Using Trust and Willingness to Influence the Spread of Electronic Word-of-Mouth

Chiao Da "銜nagement ReνleW Vol. 32 No. 2. 2012 pp.105-J34

信任與溝通意願對網路口碑傳播之影

響

Using Trust and Willingness to Influence the Spread of

Electronic Word-of-Mouth

楊孝康 1 Conna Yang

國立交通大學經營管理研究所

Institute of 8usiness and Management, National Chiao Tung University 張力元 CharlesV. Trappey

國立交通大學管理科學系

Department ofManagement Science, National Chiao Tung University

摘要:現代消費者時常使用網絡為尋找商品資訊的管道 。 網路上的購物心得 或商品使用成想成了增加產品知名度的有效方式 。 網路口碑的重要性就在於 能夠打響一個產品的知名度並增加顧客忠誠度 。 這不但是最省錢， 也是最能 快速達到成效的一個手段 。 如今在網絡發達的社會，除了傳統的面對面的口 碑傳播，我們更想要了解網路口碑的影響與傳遞 。 本研究對於人與人之間分 享網絡口碑的原因作了一個探討，去測量個人的信任程度 ， 是否會影響他願 意去溝通以及對話的結果 。 另外，本研究也將關條強度當作一個重要的變數 納入研究範園 。在統計方法上，本研究採用結構方程模式針對 257 名受測者 所填寫的問卷進行分析 。 研究結果顯示，信任與網路口碑活動及正面網路口 碑有顯著的負向關條 。 又，溝通意願與網絡口碎的三個不同構念皆有顯著的 關條 。 最後，本研究在管理意泊上提出網路口碑行銷如何研究的勞現應用在 實務上，進而達到網絡口碑活動的綜效 o 關鍵詞:網路口碑;信任;溝通意願

I Corresponding author: Institute of Business and Management, National Chiao Tung Universi旬， Hsinchu City, Taiwan, E-rnail: conna.yang@gmail.com

106 Using Trusl and Wil/ingness 10 Injluence Ihe Spread 0/ Electronic Word-of-Moulh

Abstract : Modem consumers use the lntemet as a ubiquitous channel for seeking

infonnation and sharing experiences prior, duri嗯， and after purchasing.

Post-purchase statements made by electronic word-of-mouth have become a quick

and efficient means to increase product awareness among consumers. Using

electronic word-of-mouth is important for developing brands, increasing e-commerce website loyalty, and cuing consumer behavior. This research demonstrates how willingness to communicate and individual trust inf1uences electronic word-of-mouth. Tie strength is the construct used to measure interactions between trust and willingness among 257 survey respondents Structural equation modeling was applied to test tbe model and hypotheses. The research results support the hypothesis that there is a significant relationship between trust and electronic word of moutb. The results also show that

willingness to communicate has significant relationships with electronic word-of mouth-activity, positive electronic word-of-mouth, and negative electronic

word-of-mouth. Finally, the study presents the managerial implications of using

electronic word of mouth for lntemet marketing and maintaining consumer

loyalty

Keywords: Electronic Word-of-Mouth; Trust; Willingness to Communicate

1. Research Background

The Intemet provides a means of communication that substitutes traditional ways of promoting products, brands, and services. Passing infonnation and

sharing experiences through word-of-mouth (WOM) is well known as an efficient

means of spreading commercial messages. Bone (1995) notes a growing research interest in measuring WOM antecedents and effects. She assumed that

consumer's pre-usage attitudes toward a product are significantIy inf1uenced by WOM messages. WOM effects have become increasingly important with the emergence of the Intemet and frequently referred to as electronic word-of-mouth (eWOM) effects. Stauss (1997) first studied the topic and defined eWOM as

Intemet based customer communication that is categorized as a type of WOM behavior. A clear and specific definition was provided by Hennig-Thurau et al.

Chiao Da Managemenl Reνiew Vol. 32 No. 2. 2012 107

(2004) who used Stauss's discussion to define the term. Thus, electronic word-of-mouth (eWOM) is any post-purchase statement made by former customers about a product or company, which is available to a multitude of people and institutions via the Intemet.

The relationship and tie strength between the message spreader and message recipient has emerged as an important consumer research topic. Wirtz and Chew (2002) conducted a study on tie strength as an antecedent of WOM. Results showed a significant main effect between tie strength and WOM. A study by Van Hoye and Lievens (2007) pointed out that tie strength levels influence how the recipient perceives messages. WOM from a strongly tied person tends to be perceived as more reliable than messages from random sources. There are many questions to be answered and applications to be developed by applying tie-strength to electronic content. Marketing practitioners are particularly interested in how eWOM works and deriving new methods for determining behavioral antecedents

This research evaluates Intemet based eWOM communications and measures trust and willingness to communicate in relation to electronic messaging. 80th factors are hypothesized to influence consumer behaviors across a very broad range of commercial messages incIuding product Iikes and dislikes, favorite fashions

,

complaints

,

and future purchase intentions. Varying among industries

,

the strength of WOM is hypothesized to significantly influence the purchase intentions of consumers. The Intern巴tprovides a unique platform for measuring eWOM in comparison to traditional WOM (where information is passed between personal physical communication). Since electronic messages are spread instantaneously among a large and global network of people and are stored online

,

eWOM eliminates geographic distance, time barriers, and social groups. A new and boundary-free venue to express opinions, structure beliefs and attitudes, and often, experience pleasure by casually chatting with others on Intemet is also provided by eWOM. The eWOM activities are often multi-tasked with other online activities. Since consumers willingly share their personal views about products purchased using the Jntemet, willingness to communicate should influence the frequency to spread messages electronically. Consumers often

108 Using Trusl and Willingness 10 Injluence the Spread oJ Electronic Word可:Mouth

disclose infonnation about their shopping behavior if they trust the online media site or if they are communicating with friends in a familiar social network.

Therefore

,

trust should be relevant to the analysis of eWOM. This research examines whether willingness to communicate is a significant mediator between trust and eWOM to detennine if there is theoretical support for explaining and predicting the effects of eWOM

2. Literature Review

The literature review provides the foundation for the research questions and includes a review of the constructs used in previous studies. The constructs of this study build upon the focal aspects of each domain which includes electronic word-of-mouth, willingness to communicate

,

trust

,

and tie strength. The literature review defines these constructs and relates these terms to a research 企amework.

2.

1.

Electronic Word-of-Mouth (eWO

M)

Word-of-mouth (WOM) communications exist across different cultures and countries and facilitate infonnation flow (especially product and service related infonnation) between people. From a marketing and advertising perspective

,

WOM is one of the most important mechanisms used to influence the purchase intentions of consumers. WOM is thought to be an outcome of consumer experiences with products and services (Bu位le， 1998). Both positive and negative WOM significantly influence consumer decision making and is experienced at least daily in developed societies where choice prevails. For example， 合iends or family may provide infonnation that inf1uences a purchase decision or a consumer may take a stance toward a product

, company

or brand and aggressively spread the message through social networks. WOM is dynamic and unpredictable but there are several explanations for the phenomena. Hirschman (1970) was one of the first to explain that a disappointed consumer will typically react in one or a combination of three ways. The consumer might break the relationship, complain to others, or make their dissatisfaction known to the suppliers or company. Out of

Chiao Da Managemenf Review Vol. 32 No. 2, 2012 109

these three, complaining to others about an unsatisfied product experience, which is known in the literature as negative WOM (Buttle

,

1998)

,

has the greatest impact on a company's reputation and sales. A consumer's desire to purchase leads to a search of shared experiences which influences the outcome of the purchase (Harrison-Walker

,

2001). According to Flynn, Goldsmith, and Eastman

(1996)

, individuals usually keep 10% of their

good or bad emotional experiences to themselves whereas the m句 ority of the experience will be shared with others through conversations. After acquiring WOM information, consumers may exclude their own opinions and adopt tbe attitudes and opinions of others (Godes and Mayzlin

,

2004). The WOM effect is stronger when consumers are faced with ambiguous information while making purchase decisions (Bone

,

1995).

New media platforms such as the Intemet offer different ways to spread word-of-mouth (Table 1). The Intemet is the main po口al for eWOM and has greatly changed marketing communications. SimiLar to traditional WOM

,

eWOM has a higher reLiability for changing consumers' attitudes than other marketing activities on the Intemet (Bickart and SchindLer, 200 L). The Intemet provides different channels for consumers to communÌcate through social interfaces, such as e-maiL

,

blogs, forums, and websites. The traditional spread of WOM is limited by barriers such as distance, time, and cost of disseminating information (e.g., long distance phone caLls). The Intemet enabLes consumers to share their opinions and experiences with a multitude of other consumers and most importantly, the message is stored (Hennig-Thurau et al., 2004). Stored information on websites and blogs enables consumers to gather shared experiences and leam more about a product before making a purchase decision (Doh and Hwang, 2009)

For traditional WOM, the source of information is tangible and people communicate in person. For electronic word-of-mouth, there are more channels and formats for information communication. One of the most popular channels in Taiwan is the Bulletin Board System (BBS) where users exchange ideas and opinions (Wu and Lee, 1999). The most popular BBS in Taiwan is called PTT,

which bosts over 800

,

000 users per day. The BBS is a useful online network system that combines online chat, news

,

and discussion articles (Chen et al.,

110 Using Trust and Willingness 10 Injluence Ihe Spread 01 Electronic Word-of-Mouth

but grew in number and popularity and gradually started to include general segments ofthe population. Registered users often access different boards to share their purchase experience for products such as digital cameras, makeup, furnitu凹， books, and food

Weblogs are personal web pages that are frequently updated with new articles by bloggers (electronic message senders) who include photos

,

music

,

personal comments, and links to other Internet sites (Nardi, Schiano, and Gumbrecht, 2004). New forms of micro-blogging simpli即如此tions and make it easier for users to share their status, post short messages, and upload pictures through mobile phones or the Internet. Micro-blogs lower the required time to share information and increases the frequency of postings. A traditional weblog might add one or more articles per day, whereas micro-bloggers renew their status several times a day. ln addition, instant messaging is a type of technology on the Web that allows users to send and receive short text based messages and check the status of 企iendswho are online and available for conversations (Cameron and Webst缸， 2005)

Table 1

Traditional and Electronic Word-of-Mouth Different types of word-of-mouth

Traditional Word-of-Mouth

Elec甘onic Word-oιMouth

Communication platforms

Person to Person Contacts, Traditional Phone Bulletin Board System, Bloggin皂，Micro-blogging, Email, Websites, Intemet Phone

2.2.

WiUingness to Communicate (WTC)

Written and spoken words are the most important ways people communicate with each other. Each person has different tendencies to speak, write, or communicate with others. People speak when they are willing to communicate (McCroskey and Baer, 1985). Willingness to communicate is defined as a personality trait that underlies the communication process (MacIntyre

,

1994)

Chiao Da Management Review Vol. 32 No. 2. 2012 l l l

lndividuals that are highly willing to communicate will have enhanced listening comprehension and more frequent conversations. Also, people who are more willing to communicate are better liked by family and friends compared to those who are quiet and keep to themselves. The research of Rocca and Martin (1998) shows that this is especially prevalent in workplaces where those more willing to communicate are favored by coworkers and supervisors.

Willingness to communicate was first proposed by McCroskey and Baer in 1985 and as a means to measure the predisposition towards verbal communication They define this construct as the intention to initiate communication when given the oppo口unity. Willingness to communicate is also positively related to whether a person chooses to listen and interact with others. 1n Morgan and Miller's (2002) study conceming organ donations

,

willingness to communicate with family members about the subject was used to predict organ donation behavior. Different cultural environments influence the formation of commllnication behaviors (Barrac\ough, Christophel, and McCrosk句， 1988). Willingness to communicate is dependent on the situation and the mood of the listener to whom the speaker is speaking to. The most freqllently referenced scale used to measure willingness to communicate was developed by McCroskey and Richmond (1987). The scale measllres an individual's communication behavior under four situations and for three different types of listeners. Although other personality variables impact one's communication and speaking behavior, willingness to communicate is believed to be the focal factor. Communication is considered ubiquitous with the Intemet and individllals are now more willing to communicate with others, even strangers, lIsing the new electronic communication tools. For instance

,

the younger generation in particular appears to consider texting as important as talking with the prevailing attitude that is no longer necessary for most conversations to be face to face.

2.3. Trust

Trust defines whether one chooses to believe another's words or advice.

112 Using Trust and Willingness to 1'(月'uence the Sp間。d

01

Electronic Word-oj二Mouth

consumers or not. One ofthe c1assical definitions oftrust is that it is a generalized

expectancy he1d by an individual that the word of another can be relied on (Rotter

,

1971). When reliable information is consistently provided by others

,

our trust is

strengthened. The willingness to share information and the expectancies underlying the comrnunication are the foundation of tmst. The role of tmst is an important factor for understanding consurner behavior. Schurr and Ozanne (1985) describe tmst as the belief that one's word is reliable and that the individual will fulfill obligations in an exchange. A feeling of fear for negative outcomes or loss may occur if there is lack of tmst.

Tmst is a complex and multi-dimensional construct that has been studied by different disciplines inc1uding psychology and sociology. As long as risk and information sharing exists, the issue of tmst wiU exist because of its importance in interpersonal and business relationships. According to Rempel

,

Holmes and Zanna (1985)

,

there are four conc1usions that best define 肌s1. First

,

trust is bui1t

upon past experience and a series of events that strengthen the feeling. Second

,

dispositional behaviors are tests of 甘ustworthiness. The third conc1usion is that one must be willing to accept risk, whether it is the risk of being hurt, or the potential to lose present benefits for future gains. The last conc1usion is that trust can be measured by the degree of c10seness between individuals.

The method for modeling trust inc1udes how trust is defined

,

the origin of

tm哎， and why tmst changes over time (Chanley

,

Rudolph

,

and Rahn

,

2000;

Mayer

,

Davis

,

and Schoorman

,

1995; Sme1tzer

,

1997). Tmst begins at time zero when there are strangers and a lack of communications. Tmst experiments are used to explain and explore how initial tmst develops into re1ationships. Several studies show that trust deve10ps over time (Wilson

,

Straus

,

and McEvily

,

2006)

,

and when people evaluate the circumstances and the individuals before developing tmst (Salam

,

Iyer

, Palv間

，and Singh

,

2005). This research focuses on trust as defined by Rotter (1967) which is the expectancy held by an individual or

a party that the word, promise, verbal or written statement of another individual or party can be relied on. Tmst defmed in this manner focuses on the interaction between two individuals. Since the transfer of eWOM encompasses personal

Chiao Da Management Reνiew Vol. 32 No. 2, 2012 113

experience (Richins and Root品laf銜， 1988), it is inc1uded in the forrnation of the

research questions

2

.4. Ti

e Strength (TS)

Tie strength defines the social relationship of individuals and categorizes the

frequency of contact, how well others in the network are known, and the potency

of the bonds between members of a network (Granovetter

,

1973). Strong tie

sources are found among family and friends

,

wh巳reasweak ties are found among

s甘angers (Duhan, Johnson, Wi1cox, and Harrell, 1997). The influence of tie

strength can affect consumer decision making

,

WOM

,

and intergroup interactions

(Mittal, Huppertz, and Khare, 2008). WOM communication is a social behavior that includes the exchange of inforrnation between individuals. Previous research

shows that tie strength influences the flow of inforrnation and individuals who are

strongly tied have more opportunities to interact and share inforrnation (Gotlieb

,

Grewal, and Brown, 1994). On the con甘a旬， individuals in weakly tied relationships have few chances to speak directly. People with weak ties however

are more likely to build networks with other social groups to gain inforrnation

(Granovetter

,

1973)

Granovetter (1973) proposes a theory of strength-of-weak-ties, meaning that weak ties are more wide ranging and therefore more likely to act as

communication mediators than strong ties. This theory was initially used to study

employment in workplaces and show how weak ties assist individuals in the

process of job hunting where there is an overlap in inforrnation. Weak ties help individuals gain access to inforrnation that is beyond their social network,

whereas strong ties are useful for task-oriented needs (Haythomthwaite, 2001)

Drawing upon Granovett仗 's (1973) results on the strength of weak ties

,

Levin

and Cross (2004) produced research results showing that weak ties provide

non-redundant inforrnation. Furth缸，having strong bonds of trust towards another

individual allows the individuals to share complaints or negative comments since

they are confident the listener wiU not perceive the news in a bad way. Based on

these results, the research framework for this study proposes that strong ties have

114 Using Trus1 and Willingness 10 Influence 1he Spread 01 Electronic Word-o.fMouth

3.

Research Framework

ln this section, the proposed research framework is constructed based upon the results cited in the literature review. A total of ten hypotheses are proposed and tested in this study. The research framework defines the hypotheses and the

initial causal relationship model to be tested

3.

1.

Proposed Research Framework

The proposed research framework explores the interaction between

electronic word-of-mouth

,

willingness

,

trust

,

and tie strength. Past research supports relationships between these constructs. In Ben-Ner and Putterman's

(2009) study on trust and communication, the results show that if one desires to

establish a relationship

,

then individuals must take the communication process seriously. Thus, the depth of communication promotes trustworthy behaviors Anderson and Weitz (l989) have demonstrated the interaction between trust and cornmunication across different fields of study and state that communication is critical to building a trusting relationship which will create stability. Studies in cornmunication between siblings show that solidarity has a strong and positive

relationship with the depth of cornmunication (Rocca and Ma此in ， 1998). One of

the items on the solidarity scale of Rocca and Martin (l998)'s study measures

how much the other person is trusted. Siblings with high levels oftrust are willing to communicate across deeper and wider ranges of subj 恥的， they conclude that trust between individuals will also affect the depth of communication. However, few studies explore the relationship between the level 01' trust and the willingness

to communicate. Although the content of communication is not completely associated with personal privacy or other matters

,

it is reasonable to believe that the more one is willing to be exposed

,

the more willing they are to share

information with others. Therefore this sωdy proposes several related hypotheses

to be tested

Hl: Trust has a s智n伊'cant and positj}'e relationship with an

Chiao Da Management Review Vol. 32 No. 2. 2012 115

WOM is an initiative action where an individual chooses to express his

opinions or share past experiences with others. WOM is an action that depends on

the individual's intention to comrnunicate so an individual's willingness to communicate with another person should be positively related with it. As Richmond and Roach (1992) state, willingness to comrnunicate is the

communication personality construct that permeates every facet of an individual's

life. The following hypotheses are thus derived

H2a: Trust has a sign伊cant and positive relationship with an individual's electronic word-of-mouth activity.

H2b: Trust has a significant and positive relationship with an individual's positive electronic word-o

f-

mouth.

H2c: Trust has a sign伊'cantand positive relationship with an individual's negative electronic word-妒mouth.

H3a: An individual's willingness to communicate has a significant and positive relationship with electronic word-of-mouth activi秒﹒

H3b: An individual's willingness to communicate has a sign伊cantand positive relationship with positive electronic word-of-mouth. H3c: An individual's willingness ω communicate has a significant and

positive relationship with negative electronic word-of-mouth. WOM is a social behavior that involves comrnunicators with different levels

of tie strength (Wirtz and Chew, 2002). An individual might interact with those with whom they hold strong ties (e.g. family members) or weak ties (e.g. a person encountered on Intemet). The different levels of tie strength and the relationship to the spread of eWOM have been studied

,

yet few researchers have adapted

Granovetter's theory to explain the power ofweak ties and WOM. The strength of social relationships between individuals

,

or tie strength, also applies to the way the lntemet expands our social networks. When exploring the interaction between

tie strength and eWOM

,

Granovetter's (1973) research implies that there should be a negative relationship between tie s悅ngthand eWOM activities. Furthermore,

116 Using Trusl and Willingness 10 Injluence Ihe Spread

0

1

Electronic Word可:Mouth

individuals that are weakly tied usually use passive media to contact or share

infonnation with each other (Haythomthwaite

,

2001). Yet

,

since negative reviews

and complaints about a service or product might lead to negative personal images of the individual to the receiver

,

online users will likely be more willing to spread negative eWOM with strongly tied individuals whom they trust. In line with this reasoning it is hypothesized that

H4a: An individual's tie strength with another individual on the Jnternet has a significant and negative relationship with eWOM activity. H4b: An individual's tie strength with another individual on the Jnternet

has a sign伊cantand negative relationship with positive eWOM.

H4c: An individual's tie strength with another individual on the Jnternet has a sign研cantand positive relationship with negative eWOM.

Ten proposed hypotheses are derived from the literature review (Table 2). Trust has a positive effect on willingness to comrnunicate and willingness to

comrnunicate acts as the mediator between trust and electronic word-of-mouth.

Finally, tie strength has a negative and significant effect on electronic

word-of-mouth. These foundation hypotheses are used to build the fundamental

model constructed for testing how trust and willingness to comrnunicate

influences the spread of electronic word-of-mouth (Figure 1).

4. Methodology

In the previous section, the research framework and proposed hypotheses

were derived from the literature review. This section focuses on the methodology for model construction and hypotheses testing. The scales used for measurement are adapted from previous research to provide a reliable means to evaluate the constructs of interest. Two pretests were conducted before distributing the final questlOnnalre to partlclpants

Chiao Da Management Review Vol. 32 No. 2, 2012 117 Proposed Hypothesis Hypothesis 1 Hypothesis 2a Hypothesis 2b Hypothesis 2c Hypothesis 3a Hypothesis 3b Hypothesis 3c Hypothesis 4a Hypothesis 4b Hypothesis 4c Table 2 Proposed Hypotheses Content of Hypothesis

Trust has a significant and positive relationship with an individual's willingness to communicate

Trust has a significant and positive relationship with an individual's

eWOM activity

Trust has a significant and positive relationship with 組 individual's positive eWOM

Trust has a significant and positive relationship with an individual's negative eWOM

An individual's willingness to communicate has a significant and positive relationship with eWOM activity

An individual's willingness to communicate has a signi日cant and positive relationship with positive eWOM

An individual's willingness to communicate has a significant and positive relationship with negative eWOM

An individual's tie strength with another individual on the lntemet has

a Sl伊 1日 cantand negative relationship with eWOM activity

An individual's tie strength with another individual on the lntemet has a significant and negative relationship with positive eWOM

An individual's tie strength with another individual on the lntemet has a significant and positive relationship with negative eWOM

4.

1.

Operational Definitions

lmportant works related to this study have provided the constructs from which the framework and hypotheses are derived. For example

,

MacIntyre et al. (2001)'s study on wil\ingness to communicate and Rotter's (1971) generalized trust scale are key references for questionnaire construction. The key operational definitions ofthe constructs are cited in Table 3

118 Using Trust and Willingness ω Injluence the Spread 01 Electronic Word-of-Mouth

Figure 1

Proposed Research Framework

H3a+ H3b十 H3c+ H2b+ H2a+ 4.2. Data Collection

Considering that the Intemet usage rate is highest among students and teachers, the population of this study targets a sample of respondents in Taip剖，

Taiwan. Questionnaires were handed out in campus classrooms (including 5 national universities and 3 private universities) with the help of friends and postings on PTT, the largest online chat forum in Taiwan. All items utilized a five-point Likert scale ranging from 1 being

“

strongly disagree" and 5 being

“strongly agree." All items were designed in English and then translated into traditional Chinese by a native speaker fluent in both languages to ensure the content and meanings were consistent. Back translation was provided by another bilingual speaker who compared the translation to the original scale.

Chiao Da Managemenl Reνiew Vol. 32 No. 2, 2012 119

Table 3

Operational Definition of Constructs

Construct Operational Definition Key References Trust The degreeωwhich an Gurtman (1992);

individual believes that the Rotter (1971)

smωn旬 benevolenc唔 or

truthfulness of others can generally be relied on

Willingness to Communicate The level of intention an Maclntyre el a1. (2001);

Tie Strength

Electronic Word-o 仁 Mouth Activity

Positive Electronic Word-of-Mouth Negative Electronic Word-of-Mouth

individual has to initiate a McCroskey and Richmond conversation when given a (1987)

choice

The c10seness of the Duhan el al. (1997);

relationship between two Granovetter (1973)

individuals

The quantity and frequency of Harrison-Walker (2001)

inforrnation the sender provides during the electronic

word-。仁 mouth

commuDlcatlon process

Positive statements of praise Cheung el al. (2008);

made by a customer about a Hennig-Thurau el al. (2004)

product or brand

Negative statements or Cheung el al. (2008);

complaints made by a Hennig-Thurau et al. (2004)

customer about a product or brand

The official questionnaire is divided into 5 sections: electronic word-of-mouth, trust, willingness to communicate, tie strength, and demographic variables. Electronic word-of-mouth activities were measured using Churchil\'s

(\979) word-of-mouth scale which was modified by Harrison-Walker in 2001 Positive electronic word-of-mouth was measured using 3 items from the loyalty scale and electronic negative word-of-mouth used the scale 什om Liu and McClure's (2001) research on cross-cultural customer complaint behavior. To improve the efficiency and increase the unity of the questionnaire

,

scales were modified from 7 points to 5 points. Rot1er's Trust scale

,

which was developed in

120 Using Trusl and Willingness 10 lnjluence Ihe司'pread 0/ Electronic Word-oj二Moulh

scale to measure willingness to communicate as a personality-based trait that is consistent across different receivers and different cornmunication contexts. This scale uses 20 items

,

with 8 items acting as fillers. The willingness to cornmunicate scale includes four cornmunication contexts and 也ree different receivers. The study measured each individual's general willingness to cornmunicate. Hansen's 1999 two item scale measured the closeness of working relationships and the frequency of contacts. These measures serve as a measure of tie strength

4.3. Pretest

Two pretests were conducted before the official questionnaire was handed out to participants. Our study tested the reliability of the scales using a standard Cronbach alpha of 0.70. In the first pretest

,

many participants responded that the items of the trust scale were too hard to understand and were confusing. Therefore

,

the internal consistency received a fairly low score ofO.59. The original electronic

word-of-mouth scale consisted of 2 dimensions which are word-of-mouth activity

and positive word-of-mouth. Along with the replacement of a new positive

word-of-mouth scale that consisted of 3 items, 2 items measuring negative

word-of-mouth were added to strengthen the results of the study. Willingness to

communicate achieved a high alpha ofO.80 and tie strength's alpha was 0.91.

For the second pretest, the revised questionnaire was used. Demographic questions were shown to be too sensitive and decreased the retum rate. Therefore, the demographic variables were moved to the end of the questionnaire. Second

,

elec仕onicword-of-mouth was rated with a total of 8 items incIuding two newly

added scales measuring negative electronic word舟f-mouth. Furthermore,

additional directions were added to each section to help participants navigate the questionnaire and mark responses to the questions. The resuIts for the second pretest passed the bar of 0.70 for internal consistency except for trust. Four items that were negatively correlated with the scale were removed

,

which resulted in a Cronbach's alpha ofO.71. The final questionnaire distributed to the target sample had a total of 48 questions (Table 4).

Chiao Da Managemenl Review Vol. 32 No. 2, 2012 121 Construct Trust Trust 1 Trust 2 Trust 3 Trust 4 Trust 5 Trust 6 Trust 7 Trust 8 Trust 9 Trust 10 Trust 11 Trust 12 Trust 13 Trust 14 Trust 15 Trust 16 Trust 17 Trust 18 Trust 19 Trust 20 Willingness to WTCl Communicate (WTC) WTC2 WTC3 Table 4

Measurement Items of Constructs Measurement Item Hypocrisy is on the increase in our society

In dealing with strangers one is better off to be cautious until they

have provided evidence that they are trustworthy

This countJγhas a dark future unless we can attract be仕:erpeople into politics

U sing the honor syste咽。fnol having a teacher present during exams would probably result in increased cheating

Parents usually can be relied upon to keep their promises The United Nation will never be an effective force in keeping world peaCe

The judiciary is a place where we can all get unbiased treatment

Most people would be horrified if they knew how rnuch news the public hears and sees is distorted

Even though we have reports in newspapers, radio, and T.V., it is

hardωget objective accounts of public events

The future seems very promising

If we really knew what was going on in intemational politics, the public would have reason to be more frightened than they now

seem to be

Most elected public officials are really sincere in

their campaign prom凹的

Many major national sports contests are fixed in one

way or another

Most experts can be relied upon to tell the truth about

the limits oftheir knowledge

ln these ∞mpetitive times one has to be alert or someone is likely to take advantage of you

Most salesmen are honest in describing their products

Most students in school would not cheat even ifthe于

were sure of getting away with it

Most repairmen will not overcharge even ifthey think

you are ignorant of their specialty

Large shares of accident claims filed against insurance companies are phony

Most people answer public opinion polls honestly Talk with a train station attendant

Talk with a physician

122 Using Trust and WilIingness to lnjluence the 鈔read

01

Electronic Word-of-Mouth Tie Strength (TS) Elec甘omc Word-of-Mouth Activity (eWOM Activity) Positive Electronic Word-。ιMouth (Positive eWOM) Negative Electronic Word-。ιMouth (Negative eWOM)

WTC 4 Talk with an acquaintance while standing in line WTC 5 Talk with a salesperson in a store

WTC 6 Talk in a large meeting offriends WTC 7 Talk with a police officer

WTC 8 Talk in a small group of strangers

WTC 9 Talk with a friend while standing in line WTC 10 Talk with a waiter/waitress in a restaurant

WTC II Talk in a large meeting of acquaintances WTC 12 Talk with a s仕angerwhile standing in line WTC 13 Talk with a secretary

WTC 14 Present a talk to a group of friends

WTC 15 Talk in a small group of acquaintances WTC 16 Talk with a garbage collector WTC 17 Talk in a large meeting of strangers

WTC 18 Talk with a spouse (or girllboy 企iend) WTC 19 Talk in a small group of 企iends WTC20 TS 1 TS 2 eWOM Activity 1 eWOM Activity 2 eWOM Activity 3 eWOM 主旦旦立主 Positive eWOMI Positive eWOM2 Negative eWOMI Negative eWOM2

Present a talk to a group of acquamtances

How close were you with the person you most re心ently sent an

electronic message to?

How often did you communicate with that person?

1 mention the products 也 at1 use to others on the lnternet quite

frequently

There are certain products that 1 talk more about 0叫 theInternet than in person to others

When [ use the Internet, 1 seldom miss the opportunity to talk about a certain product

1 will use the lnternet to recommend a product 1 like

On the Internet, 1 will give positive reviews about a product 1 like On 也eInternet, 1 will encourage others to buy or use products that I like

On the Internet, 1 will share my experience about products that 1 do not like

On the Intern帥，1 wilJ try to convince others not to use or purchase produ峙的 1do not like

Chiao Da Managemenl Review Vo/. 32 No. 2, 2012 123

5.

Research Results and Data Analysis

A total of 257 valid questionnaires were retrieved. Using AMOS version 17

software (Arbuckle

,

2008)

,

statistical analysis was conducted to examine and test the hypotheses. Structural equation modeling was used to examine the model

goodness of fit and to test the hypotheses

5.

1.

Sample Description

The respondents included 113 males and 144 females. The respondents were

mostly young people in the age group of 19 to 23 (47%) and 24 to 28(41.63%) which matches the goal of the study to survey students and teachers at 8

universities that host the original comrnunity of web-based communications.

5.2.

Structural Equation Modeling for Data Analysis

Structural Equation Modeling (SEM) was used to test the model and

hypotheses derived from the Iiterature. SEM is a powerful second generation statistical analysis method used for analyzing and cr巴atingcausal models. This two step approach first tests the measurement model, then the structural model. The aim of the measurement model is to evaluate the reliability, validity, and goodness of fit. Confirmatory Factor Analysis (CF A) relies upon past studies which introduced factor analysis as an extension of maximum likelihood. CF A was performed and the results were used for further analysis (Table 5). Several

items were removed since they did not meet the 0.5 criteria for reliability. Twelve

items of the trust scale did not meet the loading standard criteria and were

removed. Six items of the willingness to communicate scale scored low in the CF A. A total of 20 items were removed before conducting the measurement model analysis. The trust scale yielded a Cronbachαof 0.881, scoring much higher than the 2 pretests conducted. Furthermore

,

the 12 items for willingness to

comrnunicate had an intemal consistency of 0.797, while tie strength had the highest intemal consistency with an alpha of 0.893

124 的 ing Trusl and Willingness (0 lnßuence the Spread 01 Electronic Word-of-Mouth

Table 5

Results of CF A and Validity After Removing U nsuitable Items

Item Factor Cronbach Composite Average Variance

Loading Alpha Reliability Extracted

Trust I 501 。 251 Trust 2 870 。 757 Trust 3 729 0.532 Trust 6 501 。 251 Trust 9 831 0.691 Trust I1 688 。 474 Trust 15 788 0.621 0.8755 。 5107 WTC'6 546 。 298 WTC I1 624 0.389 WTC 14 611 0.373 WTC 15 578 0.460 WTC 19 507 0.257 WTC20 624 0.389 。 7708 0.3612 Tie 1 g85 0.783 Tie2 子 11 0.830 0.8929 0.8066 eWOMact. 1 806 。 650 eWOM act. 2 g33 0.694 eWOM act. 3 537 0.406 eWOMact.4 677 0.458 。 8295 0.5519 Pos. eWOM 1 787 0.619 Pos. eWOM 2 741 0.549 。 7374 0.5842

Neg. eWOM 1 glO 0.656

Ne昌 eWOM2 829 0.687 0.8036 。 6717

Note : 'WTC is the abbreviation for W ilIing且essto Communicate

Table 6

Discriminant Validity Comparing SQRT A VEs and Correlation Coefficients

Tie eWOM Positive Negative

strength Activity eWOM eWOM

Trust WTC Trust 0.715 WTC' 0.030 。.601 Tie -0.004 0.096 0.898 Strength eWOM -0.110 0 .3 03 艸 _{。 239** 0.743} Activity Positive -0.128* 。 282** 0.198** 0.625** 0.764 eWOM Negative -0.057 0.199** 。 232** 0.455** 0.444** 0.820 eWOM

Note : ** Correlation is significant at the 0.01 level * Coπelation is significant at the 0.05 level

ChiaoDa J.伽nagementReνiew Vol. 32 No. 2, 2012 125

Table 7

Fit Indices for Measurement and Structural Model

χ2Id.f. GFI AGFI PGFI TLI CFI RMSEA

Suggested Criteria < 3 0.9 0.8 > 0.5 > 0.9 > 0.9 < 0.08 Measurement Modcl Results 1.668 0.895 0.865 0.697 0.934 0.944 0.051 Structural Model Results 2.278 0.856 0.818 0.679 0.874 0.891 0.070 Note : Italicized figures are indices that do not meet the suggested criteria

E.e -E ES VME a n A 8l

eu

喇

h Td e3 4E ‘ l u es ρtw R Hypothesis Path HI Trust-州 TC'

H2a Trust• eWOM Activity H2b Trust• Positive eWOM H2c Trust• Negative eWOM

H3a WTC• eWOM Activ旬

H3b WTC• Positive eWOM H3c WTC• Negative eWOM

H4a Tie Strengtb-+eWOM Activity H4b Tie Strength-•Positive eWOM

H4c Tie Strengtb• Negative eWOM

Coefficients S.E. P-value 0.036 0.065 0.619 -0.133 0.056 0.047* -0.155 0.065 0.034* -0.085 0.082 0.194 0.347 0.080 0.000*** 0.419 0.089 0.000*** 0.221 0.105 0.003** 0.264 0.039 0.000*** 0.225 0.045 0.005** 0.270 0.054 0.000*** Note : *P-value < 0.05. **P-value < 0.01. ***P-value < 0.001

'WTC is the abbreviation for Wi1lingness to Communicate 'WTC is the abbreviation for WiUingness to Communicate

Discriminant analysis was used to measure the difference between constructs (Kerlinger and Lee, 2000; Premkumar and Roberts, 1999). The results for the

discriminant validity analysis are shown in Table 6. A verage variance extracted (A VE) represents the amount of variance in the observed items accounted for by

126 Using Trust and Wi/lingness 10 lnßuence the Spread 01 Electronic Word-of-Mouth the unobserved variable. The square root of the A VE of each construct should be higher than each of the correlations. The results provide support that the

constructs are distinct. After conducting the reliability and validity tests, the measurement model was tested to examine whether the model was a good fit for the sample data in Table 7. Several common indices were applied and the

following results were noted: the chi-square/degrees of freedom (X2/叫f= 1.668)

,

the goodness of fit index (GFI = 0.895), the adjusted goodness of fit index (AGFI = 0.865)

,

the parsimony goodness of fit index (PGFI = 0.697)

,

the Tucker-Lewis index (TLI = 0.934), the comparative fit index (CFI = 0.944), and the root mean

square error of approximation (RMSEA = 0.051). The goodness of fit index was

be\ow the suggested criteria of 0.05. However, no further modification or changes were made to improve the goodness of fit since changes would be based on statistical data instead of theoretical data. Research should be based on theoretical

studies and past literature to ensure trustworthy and concrete results. Therefore,

we continued with the second part of structural equation modeling which is the

testing of the hypotheses

The test of the structural model determines whether there are significant relationships in the proposed causal model. The goodness-of-fit indices in Table 7 for the structural model show a slight lack of fit. Table 8 and Figure 2 depict the critical results of this study. Trust has a significant and negative effect on both

eWOM activity 伊= -0.1 刃，p < 0.05) and positive eWOM (戶=-0.155

,

p < 0.05)

which partially supports Hypotheses 2a and 2b. Willingness to communicate has a significant and positive influence on eWOM activity 伊= 0.347, p < 0.001),

positive eWOM 伊=0.419,p < 0.001), and negative eWOM 伊=0.221 ，p < 0.01)

Therefore Hypotheses 3a, 3b, and 3c are supported. Tie strength has a positive

and significant effect on all three eWOM related variables 伊=0.264, p < 0.001, β = 0.225, P < 0.01,

ß

= 0.270, P < 0.001), partially supporting H4a and 4b. As expected, Hypothesis 4c is fully supported. Unexpectedly, Hl was rejected, since no significant relationship between trust and willingness to communicate was found. Likewise, H2c was also rejected since path analysis showed no significant

Chiao Da Management Reνiew Vo/. 32 No. 2, 2012 127

Figure 2

Results of Research Framework

Tie Strength 0.347***+ 0.036 Negative eWOM -0.085 。.155*一 -0.133*

6. Conclusion and Implications

This research examines whether willingness to communicate acts as a

mediator between trust and electronic word-of-mouth. There is a significant

relationship between several variables in the proposed model. Trust has a negative

and significant effect on eWOM activity and positive eWOM. Willingness to

communicate is proven to affect the consumer's eWOM communications in all

three dimensions which are eWOM activi旬， positive eWOM, and negative

eWOM. lndividuals that one trusts are usually friends or family and can be easily

contacted. This study also demonstrates from its results that the Intemet is not the

main communication platforrn for individuals who know each other personally

and communicate frequently face-to-face. Traditional WOM and other forrns of

communication that inc\ude phone calls or personal visits are more direct. Intemet

128 Using Trust and Willingness 10 lnjluence Ihe Spread 01 Electronic Word-oj二Mouth

with. Giving positive comments and praising products on the Internet yields a desired impression and helps build a positive personal image. On the contrary

,

consurners might consider that they are creating a negative self image if they write negative reviews and complaints. The level of dissatisfaction each consumer faces varies. It is likely that many of the electronic message recipients have not had the exp閃閃閃 thatcaused and stimulated an urge for the messenger to share this inforrnation with others.

Tie strength is shown to have a significant and positive relationship with negative and positive eWOM and eWOM activity. This means that the strength of the social relationship has a direct relationship with the level of communications sharing. The stronger the tie strength between individuals

,

the more likely they will propagate eWOM on lnternet platforms. The outcome oftie strength and trust with eWOM are unexpected since the directions of influence are different from results in the literature. These results can be used as the basis for further research. First

,

the personality of each individual may moderate the amount of usage of Intemet (Guadagno, Okd吟， and Eno, 2008). Guada厚的， Okdie, and Eno (2008)

used the Big Five personality measure scale to predict personality differences and Internet usage. lndividuals that are high in openness to experience are more willing to share details of their daily Iife. Also

,

Swickert et al. (2002)'s study on lntemet use and personality showed that personality is highly correlated with electronic information exchange (email and assessing inforrnation). Thus

,

Internet users do not have to trust the communicator to be more willi月 tospread eWOM.

6.

1.

Managerial Implications

This research provides several insights on how marketers can use eWOM as part of their strategy to attract and hold consumers. Certainly

,

eWOM should not be generalized to a single target audience. When targeting a mass target market

,

the sample should be divided among consurners with strong and weak tie strengths. Since there is no signi日cantrelationship between tie strength and trust

,

it should be further studied how individuals discriminate between the lnternet users they interact with. If we understand how consurners choose others to share their eWOM, then it will be possible to develop marketing tactics which

Chiao Da Management Review Vol. 32 No. 2, 2012 129

encourage communication. Electronic messages are a new tool for the advertising

industry. Media providers freely interact with consumers and advertise through

the Internet which shortens the distance between consurners and corporations and builds a larger target audience (Hung and Li

,

2007).

6.2.

Li

mitations and Future Research

Although this study provides useful practical insights and a basis for theory

developme肘， there are limitations. First, the sample consists of colJege students and teachers in Taipei City. Since the sample reflects a smalJ portion of our

society

,

the diversity of the sample should be increased and additional hypotheses

created. For future research, different groups of lntemet users can be examined

for differences in electronic word-of-mouth sharing. Since lnternet usage behavior (e.g.

, the

amount oftime spent on the Web) influences the interest in the lntemet

(Awad and Ragowsky, 2008), usage time wilJ likely affect eWOM. Hence, online

behaviors can be examined to observe the difference between different groups of lntemet users. FinalJy

,

it would be worthwhile to examine if there are more

constructs (for example

, personality

,

and mood) that wiU effectively complete the

model.

7.

References

Anderson, E. and Wei紹，B. (1989),“Determinants of Continuity in Conventional

Industrial Channel Dyads," Marketing Science, 8(4),310-323.

Arbuckle, J. L. (2008), Amos 17 User/s Guide, Chicago, IL: SPSS

Awad, N. F. and Ragows旬， A. (2008), “Establishing Trust in Electronic

Commerce Through Online Word of Mouth: An Examination Across

Genders," Journal

01

Management lnformation 砂'stems， 24(4), 101-121

Barraclough, R. A., Christophel, D. M. andand McCroskey, J. C. (J 988),

“Willingness to Communicate: A Cross-Cultural Investigation

,

"

Communication Research Reports

,

5(2)

,

187-192.

Ben-Ner, A. and Putterman, L. (2009), “Trust, Communication and Contracts:

130 Using Trust and Wil/ingness 的 lnjluence the Spread 01 Electronic Word-of-Mouth

106-121

Bicka此， B. and Schindler, R. (2001), “lntemet Forums as Influential Sources of Consumer Information," Journal of lnteractive Marketing, 15(3), 31-40 Bone, P. (1995), “Word-of-Mouth Effects on Short-Term and Long-Term

Product Judgments," Journal of Business Research, 32(3),213-223

Buttle, F. (1998), “Word of Mouth: Understanding and Managing Referral Marketing," Journal of Strategic Marketing, 6(3), 241-254

Cameron, A. and Webster, 1. (2005), “Unintended Consequences of Emerging Communication Technologies: Instant Messaging in the Workplace

,"

Computers in Human Behavior, 21 (1), 85-103

Chanley, V. A., Rudolph, T. 1. and Rahn, W. M. (2000), “The Origins and Consequences of Public Trust in Govemment: A Time Series Analysis

,"

Public Opinion Quarter鈔， 64(3),239.

Chen, H. S., Guo, F. R., Liu, C. T., Lee, Y. 1., Chen, 1. H., Lin, C. c., Hou, S. M. and Hsieh, B.S. (1998),“Integrated Medical Informatics With Small Group Teaching in Medical Education " lnternational Journal of Medical lnformati白， 50(1-3)，59-68.

Cheung, C. M. K., Lee, M. K. O. and Rabjohn, N. (2008), "The Impact of Electronic Word-of-Mouth: The Adoption of Online Opinions in Online

Customer Communities," lnternet Research, 18(3),229-247.

Churchill, G. A. (1979), “A Paradigm for Developing Better Measures of Marketing Constructs," Journal ofMarketing Research, 16(1),64-73. Doh, S. and Hwang, J. (2009),“How Consumers Evaluate Ewom (Electronic

Word-of-Mouth) Messages，"。缸中'sychology and Behavior, 12(2),

193-197

Duhan， 眩， Johnson, S., Wilcox, 1. and Harrell, G. (1997), “Influences on Consumer Use of Word-of-Mouth Recommendation Sources

,"

Journal of the Academy of Marketing Science, 25(4),283-295

Flynn, L., Goldsmith, R. andand Eastman, J. (1996), “Opinion Leaders and Opinion Seekers: Two New Measurement Scales

,"

Journal ofthe Academy of Marketing Science

,

24(2)

,

137-147

Chiao Da Management Reνiew Vol. 32 No. 2, 2012

Word-of-Mouth Communication," Marketing Science, 23(4),545-560

Gotli功， 1. Grewal ，眩， andand Brown, S. (1994),

131

Perceived Quality: Complementary or Divergent Cons甘ucts?" Journal 01

Applied Psychology, 79(6),875

Granove悅r， M. (1973), “The Strength of Weak Ties," American Journal 01

Sociology, 78(6), 1360-1380.

Guadagno, R. E., Okdie, B. M. andand Eno, C. A. (2008),“Who Blogs?

Personality Predictors ofBlogging

,"

Computers in Human Behavior, 24(5)

,

1993-2004.

Gurtman, M. B. (1 992) ，“Tru哎， Distru哎， and Interpersonal Problems: A

Circumplex Analysi丸" Journal 01 Personality and Social Psychology,

62(6)

,

989

Hansen, M. T. (1999),“The Search-Transfer Problem: The Role of Weak Ties in

Sharing Knowledge Across Organization Subunits

," Administrative Science

Quarter鈔， 44(1) ， 82-111

Harrison-Walker, L. 1. (2001), “The Measurement of Word-of-Mouth

Communication and an Investigation of Service Quality and Customer

Commitment as Potential Antecedents," Journal 01 Service Research, 4(1),

60

Haythomthwaite, C. (2001),“The Strength and the Impact of New Media ," Proceedings 01 the 34th Hawaii lnternational Conference on System Sciences

,

Los Alamitos

,

CA: IEEE Computer Sowety

Hennig-Thurau, T., Gwinner, K., Walsh, G. and Gremler, D. (2004),“Electronic Word of Mouth Via Consumer Opinion Platforrns 叭fhat Motivates

Consumers to Articulate Themselves on the lntemet?" Journal oj

lnteractive Marketi嗯， 18(1), 38-52.

Hirschman, A. O. (1970), Exit, Voice, and Loyalty: Responses to Decline in Firms,

Organizations, and Stat釘， Cambridge, MA: Harvard University Press

Hung, K. and Li, S. (2007),“The lnfluence of EWOM on Virtual Consumer

Communities: Social Capital, Consumer Leaming, and Behavioral

Outcomes," Journal of Advertising Research, 47(4),485-495

132 Using Trust and WilIingness 10 /，圳uence the Spread 0/ Electronic Word-of-Mouth

Fort Worth, TX: Wadsworth/Thomson Learning.

Levin, D. and Cross, R. (2004),“The Strength of Weak Ties Y ou Can Trust: The Mediating Role of Trust in Effective Knowledge Transfer," Management Science, 50(11),1477-1490

Liu, R. R. and McClure, P. (2001),“Recognizing Cross-Cultural Differences in

Consumer Complaint Behavior and Intentions: An Empirical Examination,"

Journal ofConsumer Marketing, 18(1),54-75

MacIn可re， P. D. (1994),“Variables Underlying Willingness to Comrnunicate: A Causal Analysis," Communication Research Reports, 11(2), 135-142. MacInty凹，P. 眩， Baker, S. C., Clément, R. and Conrod, S. (2001),“Willingness

to Comrnunicate, Social Support, and Language-Leaming Orientations of Immersion Students," Studies in Second Language Acquisition, 23(3),

369-388.

Mayer, R. c., Davis, J. H. and Schoorman, F. D. (1995),“An Integrative Model of Organizational Trus丸" The Academy of Management Review

,

20(3)

,

709-734.

McCroskey, J. and Baer, 1. (1985), “Willingness to Comrnunicate: The Construct and Its Measurement," Paper presented at the 71st annual meeting of the ♀1eech Communication Association

,

Denver

,

CO.

McCrosk句， J. C. and Richmond, V. P. (1987),“Willingness to Comrnunicate,"

in 1. C. McCroskey, and J. A. Daly (Eds.), Personality and Jnterpersonal Communication (129-156)

,

Newbury Park

,

CA: Sage.

Mi叫，以，Huppe巾，1. and Khare, A. (2008), "Customer Complaining: The Role of Tie S仕ength and Information Control," Journal of Retaili嗯， 84(2), 195-204.

Morgan, S. E. and Miller, 1. K. (2002), “Beyond the Organ Donor Card: The Effect of Knowledge, Attitudes

,

and Values on Willingness to Comrnunicate about Organ Donation to Family Members," Health

Communication

,

14(1), 121-134.

Nardi, B. A., Schiano, D. 1., and Gumbrecht, M. (2004), “Blogging as Social Activity, or, Would You Let 900 Million People Read Your Diary少 in Proceedings of the 2004 ACM Conference on Computer Supported

Chiao Da Management Reνiew Vol. 32 No. 2. 2012 133

Cooperative Work (222-231), New York, NY: ACM.

Premkumar, G. and Roberts, M. (1999), “Adoption of New Information

Technologies in Rural Smal\ Businesses," Omega, 27(4),467-484

Rempel, J. K., Holmes, J. G. and Zanna, M. P. (1985), “Trust in Close

Relationship丸" Journal of Personality and Social Psychology, 49(1), 95-112

Richins, M. L and Root-Shaffer, T. (1988), “The Role of Evolvement and

Opinion Leadership in Consumer Word-of-Mouth: An Implicit Model Made Explicit," Advances in Consumer Research， 的(1)， 32-36

Richmond, V. P. and Roach, K. D. (1992), “Willingness to Communicate and Employee Success in US Organizations

,

"

Journal of Applied Communication Research, 20(1), 95-115

Rocca, K. A. and Martin, M. M. (1998),“The Relationship Between Wil\ingness

to Communicate and Solidarity with Frequency, Breadth, and Depth of Communication in the Sibling Relationship," Communication Research Reports, 15( 1), 82-90

Rotter, J. (1967),“A New Scale for the Measurement of Interpersonal Trust,"

Journal of Personali紗， 35(4), 651-665

Rotter, J. (1971),“Generalized Expectancies for Interpersonal Trust," American Psychologist, 26(5),443-452

Salam, A., Iyer, L., Palvia, P. and Singh, R. (ρ20∞05)，

Commη1U仰Jn Z眩caωtiωon肘?彷s oft.的h惚eAC

M,

48(2), 72-77.

Schurr, P. and Ozann巴， J. (1985),“Influences on Exchange Processes: Buyers'

Preconceptions of a SelIer's Trustworthiness and Bargaining Toughness," The Journal ofConsumer Research, 11(4),939-953

Smeltzer, L. R. (1997),“The Meaning and Origin of Trust in Buyer Supplier Relationships," Journal ofSupply Chain Management, 33(1),40-48.

Stauss, B. (1997),“G lobal Word of Mouth," Marketing Management, 6(3), 28-30

Swicke泣， R. J., Hittner, J. 8., Harris, J. L. and Herring, J. A. (2002),

“Relationships Among lntemet Use, Personali紗， and Social Support," Computers in Human Behavior, 18(4),437-451

134 Using Trusl and Willingness 10 lnjluence Ihe Spread 0/ Electronic Word-of-Moulh

Attractiveness: Investigating If and When Word of Mouth Matters," Journal of Applied Socia! Psycho!ogy, 37(9), 2024-2047

Wilson

,

J. M.

,

Straus

,

S. G. and McEvily

,

B. (2006)

,“

All In Due Time: The Development of Trust in Computer-Mediated and Face-to-Face Teams," Organizational 8ehavior and Human Decision Processes, 99(1), 16-33 Wi巾， 1. and Chew

,

P. (2002)

,“

The Effects of Incentives

,

Deal Proneness

,

Satisfaction and Tie Strength on Word-of-Mouth Behaviour," International Journal ofService Industry Manageme肘" 13(2)

,

141-162

Wu

,

C. C. and Lee

,

G. C. (1999)

,“

Use ofBBS to Facilitate a Teaching Practicum Course

,

" Comput

ers and Education, 32(3)

,

239-247