affective commitment 10 and evaluative importance of the community in reference to himself or herself (Ellemers, Kortekaas and Ouwerkerk 1999; Bagozzi and Dholakia 2006). As social identity theory suggests, people are attracted to and identify with a social group that represents similar qualities and images to theirs and, consequently, define themselves in terms of the characteristics of typical members of the group (Tajfel and Turner 1986).
品牌驅動互動的社會維度標誌著品牌社區，因此研究關注了社會因素在消費者參與品牌社區中的作用，並確定了消費者打算繼續參與品牌社區的兩個主要前因，即品牌識別和社區識別（Bagozzi 和 Dholakia 2002,2006;Algesheimer 等人，2005 年;Peters 和 Hollenbeck，2005 年）。品牌識別是消費者將自己的自我形象與品牌形象重疊的程度，消費者將品牌作為參考點，將自己與品牌的非使用者區分開來（Escalas 和 Bettman 2005）。同樣，社區認同涉及成員對社區的共同特徵、情感承諾 10 （affective commitment） 和社區對自身重要性的認知識別（Ellemers， Kortekaas and Ouwerkerk 1999;Bagozzi 和 Dholakia 2006 年）。正如社會身份理論所表明的那樣，人們被一個代表與他們相似的品質和形象的社會群體所吸引並認同，因此，根據該群體典型成員的特徵來定義自己（Tajfel 和 Turner 1986）。

In a brand community context, a brand is the center of the consumer identification process and drives the subsequent group membership or identification among its community members. Along this logic, a brand represents a higher level of social categorization, and communities of the brand represent various subgroups. For example, a BMW driver values the brand and, therefore, is willing to exchange personal knowledge and experience with other like-minded consumers in a community. Among many communities available, the consumer may undergo a subsequent self-categorization process and select one or more particular brand communities with which he or she
在品牌社區環境中，品牌是消費者識別過程的中心，並推動其社區成員之間的後續群體成員資格或識別。按照這個邏輯，品牌代表更高級別的社會分類，而品牌的社區代表各種子群體。例如，BMW 司機重視品牌，因此願意與社區中其他志同道合的消費者交流個人知識和經驗。在眾多可用的社區中，消費者可能會經歷後續的自我分類過程，並選擇他或她與之相關的一個或多個特定品牌社區
identifies with. In fact, prior studies indicated that satisfied consumers who enjoy a strong relationship with a brand are more likely to make extra efforts to seek out other consumers to share their passion for the brand (Algesheimer et al. 2005; Peters and Hollenbeck 2005). Therefore, in a brand community setting, a member’s identification with the brand precedes and contributes to his or her identification with the affiliated brand community. The following hypothesis examines the relationship between brand identification and community identification in the context of online brand communities.
認同。事實上，先前的研究表明，與品牌建立牢固關係的滿意消費者更有可能付出額外的努力來尋找其他消費者來分享他們對品牌的熱情（Algesheimer 等人，2005 年;Peters 和 Hollenbeck，2005 年）。因此，在品牌社區環境中，成員對品牌的認同先於並有助於他或她對附屬品牌社區的認同。以下假設考察了在線品牌社區背景下品牌識別和社區識別之間的關係。

H1: Online brand community members’ identification with the brand is positively related to their identification with the brand community.
H1：在線品牌社區成員對品牌的認同與他們對品牌社區的認同呈正相關。
When a consumer recognizes a brand as possessing a high degree of similarity to the consumer’s self-image, the consumer exhibits strong identification with the brand which leads to his or her loyalty to the brand (Fournier 1998). Indeed, Algesheimer et al. (2005) found a consumer’s relationship with a brand as an influential antecedent to his or her brand loyalty. That is, the more a consumer values the self-brand connection, the more he or she would commit to the brand and continue purchasing and using it. Hence, the following hypothesis is stated to examine the predicted relationship between brand identification and brand loyalty among online brand community members.
當消費者認識到一個品牌與消費者的自我形象具有高度相似性時，消費者就會表現出對該品牌的強烈認同感，從而導致他或她對品牌的忠誠度（Fournier 1998）。事實上，Algesheimer 等人（2005 年）發現消費者與品牌的關係是他或她品牌忠誠度的一個有影響力的前提因素。也就是說，消費者越重視自我品牌的聯繫，他或她就越會致力於該品牌並繼續購買和使用它。因此，提出了以下假設來檢驗在線品牌社區成員的品牌識別和品牌忠誠度之間的預測關係。

H2: Online brand community members’ brand identification is positively related to their brand loyalty.
H2：在線品牌社區成員的品牌識別度與他們的品牌忠誠度呈正相關。

A number of studies have suggested that consumers with strong loyalty to a brand tend to create and spread information in favor of the brand (Dick and Basu 1994; Gounaris and Stathakopoulos 2004). As brand communities are comprised of fans of a brand who are loyal toward the brand, it is highly likely that members of brand communities actively engage in positive eWOM with regard to the brand. Especially when the brand is relevant to a consumer’s self-concept, the consumer is more likely to connect with the brand and provide favorable eWOM about it (Chung and Darke 2006; Allsop et al. 2007). While generally satisfied with a brand, consumers might not simply supply positive eWOM in support of the brand since such behavior requires considerable effort (Chung and Darke 2006). Highly loyal and motivated consumers such as brand community members, however, are more likely to take the effort to engage in eWOM for their favorite brand (Chung and Darke 2006). Further, the online setting with a readily available social network and various communication tools facilitates their participation in positive eWOM for the brand.
許多研究表明，對品牌有強烈忠誠度的消費者傾向於創造和傳播有利於該品牌的資訊（Dick 和 Basu 1994;Gounaris 和 Stathakopoulos 2004 年）。由於品牌社區由忠於該品牌的品牌粉絲組成，因此品牌社區的成員很可能會積極地參與與該品牌相關的積極網路口碑。特別是當品牌與消費者的自我概念相關時，消費者更有可能與品牌建立聯繫並提供關於它的有利的 eWOM（Chung 和 Darke 2006;Allsop 等人，2007 年）。雖然消費者通常對品牌感到滿意，但消費者可能不會簡單地提供積極的網路口碑來支援該品牌，因為這種行為需要付出相當大的努力（Chung 和 Darke 2006）。然而，高度忠誠和積極性的消費者，如品牌社區成員，更有可能努力為他們最喜歡的品牌參與網路口碑（Chung 和 Darke 2006）。此外，在線環境、現成的社交網路和各種通信工具有助於他們參與品牌的正面網路口碑。

Following this logic, there is good reason to believe that members of an online brand community enthusiastically engage in eWOM in multiple ways. Noteworthy here is that, unlike traditional WOM research that assumes customary functions of individuals such as opinion leaders and opinion seekers, we conceptualize that a brand community member can take on multiple roles in eWOM. Within a social network, an individual can give information to others, obtain advice from others and pass it along to others (Allsop et al. 2007). Similarly, as social exchange is central to brand communities consisting of consumers with common interests and topics, members of brand communities could serve as givers, seekers and disseminators of information and opinions all together. Since consumers who love a brand tend to share their positive brand experience with others and spread the good word for the brand (Carroll and Ahuvia 2006), eWOM generated by
按照這個邏輯，我們有充分的理由相信在線品牌社區的成員以多種方式熱情地參與網路口碑。值得注意的是，與傳統的口碑研究不同，傳統的口碑研究承擔了個人（如意見領袖和意見尋求者）的慣常職能，我們概念化了品牌社區成員可以在網路口碑中扮演多種角色。在社交網路中，個人可以向他人提供資訊，從他人那裡獲得建議並將其傳遞給他人（Allsop 等人，2007 年）。同樣，由於社會交換是由具有共同興趣和話題的消費者組成的品牌社區的核心，因此品牌社區的成員可以共同充當信息和意見的提供者、尋求者和傳播者。由於熱愛品牌的消費者傾向於與他人分享他們積極的品牌體驗，並傳播該品牌的好話（Carroll 和 Ahuvia 2006），因此由
brand community members may not be limited to within the community. These brand enthusiasts are expected to pass the information they acquire from their peer community members on to others who do not belong to the community as well. Thus, the following set of hypotheses is formulated to investigate the relationship between brand loyalty and eWOM intention of online brand community members - information giving, seeking and passing.
品牌社區成員可能不限於社區內。這些品牌愛好者應該將他們從同行社區成員那裡獲得的信息傳遞給不屬於社區的其他人。因此，制定了以下一組假設來研究品牌忠誠度與在線品牌社區成員的 eWOM 意圖之間的關係 - 信息給予、尋求和傳遞。

H3: Online brand community members’ brand loyalty is positively related to their eWOM intentions (a) to give and (b) to obtain information within the brand community as well as © to pass along information to others outside of the community.
H3：在線品牌社區成員的品牌忠誠度與他們的網路口碑意圖呈正相關 （a） 給予和 （b） 在品牌社區內獲取資訊，以及©將資訊傳遞給社區外的其他人。

In addition to the brand-related factors, the social aspect of brand communities should also influence the extent to which their members engage in eWOM. Previous studies suggest that community identification leads to positive behavioral outcomes toward the community. For instance, members of an online community post comments to maintain their status in the community, to foster their relationships with other community members and to receive social approval from the community (Hennig-Thurau et al. 2004). However, due to the anonymous, multi-directional nature of the mediated communications in the online environment, consumers’ engagement in eWOM in online brand communities may rely on their trust in other members. Although trust is important to any type of social group, it may play a more essential role in communications within an online-based community (Greenberg, Greenberg, and Antonucci 2007). In a virtual community environment wherein face to face contact or physical evidence is unavailable to provide a sufficient guarantee that other community members will obey the community rules, trust perception acts as a cue for community members in determining confidence to continue their interactions with one another (Jarvenpaa, Knoll, and Leidner 1998). Trust in virtual groups or communities has been found to significantly influence the way information sharing and communication operates among their members (Mayer, Davis, and Schoorman 1995; Jarvenpaa et al. 1998; Jarvenpaa and Leidner 1999; Ridings, Gefen, and Arinze 2002).
除了與品牌相關的因素外，品牌社區的社交方面也應該影響其成員參與網路口碑的程度。以前的研究表明，社區認同會導致對社區的積極行為結果。例如，在線社區的成員發表評論是為了保持他們在社區中的地位，促進他們與其他社區成員的關係，並獲得社區的社會認可（Hennig-Thurau 等人，2004 年）。然而，由於在線環境中仲介通信的匿名性、多向性，消費者在在線品牌社區中參與網路口碑可能依賴於他們對其他成員的信任。儘管信任對任何類型的社會群體都很重要，但它可能在基於在線的社區內的交流中發揮更重要的作用（Greenberg、Greenberg 和 Antonucci 2007）。在虛擬社區環境中，面對面的接觸或物證無法提供足夠的保證，證明其他社區成員會遵守社區規則，信任感知可以作為社區成員確定繼續彼此互動的信心的線索（Jarvenpaa、Knoll 和 Leidner 1998）。已發現對虛擬團體或社區的信任對其成員之間信息共用和通信的運作方式產生重大影響（Mayer、Davis 和 Schoorman 1995;Jarvenpaa 等人，1998 年;Jarvenpaa 和 Leidner 1999;Ridings、Gefen 和 Arinze 2002）。

Trust is generally conceptualized as being comprised of cognitive and affective components (Lewis and Weigert 1985; McAllister 1995). Cognition-based trust is derived from the calculative and rational characteristics such as reliability, competence and responsibility, demonstrated by those who are trusted, namely, trustees. On the other hand, affect-based trust involves the emotional elements and social skills of the trustees. Care and concern for the welfare of others is an example of affect-based trust. Both types of trust are evidenced in online community settings and found to facilitate information sharing among community members (Ridings et al. 2002).
信任通常被概念化為由認知和情感成分組成（Lewis 和 Weigert 1985;McAllister 1995 年）。基於認知的信任源自於被信任的人（即受託人）所表現出的計算和理性特徵，例如可靠性、能力和責任感。另一方面，基於情感的信任涉及受託人的情感元素和社交技能。關心和關心他人的福祉是基於情感的信任的一個例子。這兩種類型的信任在在線社區環境中都有所證明，並被發現有助於社區成員之間的信息共用（Ridings 等人，2002 年）。

Trust formation in a group setting is initiated by a self-categorization process in which individuals belonging to a social group form their presumptive trust in other group members simply on the basis of their shared membership (Kramer 1999). In the absence of individuating cues in a virtual environment, individuals often turn to limited symbolic cues to form stereotypical impressions of others (Lea and Spears 1992; Walther 1997; Postmes, Spears, and Lea 1998). In an online brand community, the brand serves as a symbolic object that bonds individual community members with a shared identity and fosters the community members’ feeling of belonging. Because
在群體環境中，信任的形成是由一個自我分類過程開始的，在這個過程中，屬於一個社會群體的個人僅僅根據他們的共同成員身份就形成了他們對其他群體成員的推定信任（Kramer 1999）。在虛擬環境中缺乏個人化線索的情況下，個體通常會求助於有限的象徵線索來形成對他人的刻板印象（Lea 和 Spears 1992;Walther 1997 年;Postmes、Spears 和 Lea 1998）。在線品牌社區中，品牌作為一個象徵性物件，將社區成員與共同的身份聯繫在一起，並培養社區成員的歸屬感。因為
individuals tend to favor and idealize their in-group members, the more an individual member identifies with the online brand community, the greater his or her trust in fellow members of the community should be (Brewer 1996). Following this logic, the next hypothesis is put forth to examine the relationship between online brand community members’ identification with the community and the two types of trust they have in peer members of the community.
個人傾向於支援和理想化他們的內部群體成員，個人成員對在線品牌社區的認同度越高，他或她對社區其他成員的信任就越大（Brewer 1996）。按照這個邏輯，提出了下一個假設，以檢查在線品牌社區成員對社區的認同與他們對社區同伴成員的兩種信任之間的關係。

H4: Online brand community members’ community identification is positively related to their (a) cognitive-based community trust and (b) affect-based community trust.
H4：在線品牌社區成員的社區認同與他們 （a） 基於認知的社區信任和 （b） 基於情感的社區信任呈正相關。

Trust is also an important factor in the eWOM process. A study of members of 36 online communities of various interests found that both cognition- and affect-based trust were significant predictors of the community members’ desire to exchange information with one another (Ridings et al. 2002). Cognition-based trust in fellow community members increased the value of information originated by them due to the perceived competence and expertise, which in turn encouraged information seeking from them. Community members were also more motivated to provide their opinions and thoughts when they believed feedback from other knowledgeable members was valuable. Likewise, affect-based trust in other community members enhanced information exchanges within the community because the value of information obtained from the members was augmented when the information providers were viewed as honest and sincere. Equally, such trust increased the members’ willingness to offer information as they did not need to be concented about oihers’ opportunistic behaviors.
信任也是網路口碑流程中的一個重要因素。對 36 個不同興趣的在線社區成員的研究發現，基於認知和情感的信任都是社區成員相互交換資訊願望的重要預測因素（Ridings 等人，2002 年）。由於感知的能力和專業知識，對社區成員的基於認知的信任增加了他們所產生信息的價值，這反過來又鼓勵了他們尋求資訊。當社區成員認為來自其他知識淵博的成員的反饋很有價值時，他們也更有動力提供自己的意見和想法。同樣，對其他社區成員的基於情感的信任增強了社區內的資訊交流，因為當資訊提供者被視為誠實和真誠時，從成員那裡獲得的信息的價值就會增加。同樣，這種信任增加了成員提供資訊的意願，因為他們不需要專注於其他人的機會主義行為。

Past research suggests that people are more likely to disseminate information when they perceive the source of information to be trustworthy than when they receive information from less credible sources (Phelps et al. 2004). That is, when a source providing information or opinion is seen as possessing expertise and honesty, the message from the source tends to be of perceived value and passed along to others (Allsop et al. 2007; Brown, Broderick, and Lee 2007; Chiu et al. 2007; Smith et al. 2007). By the same token, both cognition- and affect-based trust perceptions of the community in general may determine whether or not members of a brand community act on the information they acquire from peer members of the community. Therefore, the following set of hypotheses is postulated to delve into the effects of both types of trust on eWOM intention.
過去的研究表明，當人們認為資訊來源值得信賴時，比從不太可信的來源接收資訊時，他們更有可能傳播資訊（Phelps 等人，2004 年）。也就是說，當提供資訊或意見的來源被視為擁有專業知識和誠即時，來自該來源的資訊往往具有感知價值並傳遞給其他人（Allsop 等人，2007 年;Brown、Broderick 和 Lee 2007;Chiu 等人，2007 年;Smith 等人，2007 年）。同樣，基於認知和情感的社區信任感知通常都可能決定品牌社區的成員是否根據他們從社區的同行成員那裡獲得的信息採取行動。因此，假設以下一組假設來深入研究這兩種類型的信任對 eWOM 意圖的影響。

H5: Online brand community members’ cognition-based community trust is positively related to their eWOM intentions (a) to give and (b) to obtain information within the brand community as well as © to pass along information to others outside of the community.
H5：在線品牌社區成員基於認知的社區信任與他們在網路口碑上的意圖呈正相關 （a） 給予和 （b） 在品牌社區內獲取資訊，以及©將資訊傳遞給社區外的其他人。

H6: Online brand community members’ affect-based community trust is positively related to their eWOM intentions (a) to give and (b) to obtain information within the brand community as well as © to pass along information to others outside of the community.
H6：在線品牌社區成員基於情感的社區信任與他們的在線口碑意圖呈正相關 （a） 給予和 （b） 在品牌社區內獲取資訊，以及©將資訊傳遞給社區外的其他人。

To examine the proposed model and hypothesized relationships, an online survey of brand community members was carried out. Previous studies showed that vibrant brand communities usually evolve around brands with a strong image and a rich and lengthy heritage such as Jeep, Apple and Harley-Davidson (Muniz and O’Guinn 2001). Along this line, members of MINI Cooper online communities were selected for the sample in this study. MINI Cooper possesses a unique set of attributes that attracts a strong group of
為了檢查所提出的模型和假設的關係，對品牌社區成員進行了在線調查。先前的研究表明，充滿活力的品牌社區通常圍繞著具有強大形象和豐富而悠久傳統的品牌發展，例如 Jeep、Apple 和 Harley-Davidson（Muniz 和 O'Guinn 2001）。沿著這條思路，MINI Cooper 在線社區的成員被選為本研究的樣本。MINI Cooper 擁有一系列獨特的品質，吸引了眾多
enthusiastic customers in the small car segment, and a large number of communities for this brand reside online (Dyer 2006).
小型車領域的熱情客戶，以及該品牌的大量社區都居住在網上（Dyer 2006）。

Using popular search engines, including Yahoo, Google and MSN, a comprehensive catalog of MINI Cooper online communities was first obtained. After contacting the identified communities, we successfully gained permission from five communities to post the survey link along with an invitation message to their discussion forums. All of the five communities met the following requirements of being an ‘active’ online community suggested in the literature (Ridings et al. 2002, 282): (1) the bulletin board must have at least 10 postings per day for each of three days chosen at random; (2) the bulletin board must have at least 15 different individuals posting over a randomly selected three-day period; and (3) at least $80\mathrm{\%}$$80\mathrm{\%}$80%80 \% of the postings must have at least one reply for each of the three days chosen at random. A total of 222 surveys were collected. After incomplete surveys were eliminated, 172 usable surveys were included in the final sample for analysis. Since valid information on the number of community members was unavailable, a response rate was not estimated.
通過使用流行的搜尋引擎，包括 Yahoo、Google 和 MSN，首先獲得了 MINI Cooper 在線社區的綜合目錄。在聯繫了已確定的社區后，我們成功獲得了 5 個社區的許可，將調查連結連同邀請消息發佈到他們的論壇。這五個社區都滿足文獻中建議的成為「活躍」在線社區的以下要求（Ridings 等人，2002,282）:(1） 公告板必須在隨機選擇的三天中每天至少有 10 個帖子;（2） 公告板必須在隨機選擇的三天內至少有 15 個不同的個人發帖;（3） 至少 $80\mathrm{\%}$$80\mathrm{\%}$80%80 \% 每個帖子必須在隨機選擇的三天中的每一天至少有一個回復。共收集了 222 份調查。在排除不完整的調查後，將 172 份可用的調查納入最終樣本進行分析。由於無法獲得有關社區成員數量的有效資訊，因此未估計回應率。

Of the 172 respondents, $73.3\mathrm{\%}$$73.3\mathrm{\%}$73.3%73.3 \% were male and $26.2\mathrm{\%}$$26.2\mathrm{\%}$26.2%26.2 \% female, with the largest portion ( $32.6\mathrm{\%}$$32.6\mathrm{\%}$32.6%32.6 \% ) between the ages of $36-50$$36-50$36-5036-50 and married ( $68.0\mathrm{\%}$$68.0\mathrm{\%}$68.0%68.0 \% ). The majority of the respondents classified themselves as Caucasian ( $88.4\mathrm{\%}$$88.4\mathrm{\%}$88.4%88.4 \% ), holding a college degree ( $39.5\mathrm{\%}$$39.5\mathrm{\%}$39.5%39.5 \% ) and earning a household income greater than $\mathrm{\$}75,000(58.7\mathrm{\%})$$\mathrm{\$}75,000(58.7\mathrm{\%})$$75,000(58.7%)\$ 75,000(58.7 \%). As for community website usage, $50\mathrm{\%}$$50\mathrm{\%}$50%50 \% of the respondents reported visiting their community website daily, and $41.9\mathrm{\%}$$41.9\mathrm{\%}$41.9%41.9 \% and $31.9\mathrm{\%}$$31.9\mathrm{\%}$31.9%31.9 \% were found to stay on the website for half an hour to one hour and one to two hours, respectively, on an average day. Also, $59.9\mathrm{\%}$$59.9\mathrm{\%}$59.9%59.9 \% made a new connection or more in the community, and the average number of newly formed relationships was 16 . Table 1 presents demographic characteristics of the survey respondents.
在 172 名受訪者中， $73.3\mathrm{\%}$$73.3\mathrm{\%}$73.3%73.3 \% 男性和 $26.2\mathrm{\%}$$26.2\mathrm{\%}$26.2%26.2 \% 女性，其中最大比例 （ $32.6\mathrm{\%}$$32.6\mathrm{\%}$32.6%32.6 \% ） 介於年齡 $36-50$$36-50$36-5036-50 和已婚 （ $68.0\mathrm{\%}$$68.0\mathrm{\%}$68.0%68.0 \% ） 之間。大多數受訪者將自己歸類為白種人 （ $88.4\mathrm{\%}$$88.4\mathrm{\%}$88.4%88.4 \% ），持有大學學位 （ $39.5\mathrm{\%}$$39.5\mathrm{\%}$39.5%39.5 \% ） 且家庭收入大於 $\mathrm{\$}75,000(58.7\mathrm{\%})$$\mathrm{\$}75,000(58.7\mathrm{\%})$$75,000(58.7%)\$ 75,000(58.7 \%) 。至於社區網站的使用方式， $50\mathrm{\%}$$50\mathrm{\%}$50%50 \% 受訪者表示每天都會訪問他們的社區網站， $41.9\mathrm{\%}$$41.9\mathrm{\%}$41.9%41.9 \% 並且 $31.9\mathrm{\%}$$31.9\mathrm{\%}$31.9%31.9 \% 被發現平均每天在網站上停留的時間分別為半小時至一小時和一至兩小時。此外， $59.9\mathrm{\%}$$59.9\mathrm{\%}$59.9%59.9 \% 在社區中建立了新的聯繫或更多，新建立的關係的平均數量為 16。表 1 顯示了調查受訪者的人口統計特徵。

The measures employed in this study were borrowed from the existing literature and adapted to the context of the study. Using a three-item, seven-point, Likert-type scale ranging from (1) strongly disagree to (7) strongly agree, Brand Identification, the degree of a consumer’s self-brand association, was assessed (Algesheimer et al. 2005). Community Identification, a consumer’s self-awareness and affective commitment to the community was measured via a five-item, seven-point, Likert-type scale, with 1 being ‘strongly disagree’ and 7 being ‘strongly agree’ (Algesheimer et al. 2005). Brand Loyalty was gauged via a three-item, seven-point, Likert-type scale examining a consumer’s purchase intention of brand-related products (Algesheimer et al. 2005). Trust among brand community members was conceptualized as a belief with two dimensions: CognitionBased Community Trust and Affect-Based Community Trust. Each type of trust perception was assessed with a six-item, seven-point, Likert-type scale (Ridings et al. 2002).
本研究中採用的措施借鑒了現有文獻，並適應了研究的背景。使用從 （1） 非常不同意到 （7） 非常同意的三項七點李克特式量表，評估了品牌識別，即消費者自我品牌聯想的程度（Algesheimer 等人，2005 年）。社區認同，即消費者的自我意識和對社區的情感承諾，是通過一個包含五個專案、七點的李克特式量表來衡量的，其中 1 個是“非常不同意”，7 個是“非常同意”（Algesheimer 等人，2005 年）。品牌忠誠度是通過一個包含三個專案、七點的李克特式量表來衡量的，該量表檢查了消費者對品牌相關產品的購買意向（Algesheimer 等人，2005 年）。品牌社區成員之間的信任被概念化為具有兩個維度的信念：基於認知的社區信任和基於情感的社區信任。每種類型的信任感知都使用六項、七點的李克特式量表進行評估（Ridings 等人，2002 年）。

Two items measured Intention to Give Information, a community member’s intention to share their knowledge, and three items assessed Intention to Obtain Information, a community member’s intention to acquire information and advice related to brand-related topics on the brand community website (Ridings et al. 2002). Both constructs were gauged
兩個項目測量了提供資訊的意圖，即社區成員分享其知識的意圖，三個專案評估了獲取資訊的意圖，即社區成員在品牌社區網站上獲取與品牌相關主題相關的資訊和建議的意圖（Ridings 等人，2002 年）。兩種結構都經過了測量

Table 1. Characteristics of the respondents.
表 1.受訪者的特徵。

Notes: Total $N=172$$N=172$N=172N=172.  注： 總計 $N=172$$N=172$N=172N=172 .
${}^3$${}^3$^(3){ }^{3} Number of missing data $=1$$=1$=1=1.
${}^3$${}^3$^(3){ }^{3} Number of missing data （缺失數據 $=1$$=1$=1=1 數） 。
${}^1$${}^1$^(1){ }^{1} Number of missing data $=1$$=1$=1=1.
${}^1$${}^1$^(1){ }^{1} Number of missing data （缺失數據 $=1$$=1$=1=1 數） 。
"Number of missing data $=3$$=3$=3=3.
遺失資料 $=3$$=3$=3=3 數 .
${}^{\mathrm{d}}$${}^{\mathrm{d}}$^(d){ }^{\mathrm{d}} Number of missing data $=3$$=3$=3=3.
${}^{\mathrm{d}}$${}^{\mathrm{d}}$^(d){ }^{\mathrm{d}} Number of missing data （缺失數據 $=3$$=3$=3=3 數） 。
${}^{\mathrm{e}}$${}^{\mathrm{e}}$^(e){ }^{\mathrm{e}} Number of missing data $=3$$=3$=3=3.
${}^{\mathrm{e}}$${}^{\mathrm{e}}$^(e){ }^{\mathrm{e}} Number of missing data （缺失數據 $=3$$=3$=3=3 數） 。
with a seven-point, Likert-type scale with the endpoints of 1 being ‘strongly disagree’ and 7 being ‘strongly agree’. A four-item, seven-point, Likert-type scale estimated Intention to Pass along Information, a community member’s intention to disseminate the information acquired from the brand community to other non-members. Specific measurement items appear in Table 2.
採用七點李克特式量表，終點 1 為“非常不同意”，7 為“非常同意”。一個由四個項目組成的7點李克特式量表估計了傳遞資訊的意向，即社區成員將從品牌社區獲得的信息傳播給其他非成員的意圖。具體測量專案如表 2 所示。

Before testing the hypothesized causal relationships, a confirmatory analysis using AMOS 7 assessed if the multi-item scales achieved satisfactory levels of reliability and whether
在測試假設的因果關係之前，使用 AMOS 7 的驗證性分析評估了多專案量表是否達到令人滿意的可靠性水準，以及

Table 2. Measures, factor loadings, descriptive statistics and reliability.
表 2.測量、因數載荷、描述性統計和可靠性。

Table 2 - continued
表 2 - 續

Notes:  筆記：
*All coefficients are significant at $p<.001$$p<.001$p < .001p<.001 and generated from a final confirmatory factor analysis after eliminating the last item of affect-based trast ${}^{\mathrm{h}}$${}^{\mathrm{h}}$^(h){ }^{\mathrm{h}}.
*所有係數在 $p<.001$$p<.001$p < .001p<.001 消除基於情感的 trast ${}^{\mathrm{h}}$${}^{\mathrm{h}}$^(h){ }^{\mathrm{h}} 的最後一項后，從最終驗證性因數分析中生成。
${}^a$${}^a$^(a){ }^{a} Reference indicator.
${}^a$${}^a$^(a){ }^{a} 參考指示符。
${}^5$${}^5$^(5){ }^{5} The item is reverse coded and not included in the final measurement model due to the loading $<.40$$<.40$< .40<.40. The coefficients are obtained from an initial confirmatory analysis.
${}^5$${}^5$^(5){ }^{5} 該專案被反向編碼，由於載入 $<.40$$<.40$< .40<.40 ，未包含在最終測量模型中。係數是從初始驗證分析中獲得的。
factor loadings significantly related to their corresponding constructs. With the use of existing scale items adapted from the literature for measuring the latent variables, the observed variables in the model were constrained to load on only one factor and the error terms were not allowed to co-vary. Results demonstrated that all of the items significantly loaded on the relevant construct ( $p<.01$$p<.01$p < .01p<.01 ). The factor loading of .317 for one indicator of affect-based-trust, however, was low and was accordingly eliminated. All the other indicators with loadings larger than .60 were deemed as appropriate (Hair et al. 1998). Reliabilities of all resulting scales were assessed and found to be internally consistent. Subsequently, an index score measuring each of the variables was created by averaging up the corresponding item scores. Table 2 presents the factor loadings and reliability coefficients for all variables in the model. The correlations for the variables in the model are reported in Table 3.
因數載荷與其相應的結構顯著相關。通過使用從文獻改編的現有尺度項目來測量潛在變數，模型中的觀測變數被約束為僅在一個因數上載入，並且不允許誤差項協變。結果表明，所有項目都顯著載入到相關結構 （ $p<.01$$p<.01$p < .01p<.01 ） 上。然而，對於基於情感的信任的一個指標，.317 的因數負載很低，因此被消除了。載荷大於 .60 的所有其他指標都被認為是合適的（Hair 等人，1998 年）。評估了所有結果量表的可靠性，發現其內部一致性。隨後，通過平均相應的專案分數來創建衡量每個變數的指數分數。表 2 顯示了模型中所有變數的因數載荷和可靠性係數。表中各變數的相關性如表 3 所示。

The proposed model and hypothesized linkages among the variables within the model were tested with path analysis using AMOS 7 . The large number of measured variables in the model but the relatively small sample size suggested a path analysis. A sufficient number of cases, 15 or more per measured variable, are recommended for structural equation modeling (Stevens 1996), but the proposed model has a total of 31 indicators while the sample included 172 cases. Following the procedure suggested in the literature (Joreskog and Sorbom 1993; see Goldsmith, Lafferty, and Newell 2000; Goldsmith 2002 for examples), a path analysis with index scores of all variables assessed the causal linkages in the proposed model.
使用 AMOS 7 通過路徑分析測試了所提出的模型和模型中變數之間的假設聯繫。模型中的測量變數數量眾多，但樣本量相對較小，這表明需要進行路徑分析。建議為結構方程建模提供足夠數量的案例，每個測量變數 15 個或更多案例 （Stevens 1996），但所提出的模型共有 31 個指標，而樣本包括 172 個案例。遵循文獻中建議的程式（Joreskog 和 Sorbom 1993;參見 Goldsmith、Lafferty 和 Newell 2000;例如，Goldsmith 2002），所有變數的指數分數的路徑分析評估了所提出的模型中的因果關係。

The results showed that, of the 13 hypothesized causal relationships, 10 were statistically significant ( $p<.05$$p<.05$p < .05p<.05 ) and one was marginally significant ( $p<.1$$p<.1$p < .1p<.1 ) in the expected direction. In support of H1 and H2, community members’ identification with the brand appeared to have a positive, significant impact on brand loyalty (standardized path coefficient $=.60,p<.001$$=.60,p<.001$=.60,p < .001=.60, p<.001 ) and community identification (standardized path coefficient $=.48,p<.001$$=.48,p<.001$=.48,p < .001=.48, p<.001 ). In examining the relationships of brand loyalty to eWOM intention, brand loyalty was found to have a significant, positive effect on community members’ intentions to obtain information (standardized path coefficient $=.19,p<.01$$=.19,p<.01$=.19,p < .01=.19, p<.01 )
結果表明，在 13 個假設的因果關係中，10 個在預期方向上具有統計學意義 （ $p<.05$$p<.05$p < .05p<.05 ），1 個在預期方向上具有輕微顯著性 （ $p<.1$$p<.1$p < .1p<.1 ）。為了支援 H1 和 H2，社區成員對品牌的認同似乎對品牌忠誠度（標準化路徑係數 $=.60,p<.001$$=.60,p<.001$=.60,p < .001=.60, p<.001 ）和社區認同（標準化路徑係數）產生了積極、顯著的影響 $=.48,p<.001$$=.48,p<.001$=.48,p < .001=.48, p<.001 。在研究品牌忠誠度與網路口碑意圖的關係時，發現品牌忠誠度對社區成員獲取資訊的意願有顯著的積極影響（標準化路徑係數 $=.19,p<.01$$=.19,p<.01$=.19,p < .01=.19, p<.01 ）

Table 3. Correlations among variables.
表 3.變數之間的相關性。

Note: ${}^{4}p>.05$${}^{4}p>.05$^(4)p > .05{ }^{4} p>.05, all other coelficients are significant at $p<.001$$p<.001$p < .001p<.001 (two-tailed).
注： ${}^{4}p>.05$${}^{4}p>.05$^(4)p > .05{ }^{4} p>.05 ，所有其他腔體在 （雙尾） 處 $p<.001$$p<.001$p < .001p<.001 均顯著。
and to pass along information (standardized path coefficient $=.35,p<.001$$=.35,p<.001$=.35,p < .001=.35, p<.001 ) while the impact of brand loyalty on the intention of offering information was not significant ( $p>.1$$p>.1$p > .1p>.1 ). Thus, H3b and H3c were supported whereas H3a was disconfirmed.
並傳遞資訊（標準化路徑係數 $=.35,p<.001$$=.35,p<.001$=.35,p < .001=.35, p<.001 ），而品牌忠誠度對提供資訊意圖的影響並不顯著 （ $p>.1$$p>.1$p > .1p>.1 ）。因此，H3b 和 H3c 得到支援，而 H3a 被否定。

As predicted in H 4 a and H 4 b , community identification exerted a significant, positive influence on both cognition-based trust (standardized path coefficient $=.53,p<.001$$=.53,p<.001$=.53,p < .001=.53, p<.001 ) and affect-based trust (standardized path coefficient $=.55,p<.001$$=.55,p<.001$=.55,p < .001=.55, p<.001 ). The last sets of hypotheses examined the effects of both types of trust on eWOM intention among brand community members. Confirming H5a, H5b and H5c, community members’ cognitionbased trust in the brand community significantly influenced their willingness to engage in eWOM behaviors. Specifically, cognition-based trust led to the stronger intentions to give information (standardized path coefficient $=.20,p<.05$$=.20,p<.05$=.20,p < .05=.20, p<.05 ) and to obtain information (standardized path coefficient $=.41,p<.001$$=.41,p<.001$=.41,p < .001=.41, p<.001 ) within the community as well as the greater likelihood of their passing information to non-community members (standardized path coefficient $=.29,p<.001$$=.29,p<.001$=.29,p < .001=.29, p<.001 ). When the impact of affect-based trust on eWOM intention was examined, the results were mixed. Affect-based trust was found to significantly influence the intention to pass along information (standardized path coefficient $=.24,p<.01$$=.24,p<.01$=.24,p < .01=.24, p<.01 ), lending support for H6c. The effect of affect-based trust on the intention to give information was marginally significant (standardized path coefficient $=.17,p=.06$$=.17,p=.06$=.17,p=.06=.17, p=.06 ), and its impact on the willingness to obtain information was not significant ( $p>.1$$p>.1$p > .1p>.1 ). However, H6a was marginally supported, and H6b was disconfirmed. Table 4 displays the path coefficients, both unstandardized and standardized.
正如 H 4 a 和 H 4 b 中預測的那樣，社區認同對基於認知的信任（標準化路徑係數 $=.53,p<.001$$=.53,p<.001$=.53,p < .001=.53, p<.001 ）和基於情感的信任（標準化路徑係數）都產生了顯著的積極影響 $=.55,p<.001$$=.55,p<.001$=.55,p < .001=.55, p<.001 。最後一組假設研究了這兩種類型的信任對品牌社區成員的 eWOM 意圖的影響。確認 H5a、H5b 和 H5c，社區成員對品牌社區的認知信任顯著影響了他們參與網路口碑行為的意願。具體來說，基於認知的信任導致社區內提供更強的資訊（標準化路徑係數 $=.20,p<.05$$=.20,p<.05$=.20,p < .05=.20, p<.05 ）和獲取資訊（標準化路徑係數 $=.41,p<.001$$=.41,p<.001$=.41,p < .001=.41, p<.001 ）的意願，以及他們向非社區成員傳遞資訊的可能性更大（標準化路徑係數 $=.29,p<.001$$=.29,p<.001$=.29,p < .001=.29, p<.001 ）。當檢查基於情感的信任對 eWOM 意圖的影響時，結果喜憂參半。研究發現，基於情感的信任顯著影響傳遞資訊的意願 （標準化路徑係數 $=.24,p<.01$$=.24,p<.01$=.24,p < .01=.24, p<.01 ），為 H6c 提供支援。基於情感的信任對提供資訊意願的影響略有顯著性 （標準化路徑係數 $=.17,p=.06$$=.17,p=.06$=.17,p=.06=.17, p=.06 ），而它對獲取資訊意願的影響並不顯著 （ $p>.1$$p>.1$p > .1p>.1 ）。然而，H6a 得到微弱支援，H6b 被否定。表 4 顯示了非標準化和標準化的路徑係數。

The goodness-of-fit indices suggest the model did fit the data fairly well; ${\chi }^2=51.86$${\chi }^2=51.86$chi^(2)=51.86\chi^{2}=51.86, d.f. $=11,p<.001;\mathrm{GFI}=.94;\mathrm{AGFI}=.79;\mathrm{CFI}=.92;\mathrm{RMR}=.12;\mathrm{RMSEA}=.15$$=11,p<.001;\mathrm{GFI}=.94;\mathrm{AGFI}=.79;\mathrm{CFI}=.92;\mathrm{RMR}=.12;\mathrm{RMSEA}=.15$=11,p < .001;GFI=.94;AGFI=.79;CFI=.92;RMR=.12;RMSEA=.15=11, p<.001 ; \mathrm{GFI}=.94 ; \mathrm{AGFI}=.79 ; \mathrm{CFI}=.92 ; \mathrm{RMR}=.12 ; \mathrm{RMSEA}=.15. In order to refine the originally proposed model and achieve parsimony, the two nonsignificant relationships were tested in the second phase of the analysis by systematically relaxing a restriction and examining the resultant change in chi-square. That is, the chisquare value of an alternative model without each of the non-significant relationships was re-estimated and compared to that of the original model. The results showed that the paths (1) from brand loyalty to the intention to give information (H3a) and (2) from affect-based trust to the intention to obtain information (H6b) did not yield significant changes in the chi-square of the model fit: (1) ${\chi }^2{}_{\text{difference}}=.03$${\chi }^2{}_{\text{difference }}=.03$chi^(2)_("difference ")=.03\chi^{2}{ }_{\text {difference }}=.03, d.f. $=1,p>.05$$=1,p>.05$=1,p > .05=1, p>.05; (2) ${\chi }_{\text{difference}}^2=.01$${\chi }_{\text{difference }}^2=.01$chi_("difference ")^(2)=.01\chi_{\text {difference }}^{2}=.01, d.f. $=1,p>.05$$=1,p>.05$=1,p > .05=1, p>.05. As a result, both paths, which did not significantly contribute to the improvement of the model fit, were eliminated from the model.
擬合優度指數表明該模型確實很好地擬合了數據; ${\chi }^2=51.86$${\chi }^2=51.86$chi^(2)=51.86\chi^{2}=51.86 ， d.f. $=11,p<.001;\mathrm{GFI}=.94;\mathrm{AGFI}=.79;\mathrm{CFI}=.92;\mathrm{RMR}=.12;\mathrm{RMSEA}=.15$$=11,p<.001;\mathrm{GFI}=.94;\mathrm{AGFI}=.79;\mathrm{CFI}=.92;\mathrm{RMR}=.12;\mathrm{RMSEA}=.15$=11,p < .001;GFI=.94;AGFI=.79;CFI=.92;RMR=.12;RMSEA=.15=11, p<.001 ; \mathrm{GFI}=.94 ; \mathrm{AGFI}=.79 ; \mathrm{CFI}=.92 ; \mathrm{RMR}=.12 ; \mathrm{RMSEA}=.15 .為了改進最初提出的模型並實現簡潔，在分析的第二階段通過系統地放寬限制並檢查由此產生的卡方變化來測試這兩個不顯著的關係。也就是說，重新估計沒有每個非顯著關係的備擇模型的卡方值，並與原始模型的卡方值進行比較。結果表明，路徑 （1） 從品牌忠誠度到提供資訊的意向 （H3a） 和 （2） 從基於情感的信任到獲取資訊的意圖 （H6b） 在模型擬合的卡方中沒有產生顯著變化：（1） ${\chi }^2{}_{\text{difference}}=.03$${\chi }^2{}_{\text{difference }}=.03$chi^(2)_("difference ")=.03\chi^{2}{ }_{\text {difference }}=.03 ， d.f. $=1,p>.05$$=1,p>.05$=1,p > .05=1, p>.05 ;（2） ${\chi }_{\text{difference}}^2=.01$${\chi }_{\text{difference }}^2=.01$chi_("difference ")^(2)=.01\chi_{\text {difference }}^{2}=.01 ， d.f. $=1,p>.05$$=1,p>.05$=1,p > .05=1, p>.05 .結果，這兩條路徑都被從模型中消除了，這對模型擬合的改進沒有顯著貢獻。

Table 4. Parameter estimates and model-fit statistics.
表 4.參數估計值和模型擬合統計量。

Notes: ${}^{a}p<.1,{}^{\ast }p<.05,{}^{\ast \ast }p<.01$${}^{a}p<.1,{}^{\ast }p<.05,{}^{\ast \ast }p<.01$^(a)p < .1,^(**)p < .05,^(****)p < .01{ }^{a} p<.1,{ }^{*} p<.05,{ }^{* *} p<.01, all other coefficients are not significant, $p>.1$$p>.1$p > .1p>.1.
注： ${}^{a}p<.1,{}^{\ast }p<.05,{}^{\ast \ast }p<.01$${}^{a}p<.1,{}^{\ast }p<.05,{}^{\ast \ast }p<.01$^(a)p < .1,^(**)p < .05,^(****)p < .01{ }^{a} p<.1,{ }^{*} p<.05,{ }^{* *} p<.01 ，所有其他係數均不顯著，。 $p>.1$$p>.1$p > .1p>.1
The modified model, which was more parsimonious without non-significant linkages, was chosen as the final model. The path coefficients were re-estimated and presented in Table 4. Note that in this model, the marginally significant relationship between affectbased trust and the intention to provide information in the initial testing became significant (standardized path coefficient $=.17,p<.05$$=.17,p<.05$=.17,p < .05=.17, p<.05 ). Overall, this final model accounts for the data slightly better than the original model: ${\chi }^2=49.10$${\chi }^2=49.10$chi^(2)=49.10\chi^{2}=49.10, d.f. $=13,p=<.001;\mathrm{GFI}=.94$$=13,p=<.001;\mathrm{GFI}=.94$=13,p=<.001;GFI=.94=13, p=<.001 ; \mathrm{GFI}=.94; $\mathrm{AGFI}=.83;\mathrm{CFI}=.93;\mathrm{RMR}=.12;\mathrm{RMSEA}=.12$$\mathrm{AGFI}=.83;\mathrm{CFI}=.93;\mathrm{RMR}=.12;\mathrm{RMSEA}=.12$AGFI=.83;CFI=.93;RMR=.12;RMSEA=.12\mathrm{AGFI}=.83 ; \mathrm{CFI}=.93 ; \mathrm{RMR}=.12 ; \mathrm{RMSEA}=.12. While this model, on the whole, did not fit the data at the desired level, it demonstrates the significant relationships among the key variables in the eWOM process in the online brand community context, which should serve as a useful basis for further model development in future research.
修改後的模型更簡潔，沒有不顯著的聯繫，被選為最終模型。重新估計了路徑係數，並在表 4 中列出了。請注意，在此模型中，基於情感的信任與在初始測試中提供資訊的意圖之間的邊際顯著關係變得顯著顯著（標準化路徑係數 $=.17,p<.05$$=.17,p<.05$=.17,p < .05=.17, p<.05 ）。總體而言，這個最終模型對數據的解釋略好於原始模型： ${\chi }^2=49.10$${\chi }^2=49.10$chi^(2)=49.10\chi^{2}=49.10 ， d.f. $=13,p=<.001;\mathrm{GFI}=.94$$=13,p=<.001;\mathrm{GFI}=.94$=13,p=<.001;GFI=.94=13, p=<.001 ; \mathrm{GFI}=.94 ; $\mathrm{AGFI}=.83;\mathrm{CFI}=.93;\mathrm{RMR}=.12;\mathrm{RMSEA}=.12$$\mathrm{AGFI}=.83;\mathrm{CFI}=.93;\mathrm{RMR}=.12;\mathrm{RMSEA}=.12$AGFI=.83;CFI=.93;RMR=.12;RMSEA=.12\mathrm{AGFI}=.83 ; \mathrm{CFI}=.93 ; \mathrm{RMR}=.12 ; \mathrm{RMSEA}=.12 。雖然這個模型總體上沒有達到預期的數據擬合水準，但它證明瞭在線品牌社區背景下網路口碑過程中關鍵變數之間的顯著關係，這應該為未來研究的進一步模型開發提供有用的基礎。

The current study examined the links between the relational dimensions of an online brand community and consumers’ eWOM intentions, namely their intention to exchange information within the community as well as their intent to pass along the information to non-community members. Although a consumer’s identifications with a brand and the brand community the consumer belongs to as well as brand loyalty were suggested to contribute to positive brand-focused WOM in previous research, few studies, if any, have empirically investigated how precisely WOM among brand community members takes place in an online environment. Findings of this study advance our understanding of brand-related eWOM by delineating the eWOM process and its antecedents in an online brand community setting.
目前的研究考察了在線品牌社區的關係維度與消費者的 eWOM 意圖之間的聯繫，即他們在社區內交換資訊的意圖以及他們將資訊傳遞給非社區成員的意圖。儘管在以前的研究中，消費者對品牌和消費者所屬的品牌社區的認同以及品牌忠誠度被認為有助於以品牌為中心的積極口碑，但很少有研究（如果有的話）實證調查了品牌社區成員中的口碑是如何在在線環境中發生的。這項研究的結果通過在在線品牌社區環境中描述網路口碑過程及其前因，促進了我們對品牌相關網路口碑的理解。

The results of this study highlight the consumer-brand relationship as an important starting point for eWOM and suggest that the influence of brand identification on eWOM intentions among online brand community members flows via two paths. First, a brand community member’s identification with the brand leads to enhanced brand loyalty, which in turn increases the member’s intention to engage in eWOM in favor of the brand. This study provided empirical support for the notion that a brand community is often composed of brand enthusiasts who act as evangelists for the brand. Brand community members’ affinity for the brand indeed predicts their positive eWOM intention for the brand, and their strong commitment to the brand mediates this influence of brand identification on eWOM intention.
本研究的結果強調消費者與品牌的關係是網路口碑的重要起點，並表明品牌識別對在線品牌社區成員網路口碑意圖的影響通過兩條路徑流動。首先，品牌社區成員對品牌的認同會增強品牌忠誠度，這反過來又會增加成員參與網路口碑以支援品牌的意願。這項研究為品牌社區通常由充當品牌傳播者的品牌愛好者組成這一概念提供了實證支援。品牌社區成員對品牌的親和力確實預示了他們對品牌的積極網路口碑意向，而他們對品牌的強烈承諾則仲介了品牌識別對網路口碑意向的影響。

Second, brand community members’ identification with the brand positively influences their identification with the brand community which subsequently enhances trust among the brand community members. This finding is in line with our prediction drawn from social identity theory suggesting that a sense of belonging and shared membership among brand community members enhance their feeling of trust in fellow members’ ability and benevolence. This improved trust subsequently leads to stronger eWOM intention among brand community members. This finding highlights the importance of the relational dimension of the brand community in producing positive outcomes for the brand. Not only a close consumer-brand relationship but also a strong affiliation with the brand community can contribute to eWOM in favor of the brand.
其次，品牌社區成員對品牌的認同會對他們對品牌社區的認同產生積極影響，從而增強品牌社區成員之間的信任。這一發現與我們從社會認同理論中得出的預測一致，即品牌社區成員之間的歸屬感和共同的歸屬感增強了他們對其他成員的能力和仁慈的信任感。這種信任的提高隨後導致品牌社區成員對網路口碑的意圖更加強烈。這一發現凸顯了品牌社區的關係維度在為品牌產生積極成果方面的重要性。不僅消費者與品牌建立密切的關係，而且與品牌社區的緊密聯繫，都有助於網路口碑的發展，從而有利於品牌。

An interesting finding is that trust in peer community members, both affect and cognition dimensions, is a significant mediator of community identification on eWOM and a direct predictor of eWOM intention among brand community members. While previous empirical studies suggest that community identification enhances information sharing behaviors among community members, the results of this study underscore trust, especially cognition-based trust, as a direct pathway to eWOM in terms of the three investigated dimensions (giving, receiving and passing) in an online brand community environment. In partial support of our prediction, affect-based trust only appeared to significantly predict brand community members’ intention to give information to peer members and pass information on to outsiders. A possible speculation is that the insignificant link between affect-based trust and intention to obtain information would be due in part to the nature of the car brand around which the communities are formed. A car is typically considered as a high involvement product involving a great deal of technical knowledge. The car brand communities from which the sample was drawn tended to discuss many technical issues such as car repairs and reforms, and the focus on technical knowledge may have led the community members to weigh competence of the information sources more heavily than their integrity (Levin and Cross 2004). Taken together, these findings clearly point to the importance of trust as a vital factor for facilitating eWOM in an online community setting wherein face to face interactions are absent.
一個有趣的發現是，對同伴社區成員的信任，包括情感和認知維度，是社區對網路口碑認同的重要仲介，也是品牌社區成員網路口碑意圖的直接預測因數。雖然之前的實證研究表明，社區認同增強了社區成員之間的資訊共享行為，但本研究的結果強調了信任，尤其是基於認知的信任，是在線品牌社區環境中三個研究維度（給予、接受和傳遞）的直接途徑。在一定程度上支援我們的預測，基於情感的信任似乎只顯著地預測了品牌社區成員向同行成員提供資訊並將資訊傳遞給外部人的意圖。一種可能的推測是，基於情感的信任和獲取資訊的意圖之間的微不足道的聯繫，部分是由於社區所圍繞的汽車品牌的性質。汽車通常被認為是涉及大量技術知識的高參與度產品。樣本來自的汽車品牌社區傾向於討論許多技術問題，例如汽車維修和改革，而對技術知識的關注可能導致社區成員更重視資訊來源的能力，而不是他們的完整性（Levin 和 Cross 2004）。綜上所述，這些發現清楚地表明，在沒有面對面互動的在線社區環境中，信任是促進網路口碑的重要因素。

For marketers, the results of this study suggest that online brand community members actively engage in eWOM within and beyond their communities and may serve as brand evangelists and effective message transmitters for viral marketing (Dobele, Toleman, and Beverland 2005). In light of the findings, marketers can magnify the eWOM effects and the flow of information dissemination by not only building a strong brand for which consumers develop a strong affinity but also by building a vibrant brand community in which its members bond closely together. Strategies promoting a unique community identity and brand culture that can sharply set apart in-group and out-group members, such as offering exclusive product information or product promotion to the community
對於營銷人員來說，這項研究的結果表明，在線品牌社區成員在其社區內外積极參與網路口碑，並可能成為品牌傳播者和病毒式行銷的有效資訊傳播者（Dobele、Toleman 和 Beverland 2005）。鑒於這些發現，行銷人員不僅可以建立一個消費者產生強烈親和力的強大品牌，還可以通過建立一個充滿活力的品牌社區，讓其成員緊密聯繫在一起，從而放大網路口碑效應和資訊傳播的流動。推廣獨特的社區身份和品牌文化的策略，可以明顯區分群體內和外群體成員，例如向社區提供獨家產品資訊或產品促銷
members, could potentially enhance consumers’ identification with the brand community and, in turn, increase their group-based trust. Trust development is also seen as a continued, reciprocal process based on personal interactions (Crossman and Lee-Kelley 2004). Marketers could encourage interactions among brand community members by integrating social networking features and communication tools in the development of brand community websites.
成員可能會增強消費者對品牌社區的認同感，進而增加他們基於群體的信任。信任發展也被視為一個基於個人互動的持續、互惠的過程（Crossman 和 Lee-Kelley 2004）。營銷人員可以通過在品牌社區網站的開發中集成社交網路功能和通信工具來鼓勵品牌社區成員之間的互動。

This study has a few limitations that should be acknowledged and discussed toward directions for future research. First, the sample employed in this study limits generalizability of the results. Although a huge roster of brand communities exists on the Internet, our sample comprised members of online communities of a single brand. Also, since a comprehensive list of members was not accessible, the survey data were collected on a voluntary basis. In addition, the examination of the sample characteristics revealed that the respondents were energetic members with moderate to strong engagement in the community in terms of community website usage and relationship development. While, given the rather restricted access to brand communities, the sample of active brand community members still shed light on the topic of investigation and was deemed as adequate for testing of the theoretical model, future research should employ a more representative sample to provide further corroboration of the findings.
這項研究有一些局限性，應該承認並討論未來的研究方向。首先，本研究中使用的樣本限制了結果的普遍性。儘管互聯網上存在大量的品牌社區，但我們的樣本由單個品牌的在線社區成員組成。此外，由於無法獲得完整的成員名單，因此調查數據是在自願的基礎上收集的。此外，對樣本特徵的檢查表明，受訪者是精力充沛的成員，在社區網站使用和關係發展方面對社區的參與度為中度至高度。雖然鑒於品牌社區的訪問相當有限，活躍的品牌社區成員的樣本仍然闡明瞭調查主題，並被認為足以檢驗理論模型，但未來的研究應該採用更具代表性的樣本本來進一步證實研究結果。

Along the same line, future studies should examine brand communities of various product types in order to expand the findings of this study limited to a single car brand. For instance, according to the FCB grid (Vaughn 1980, 1986), brands can be categorized based on two dimensions - involvement and think/feel. This classification suggests that consumers’ cognitive and affective efforts invested in purchase decisions for products in each quadrant are different. Brands with an emphasis on symbolic meanings or affective appeals may attract more brand community members because brand communities, in general, support brands for social values rather than utilitarian values. Such social values provide a basis for self-categorization for brand users and facilitate group rituals that define and solidify the group identity and distinguish its members apart from non-brand users (Wipperfurth 2005). Yet successful communities are often observed for brands with a utilitarian focus (e.g. Dell’s Ideastorm). Therefore, it would be helpful for future research to explore and compare communities of different brands with differing characteristics and values.
同樣，未來的研究應檢查各種商品類型的品牌社區，以便將本研究的結果擴展到單個汽車品牌。例如，根據 FCB 網格（Vaughn 1980,1986），品牌可以根據兩個維度進行分類 - 參與度和思考/感覺。這種分類表明，消費者在每個象限中投資於產品購買決策的認知和情感努力是不同的。強調象徵意義或情感訴求的品牌可能會吸引更多的品牌社區成員，因為品牌社區通常支援品牌的社會價值觀而不是功利主義價值觀。這種社會價值觀為品牌使用者的自我分類提供了基礎，並促進了定義和鞏固群體身份的群體儀式，並將其成員與非品牌使用者區分開來（Wipperfurth 2005）。然而，對於注重功利主義的品牌（例如戴爾的 Ideastorm），通常會觀察到成功的社區。因此，探索和比較具有不同特徵和價值觀的不同品牌的社區將有助於未來的研究。

Although the present study focused on eWOM among online brand community members, the magnitude and mechanism of eWOM may differ between community members and non-members. Future studies with a direct comparison of eWOM patterns between brand community members and non-members and a close examination of their multiple roles in the eWOM process should reveal useful insights. Another fruitful venue for extension of this study is investigation of the social network of brand communities. The literature suggests that individuals can develop meaningful relationships only based on mediated interactions online (Rheingold 1993). Members of a brand community may build and maintain a social network within the community and the extent and density of the network would impact on eWOM. Future research may delve into the structure of the social network among brand community members such as strong and weak ties and its influence on eWOM (Granovetter 1983).
儘管本研究側重於在線品牌社區成員的 eWOM，但 eWOM 的規模和機制可能因社區成員和非成員而異。未來的研究直接比較品牌社區成員和非成員之間的網路口碑模式，並仔細研究他們在網路口碑過程中的多重角色，應該會揭示有用的見解。擴展這項研究的另一個富有成效的場所是對品牌社區社交網路的調查。文獻表明，個體只能根據在線中介互動來建立有意義的關係（Rheingold 1993）。品牌社區的成員可以在社區內建立和維護社交網路，網路的範圍和密度會影響網路口碑。未來的研究可能會深入研究品牌社區成員之間的社交網路結構，例如強弱關係及其對網路口碑的影響（Granovetter 1983）。

Another limitation of this study is the assessment of eWOM intention, not actual eWOM behavior. Although intention is generally considered as a valid proxy measure of
這項研究的另一個局限性是評估網路口碑意圖，而不是實際的網路口碑行為。儘管意圖通常被認為是
behavior, it would be worthwhile to gauge actual eWOM behaviors to examine its antecedents. Tied to the issue, it is also worthwhile to conduct a longitudinal study. Community and trust development occurs over a long period of time, and brand communities in different stages may show a different pattern of eWOM among their members. A longitudinal study of a brand community could track the potential changes of the eWOM process over time as the community evolves.
行為，那麼衡量實際的 eWOM 行為以檢查其前因是值得的。與這個問題相關，進行縱向研究也是值得的。社區和信任的發展需要很長時間才能實現，處於不同階段的品牌社區可能會在其成員中表現出不同的網路口碑模式。對品牌社區的縱向研究可以跟蹤隨著社區的發展，網路口碑流程隨時間推移的潛在變化。

Yi-Hisin Yeh (MA, the University of Texas at Austin) is a doctoral candidate in the Department of Advertising at the University of Texas at Austin. Her research interest lies in online consumer behavior with a focus on consumer-brand relationships.
Yi-Hisin Yeh（德克薩斯大學奧斯丁分校碩士）是德克薩斯大學奧斯丁分校廣告系的博士生。她的研究興趣在於在線消費者行為，重點是消費者與品牌的關係。

Sejung Marina Choi (PhD, Michigan State University) is Associate Professor of Advertising at the University of Texas at Austin. Her research interests are in the areas of source credibility, consumer-brand relationships, new media and cross-cultural consumer behavior.
Sejung Marina Choi（密歇根州立大學博士）是德克薩斯大學奧斯丁分校廣告學副教授。她的研究興趣是來源可信度、消費者與品牌關係、新媒體和跨文化消費者行為等領域。

Algesheimer, R., U.M. Dholakia, and A. Herrmann. 2005. The social influence of brand community: Evidence from European car clubs. Journal of Marketing 69, no. 3: 19-34.
Algesheimer， R.、UM. Dholakia 和 A. Herrmann。2005. 品牌社區的社會影響：來自歐洲汽車俱樂部的證據。市場營銷雜誌 69，第 3 期：19-34。
Allsop, D.T., B.R. Bassett, and J.A. Hoskins. 2007. Word-of-mouth research: Principles and applications. Journal of Advertising Research 47, no. 4: 398-411.
Allsop， D.T.， B.R. Bassett 和 JA Hoskins。2007. 口碑研究：原理和應用。廣告研究雜誌 47，第 4 期：398-411。
Bagozzi, R.P., and U.M. Dholakia. 2002. The social actions in virtual communities. Jounal of Interactive Marketing 16, no. 2: 2-21.
Bagozzi， R.P. 和 UM Dholakia。2002. 虛擬社區中的社會行動。互動行銷 16，第 2 期：2-21。
-. 2006. Antecedents and purchase consequences of customer participation in small group brand communities. International Journal of Research in Marketing 23, no. 1: 45-61.
-.2006. 顧客參與小團體品牌社區的前因和購買後果。國際市場行銷研究雜誌 23，第 1 期：45-61。
Bickart, B., and R.M. Schindler. 2001. Internet forums as influential sources of consumer information. Journal of Interactive Marketing 15, no. 3: 31-40.
Bickart， B. 和 RM Schindler。2001. 互聯網論壇作為消費者資訊有影響力的來源。互動營銷雜誌 15，第 3 期：31-40。
Brewer, M.B. 1996. In-group favoritism: The subtle side of intergroup discrimination. In Codes of conduct: Behavioral research into business ethics, ed. D.M. Messick and A.E. Tenbrunsel, 160-71. New York: Russell Sage Foundation.
布魯爾，MB，1996 年。群體內偏袒：群體間歧視的微妙一面。在行為準則：商業道德的行為研究中，D.M. Messick 和 A.E. Tenbrunsel 編輯，160-71。紐約：羅素·塞奇基金會。

Brown, J., A.J. Brodcrick, and N. Lee. 2007. Word of mouth communication within online communities: Conceptualizing the online social network. Journal of Interactive Marketing 21, no. 3: 2-20.
布朗，J.，AJ 布羅德里克和 N. Lee。2007. 在線社區內的口碑傳播：概念化在線社交網路。互動營銷雜誌 21，第 3 期：2-20。
Carroli, B.A., and A.C. Ahuvia. 2006. Some antecedents and outcomes of brand love. Marketing Letters 17, no. 2: 79-89.
Carroli， BA 和 AC Ahuvia。2006. 品牌熱愛的一些前因和結果。行銷快報 17，第 2 期：79-89。
Chiu, H.-C., Y.-C. Hsieh, Y.-H. Kao, and M. Lee. 2007. The determinants of email reccivers’ disseminating behaviors on the Internet. Journal of Advertising Research 47, no, 4: 524-34.
邱，H.-C.，Y.-C.謝 Y.-H.Kao 和 M. Lee。2007. 電子郵件接收者在 Internet 上傳播行為的決定因素。廣告研究雜誌 47，第 4 期：524-34。
Chung, C.M.Y., and P.R. Darke. 2006. The consumer as advocate: Self-relevance, culture, and word-of-mouth. Marketing Letters 17, no. 4: 269-79.
Chung， CMY 和 PR Darke。2006. 作為宣導者的消費者：自我相關性、文化和口碑。行銷快報 17，第 4 期：269-79。
Crossman, A., and L. Lee-Kelley. 2004. Trust, commitment and team working: The paradox of virtual organizations. Global Networks 4, no. 4: 375-90.
Crossman， A. 和 L. Lee-Kelley。2004. 信任、承諾和團隊合作：虛擬組織的悖論。全球網路 4，第 4 期：375-90。
Dick, A.S., and K. Basu. 1994. Customer loyalty: Toward an integrated conceptual framework. Journal of the Academy of Marketing Science 22, no. 2: 99-113.
迪克，AS，和 K. Basu。1994. 客戶忠誠度：邁向集成的概念框架。營銷科學學院雜誌 22，第 2 期：99-113。
Dobele, A., D. Toleman, and M. Beverland. 2005. Controlled infection! Spreading the brand message through viral marketing. Business Horizons 48, no. 2: 143-9.
Dobele， A.、D. Toleman 和 M. Beverland。2005. 控制感染！通過病毒式營銷傳播品牌資訊。商業視野 48，第 2 期：143-9。
Dwyer, P. 2007. Measuring the value of electronic word of mouth and its impact in consumer communities. Journal of Interactive Marketing 21, no, 2: 63-79.
德懷爾，第 2007 頁。衡量電子口碑的價值及其對消費者社區的影響。互動營銷雜誌 21，第 2 期：63-79。
Dyer, E. 2006. Tripping the plastic nostalgic fantastic. New York Times, February 26.
戴爾，E. 2006 年。絆倒塑膠懷舊太棒了。紐約時報，2 月 26 日。
Ellemers, N., P. Kortekaas, and J.W. Ouwerkerk. 1999. Self-categorization, commitment to the group and group self-esteem as related but distinct aspects of social identity. European Journal of Social Psychology 29, nos 2-3: 371-89.
Ellemers， N.、P. Kortekaas 和 JW Ouwerkerk。1999. 自我分類、對群體的承諾和群體自尊作為社會身份的相關但不同的方面。歐洲社會心理學雜誌 29，第 2-3 期：371-89。
Escalas, J.E., and J.R. Bettman. 2005. Self-construal, reference groups, and brand meaning. Journal of Consumer Research 24, December: 378-89.
Escalas， J.E. 和 JR Bettman。2005. 自我建構、參考群體和品牌含義。消費者研究雜誌 24,12 月：378-89。

Fournier, S. 1998. Consumers and their brands: Developing relationship theory in consumer research. Journal of Consumer Research 24, March: 343-73.
富尼耶，S. 1998 年。消費者及其品牌：在消費者研究中發展關係理論。消費者研究雜誌 24,3 月：343-73。
Goldsmith, R.E. 2002. Explaining and predicting consumer intention to purchase over the Internet: An exploratory study. Journal of Marketing Theory and Practice 10, no. 2: 22-8.
戈德史密斯，RE 2002 年。解釋和預測消費者通過 Internet 購買的意願：一項探索性研究。營銷理論與實踐雜誌 10，第 2 期：22-8。
Goldsmith, R.E., and D. Horowitz. 2006. Measuring motivation for online opinion seeking. Journal of Interactive Advertising 6, no. 2: 1-16.
Goldsmith， RE 和 D. Horowitz。2006. 測量在線尋求意見的動機。互動廣告雜誌 6，第 2 期：1-16。
Goldsmith, R.E., B.A. Lafferty, and S.J. Newell. 2000. The impact of corporate credibility and celebrity credibility on consumer reaction to advertisements and brands. Journal of Advertising 29, no. 3: 43-54.
Goldsmith， R.E.， B.A. Lafferty 和 SJ Newell。2000. 企業信譽和名人信譽對消費者對廣告和品牌反應的影響。廣告雜誌 29，第 3 期：43-54。
Gounaris, S., and V. Stathakopoulos. 2004. Antecedents and consequences of brand loyalty: An empirical study. Journal of Brand Management 11, no. 4: 283-306.
Gounaris， S. 和 V. Stathakopoulos。2004. 品牌忠誠度的前因和後果：一項實證研究。品牌管理雜誌 11，第 4 期：283-306。
Granovetter, M.S. 1983. The strength of weak ties: A network theory revisited. Sociological Theory 1, no. 1: 201-33.
Granovetter，理學碩士 1983 年。弱關係的力量：重新審視網路理論。社會學理論 1，第 1 期：201-33。
Greenberg, P.S., R.H. Greenberg, and Y.L. Antonucci. 2007. Creating and sustaining trust in virtual team. Business Horizons 50, no. 4: 325-33.
格林伯格，PS、RH 格林伯格和 YL 安東努奇。2007. 在虛擬團隊中建立和維持信任。商業視野 50，第 4 期：325-33。
Hair, J.F., R.E. Anderson, R.L. Tatham, and W.C. Black. 1998. Multivariate data analysis with readings. 5th ed. Englewood Clifts, NJ: Prentice-Hall.
Hair， J.F.， R.E. Anderson， RL Tatham 和 WC Black。1998. 讀數的多變數數據分析。第 5 版恩格爾伍德·克利夫茨，新澤西州：Prentice-Hall。
Hennig-Thurau, T., K.P. Gwinner, G. Walsh, and D.D. Gremler. 2004. Electronic word-of-mouth via consumer-opinion platforms: What motives consumers to articulate themselves on the Internet? Journal of Interactive Marketing 18, no. 1: 38-52.
Hennig-Thurau， T.， K.P. Gwinner， G. Walsh， 和 D.D. Gremler。2004. 通過消費者意見平台的電子口碑：是什麼促使消費者在互聯網上表達自己？互動營銷雜誌 18，第 1 期：38-52。
Horrigan, J.B. 2001. Online communities: Networks that nurture long-distance relationships and local ties. http://www.pewinternet.org/pdfs/PIP_Communities_Report.pdf.
霍雷根，JB 2001 年。在線社區：培養異地戀和本地聯繫的網路。http://www.pewinternet.org/pdfs/PIP_Communities_Report.pdf。
Hung, K.H., and S.Y. Li. 2007. The influence of eWOM on virtual consumer communities: Social capital, consumer learning, and behavioral outcomes. Journal of Advertising Research 47, no. 4: 485-95.
Hung， K.H. 和 S.Y. Li. 2007 年。網路口碑對虛擬消費者社區的影響：社會資本、消費者學習和行為結果。廣告研究雜誌 47，第 4 期：485-95。
Jarvenpaa, S.L., K. Knoll, and D.E. Leidner. 1998. Is anybody out there? Antecedents of trust in global virtual teams. Journal of Management Information System 14, no. 4: 29-64.
Jarvenpaa， S.L.， K. Knoll， 和 DE Leidner。1998. 有人在嗎？全球虛擬團隊信任的前因。管理資訊系統雜誌 14，第 4 期：29-64。
Jarvenpaa, S.L., and D.E. Leidner. 1999. Communication and trust in global virtual teams. Organization Science 10, no. 6: 791-815.
Jarvenpaa， SL 和 DE Leidner。1999. 全球虛擬團隊中的溝通和信任。組織科學 10，第 6 期：791-815。
Joreskog, K.G., and D. Sorbon. 1993. Lisrel8: Structural equation modeling with the SIMPLIS command language. Chicago, IL: Scientific Software International.
Joreskog， K.G. 和 D. Sorbon。1993. Lisrel8：使用 SIMPLIS 命令語言進行結構方程建模。伊利諾伊州芝加哥：Scientific Software International。
Kramer, R.M. 1999. Trust and distrust in organizations: Emerging perspectives, enduring questions. Annual Review of Psychology 50: 569-98.
克萊默，RM 1999。組織中的信任與不信任：新興觀點，持久的問題。心理學年鑒 50：569-98。
Lea, M., and R. Spears. 1992. Paralanguage and social perception in computer-mediated communication. Journal of Organizational Computing 2, nos 3-4: 321-41.
Lea， M. 和 R. Spears。1992. 計算機介導通信中的副語言和社會感知。組織計算雜誌 2，第 3-4 期：321-41。
Levin, D.Z., and R. Cross. 2004. The strength of weak ties you can trust: The mediating role of trust in effective knowledge transfer. Management Science 50, no. 11: 1477-90.
Levin， D.Z. 和 R. Cross。2004. 您可以信任的弱紐帶的力量：信任在有效知識轉移中的仲介作用。管理科學 50，第 11 期：1477-90。
Lewis, J.D., and A. Weigert. 1985. Trust as a social reality. Social Forces 63, no. 4: 967-85.
路易斯，法學博士和 A. Weigert。1985. 信任作為一種社會現實。社會力量 63，第 4 期：967-85。
Mangold, W.G., and D.J. Faulds. 2009. Social media: The new hybrid element of the promotion mix. Business Horizons 52, no. 4: 357-65.
曼戈爾德，WG 和 DJ 福爾茲。2009. 社交媒體：促銷組合的新混合元素。商業視野 52，第 4 期：357-65。
Mayer, R.C., J.H. Davis, and F.D. Schoorman. 1995. An integrative model of organizational trust. Academy of Management Review 20, no. 3: 709-34.
Mayer， R.C.、J.H. Davis 和 FD Schoorman。1995. 組織信任的綜合模型。管理學會評論 20，第 3 期：709-34。
McAllister, D.J. 1995. Affect- and cognition-based trust as foundations for interpersonal cooperation in organizations. Academy of Management Journal 38, no. 1: 24-59.
麥卡利斯特，DJ 1995 年。基於情感和認知的信任作為組織中人際合作的基礎。管理學會雜誌 38，第 1 期：24-59。
Muniz, A.M., Jr, and T.C. O’Guinn. 2001. Brand community. Journal of Consumer Research 27, March: 412-32.
Muniz， A.M.， Jr 和 TC O'Guinn。2001. 品牌社區。消費者研究雜誌 27,3 月：412-32。
Peters, C., and C.R. Hollenbeck. 2005. A behavioural model of brand community member involvement: A conceptual framework and empirical results. Journal of Customer Behaviour 4, no. 3: 385-424.
彼得斯，C. 和 CR 霍倫貝克。2005. 品牌社區成員參與的行為模型：概念框架和實證結果。客戶行為雜誌 4，第 3 期：385-424。
Phelps, J.E., R. Lewis, L. Mobilio, D. Perry, and N. Raman. 2004. Viral marketing or electronic word-of-mouth advertising: Examining consumer responses and motivations to pass along email. Journal of Advertising Research 44, no. 4: 333-48.
菲爾普斯，J.E.，R. Lewis，L. Mobilio，D. Perry 和 N. Raman。2004. 病毒式行銷或電子口碑廣告：檢查消費者的反應和傳遞電子郵件的動機。廣告研究雜誌 44，第 4 期：333-48。
Plummer, J.T. 2007. Editorial: Word of mouth - a new advertising discipline? Journal of Advertising Research 47, no. 4: 385-6.
普盧默，JT 2007 年。社論：口碑 - 新的廣告學科？廣告研究雜誌 47，第 4 期：385-6。
Postmes, T., R. Spears, and M. Lea. 1998. Breaching or building social boundaries? Side-effects of computer-mediated communication. Communication Research 25, no. 6: 689-715.
Postmes， T.、R. Spears 和 M. Lea。1998 年。突破或建立社交界限？計算機介導通信的副作用。傳播研究 25，第 6 期：689-715。
PQMedia. 2007. Word-of-mouth marketing spending to break $\mathrm{\$}1$$\mathrm{\$}1$$1\$ 1 billion in 2007. http://www. pqmedia.com/about-press-20071115-wommf.html.
PQMedia.2007. 2007 年口碑行銷支出突破 $\mathrm{\$}1$$\mathrm{\$}1$$1\$ 1 10 億美元。http://www。pqmedia.com/about-press-20071115-wommf.html。