The current issue and full text archive of this journal is available at www.emeraldinsight.com/1066-2243.

htm

INTR 16,4

The value of participation in virtual consumer communities on brand loyalty

Rong-An Shang, Yu-Chen Chen and Hsueh-Jung Liao
Department of Business Administration, Soochow University, Taipei, Taiwan, Republic of China
Abstract
Purpose To examine the effects of consumers lurking and posting behaviors in virtual consumer communities on specic brand loyalty. Design/methodology/approach An online survey of a virtual community of Apple computer users was conducted to test a model of involvement, participation, loyalty, trust, and attitudes toward the brand exist in messages within the community. Findings The causes and effects of lurking and posting differed. Lurking contributed to brand loyalty more than posting did, and the primary purpose of lurking was to look for information regarding product function/performance, instead of satisfying consumers affective needs. Research limitations/implications As a result of the limitations in a computer mediated communication environment, virtual consumer communities should be perceived as an arena for weak-tied strangers to communicate for word-of-mouth (WOM) behavior, rather than as a brand community. Practical implications Although negative messages in a community can hurt brand image, companies should encourage the establishment of communities about their products and encourage consumers to participate in them. Originality/value Although the importance of virtual communities has been recognized, few studies have been done to examine the business value of consumer communities. Based on perspectives of brand community and WOM behavior, this paper contributes to virtual community and marketing research by clarifying the effects of consumers lurking and posting behaviors in a consumer community on brand loyalty. Keywords Internet, Brand loyalty, Buying behaviour Paper type Research paper

398

Internet Research Vol. 16 No. 4, 2006 pp. 398-418 q Emerald Group Publishing Limited 1066-2243 DOI 10.1108/10662240610690025

Introduction The Internet has become a forum for people to interact and communicate with each other. The increasing number of people chatting in cyberspace may have potential social inuences and create characteristics similar to traditional communities. These communicating groups in cyberspace have been called virtual communities. Rheingold (1993) dened a virtual community as a social aggregation of people carrying out public discussions long enough, with sufcient human feeling, to form webs of personal relationships in cyberspace. As a result, the word community no longer refers only to social groups bound together by geography. Virtual communities can be divided into commercial and non-commercial communities (Leimeister et al., 2004). Non-commercial communities, or consumer communities (Thorbjrnsen et al., 2002), are operated or founded by consumers or a third-party for information exchange or relationship establishment. These could be special forms of

brand communities focused on a specic brand, where interactions among members mainly take place in cyberspace (Muniz and OGuinn, 2001). The value of brand communities has been recognized recently (McAlexander et al., 2002; Muniz and OGuinn, 2001; Muniz and Schau, 2005). The high consumer cost of participation in brand communities, however, constrains their popularity. Virtual consumer communities may provide a feasible solution for overcoming this restraint, due to the relative ease of interaction with others in cyberspace. The effects of participation in a real activity, however, differ from the effects of participation in a cyberspace community. Consumer communities are believed to have a greater inuence on consumers and are more credible than commercial ones (Bickart and Schindler, 2001) but consumer communities may also be disadvantageous to a company because negative messages from unsatised consumers could threaten brand image (Maclaran and Catterall, 2002). The value of consumer participation in virtual consumer communities still needs to be investigated. This study examines the business value of consumer participation in virtual consumer communities by taking consumer loyalty as a surrogate. The effects of participation in the eld of marketing have not been as well understood and emphasized as in the eld of organizational behaviors or information systems. Previous studies have shown that participation can greatly affect peoples attitudes, but different kinds of participation lead to different results (Barki and Hartwick, 1989, 1994; Ives and Olson, 1984; Vroom and Jago, 1988). Consumers who participate in a consumer community focusing on a specic brand, whether posting or lurking, may change their attitude towards the brand. Researchers have suggested that consumers join communities in order to learn from others experiences or acquire information (Romm et al., 1997), and involvement has been considered as the reason why consumers look for product information (Beatty and Smith, 1987). Therefore, we conducted a survey to test a model of involvement, participation, and loyalty. First, we wanted to explore the effects of consumers posting and lurking behaviors in the virtual consumer community on brand loyalty. Second, we wanted to see whether affective or cognitive involvement could cause people to participate in the virtual consumer community. Finally, we wanted to examine whether trust and attitudes perceived in messages in the community could inuence the effects of consumers participation. The remainder of this paper is divided into six sections. The next section overviews the literature on virtual community, consumer involvement, and brand loyalty as the theoretical background of this study. Section 2 derives a research framework and discusses the hypotheses tested in this study. Section 3 introduces the research method, and Section 4 reports the results. Implications of the ndings are explained in Section 5 and Section 6 concludes the paper. Theoretical background Virtual community Virtual communities have received widespread attention during the past decade. Studies have focused on two facets of virtual communities. The rst concerns the motivations for participation. In organization or IS research, participation usually means taking part in or contributing to some specic activities or events, and is usually measured by specic behaviors, activities, and assignments (Barki and Hartwick, 1994; Vroom and Jago, 1988). Therefore, participation in a virtual community may only refer to the members posting behaviors (Mathwick, 2002). On the

The value of participation

399

INTR 16,4

400

other hand, Ives and Olson (1984) also included activities of a passive nature in their conceptualization of participation. Although not directly contributing to the community, lurking in the virtual community can also be seen as a form of passive participation (Blanchard and Markus, 2003; Nonnecke and Preece, 2000). Nonnecke and Preece (2004) found that the main purpose for both lurkers and posters is to gain a general understanding of a topic. Common interests, values, or desire among participants to discuss specic issues are also motives for participation mentioned in previous studies (Armstrong and Hagel, 1996; Rheingold, 1993). The second issue is the effects of participation in virtual communities. Empirical studies of how virtual communities inuence business are relatively rare, especially for non-commercial communities. Opinions communicated between members of a consumer community may represent a type of written word-of-mouth (WOM) (Bickart and Schindler, 2001; Hennig-Thurau et al., 2004). WOM is an important factor affecting customers attitudes and buying decisions (Buttle, 1998; Murray, 1991). Maclaran and Catterall (2002) argued that consumer communities might become threats to companies because companies have no control over negative messages from unsatised customers on the Internet. Muniz and OGuinn (2001) also asserted that brand communities operating within cyberspace could pose enormous rumor control problems. Consumer virtual communities can also be seen as a special form of brand communities. Muniz and OGuinn (2001) dened a brand community as a specialized, non-geographically bound community, based on a structured set of social relationships among admirers of a brand. It is marked by a shared consciousness, a set of rituals and traditions, and a sense of moral responsibility. Brand communities carry out important functions on behalf of the brand, such as sharing information, perpetuating the history and culture of the brand, and providing assistance. The experiences in the community may constitute a major part of the consumer experiences of the brand. Previous studies have shown that events or activities are important for the emergence of a brand community (McAlexander et al., 2002). There is a high consumer cost, however, for participation in the events or activities. On the other hand, due to the low cost of interaction with others in cyberspace, consumers can easily and more frequently share their brand feelings or experiences with others, and the feelings of belonging, trust, and obligation, as well as group symbols, culture, and rules, can be developed in the virtual community (Herring, 1996; Rheingold, 1993). Virtual consumer communities may provide a feasible alternative to brand communities for admirers of a brand. Involvement Information searching has been proposed as the main purpose for community members participation. Consumers are motivated to search for information by their involvement in an object (Bloch et al., 1986). Therefore, involvement can be seen as the ultimate motivation for consumers to participate in a virtual community. Involvement is dened as ones perception of the degree of personal relevance towards an object (Krugman, 1967; Mitchell, 1979). Product involvement concerns the characteristics of the object and the objects value to an individual (Zaichkowsky, 1986). Product involvement leads to information searching, because certain products or brands are related to an individuals ego or self-image (Bloch and Richins, 1983; Beatty and Smith, 1987). An individual may participate in a consumer community and search for information when he/she is involved in a particular brand or product.

Involvement can be cognitive or affective. Cognitive involvement is caused by a utilitarian motive, referring to an individuals concern with the cost and benets of the product or service and interest in the functional performance of the product. Affective involvement is caused by a value-expressive motive, referring to an individuals interest in enhancing self-esteem or self-conception, and in projecting his/her desired self-image to the outside world through the use of the product or service (Park and Young, 1983). While high cognitive involvement leads to intense processing of product-feature information, affective involvement is more focused on symbolic qualities and image dimensions (Mittal, 1987). 2.3 Brand loyalty Brand loyalty originally referred to consumers repeated purchasing. Repeat purchasing, however, may only indicate consumers temporary acceptance of a brand. Therefore, the concept of brand loyalty was extended to encompass both attitudinal and behavioral loyalty (Jacoby and Kyner, 1973). Behavioral loyalty means that consumers will repeatedly purchase the same brand. Attitudinal loyalty is assumed to be more stable than behavioral loyalty and represents consumers commitment or preferences when considering unique values associated with a brand (Chaudhuri and Holbrook, 2001). Oliver (1999) dened brand loyalty as a strong commitment to re-buy a preferred product or re-patronize a service consistently in the future, thereby causing repetitive purchasing of same-brand or same brand-set products, despite situational inuences. Gounaris and Stathakopoulos (2004) also argued that attitudinal loyalty should lead to an increase in behavioral loyalty. Researchers have suggested that involvement will result in customer commitment or loyalty. Howard and Sheth (1969) stated that highly involved consumers might perceive greater product importance and commit to their choice. Tyebjee (1977) demonstrated that consumers with low involvement were characterized by low brand loyalty and searched for alternatives. He argued that the higher the involvement, the greater the commitment, and hence, loyalty to the brand. Iwasaki and Havitz (1998) suggested a three-stage model to present the relationship between involvement and behavioral loyalty. The sequential psychological progress includes the formation of a high level of involvement in an activity, then the development of a psychological commitment to a brand, and nally, the maintenance of strong attitudes towards resisting changes in brand preference. They further argued that involvement would bring with it the sunk costs that would further enhance the level of commitment. Research framework Participating in a virtual community could be seen as an involving activity, motivated by consumer involvement and presenting a commitment to the brand or product, leading to increased brand loyalty. The motivations and effects of participating in the virtual community can be moderated by the trustworthiness of the community and the perceived attitude toward the brand contained in the messages within the community. We proposed a framework to illustrate the relationships between involvement, brand loyalty, virtual community participation, trust, and the perceived attitudes of the messages to clarify the value of consumer communities to companies. The research framework is shown in Figure 1.

The value of participation

401

INTR 16,4

402

People participate in virtual communities because they have common interests, values, or concerns in discussing specic issues with others (Armstrong and Hagel, 1996; Rheingold, 1993). Previous studies have suggested that participation in a virtual community is motivated by a need for information (Romm et al., 1997; Chan et al., 2004). People may join a certain consumer community because they feel the issues regarding the specic brand are relevant to them. A sense of cognitive involvement will lead a consumer to search for product-feature information (Bloch and Richins, 1983; Beatty and Smith, 1987; Mittal, 1987). A brand community is very likely an important information resource for consumers (Muniz and OGuinn, 2001). Therefore, we proposed a hypothesis that states: H1a. The degree of cognitive involvement in a brand will have a positive effect on the lurking behavior in the virtual consumer community about the brand. Affective involvement is caused by an individuals value-expression motives that focus on emotional or aesthetic appeals to express ones self-image (Park and Young, 1983). An affectively involved person is very likely emotionally bonded with an object that stands for his/her actual or ideal self-image. Affective involvement may stress a persons desire to achieve certain emotional states evoked by that object (McGuire, 1974). Highly affectively involved consumers will focus more on symbolic and image dimensions of a brand and tend to process information related to his/her self-image, and may care more about issues regarding that brand or product. Therefore, we proposed the following hypothesis: H1b. The degree of affective involvement in a brand will have a positive effect on the lurking behavior in the virtual consumer community about the brand. Posting in the virtual consumer community may represent a type of written WOM (Bickart and Schindler, 2001; Hennig-Thurau et al., 2004). Researchers have suggested that WOM behavior is motivated by involvement (Engel et al., 1993). Mathwick (2002) also proposed that consumers take part in the community for exchange or communal reasons. Exchange-oriented consumers are willing to contribute to the community because of their obligations to repay others or their expectations of receiving benets from others in the future. People looking for more information are thus more likely to post in order to sustain the exchange relationship with others and gain more information in the future. Therefore, we proposed that: H1c. The degree of cognitive involvement in a brand will have a positive effect on the posting behavior in the virtual consumer community about the brand.

Figure 1. Research framework

Posting in a virtual community provides affectively involved members a way to present themselves, build self-image, and promote their beliefs (Cheung et al., 2004 Schau and Gilly, 2003). Those who are affectively attached to a brand may enjoy meeting others with the same beliefs and sharing their experiences with them. People are more likely to help others and contribute to the virtual community for communal reasons when they feel that they belong to the same group. Therefore, we proposed that: H1d. The degree of affective involvement in a brand will have a positive effect on the posting behavior in the virtual consumer community about the brand. The interpersonal communications within a consumer community may inuence members attitudes and preferences toward brand or purchase decisions (Bickart and Schindler, 2001). A customers understanding of the product/brand will increase during this process. His/her expectations for the product/brand will more likely match actual performance; this will increase his/her satisfaction and reinforce his/her loyalty to that brand or product. Therefore, we hypothesized that: H2a. The degree of lurking behavior in a virtual consumer community about a brand will have a positive effect on the loyalty to that brand. The time and effort invested into participating in a community is a sunk cost that will turn into consumer commitment to the brand, and end with loyalty (Iwasaki and Havitz, 1998). Posting requires more effort than lurking. When a consumer has spent more time in a brand community, he/she will be more likely to have a positive attitude toward the brand, avoiding the problem of cognitive dissonance. Therefore, we proposed the following hypothesis: H2b. The degree of posting behavior in a virtual consumer community about a brand will have a positive effect on the loyalty to that brand. Whether or not a consumer trusts the members in the community may affect his/her willingness to interact with them (Yoon, 2002). Although consumers involvement in a brand can bring them to the community, people will not take part if they do not trust the community. Therefore, trust in a community might strengthen the relationship between involvement and participation. This leads to: H3. A consumers trust in a community will moderate the effect of the involvement on the participation in the virtual consumer community. Furthermore, messages circulating within a consumer community may be perceived by participants as having positive or negative attitudes towards the brand that is the focus of the community. A participants perception of the attitude of the messages in a community will also affect his/her attitude. Positive endorsements from users would help promote a company by circulating WOM advertising (Cheung et al., 2004). Similarly, negative messages from unsatised consumers could affect the attitudes of the other members, thus becoming a threat to companies (Maclaran and Catterall, 2002). Negative messages create lower loyalty in the active participants than the inactive ones. Therefore, we hypothesized that the attitude perceived in messages towards the brand would moderate the effects of participation on brand loyalty. This leads to:

The value of participation

403

INTR 16,4

H4. The perceived attitude of the messages in the virtual consumer community will moderate the effect of participation in the virtual community on brand loyalty. Studies have suggested that involvement causes psychological commitment to a brand, leading to stronger brand loyalty (Howard and Sheth, 1969; Tyebjee, 1977). Cognitive involvement promotes consumers loyalty by increasing the knowledge/information related to the brands performance or functions. On the other hand, affective involvement stresses a persons achievement of certain emotional states evoked by an object. Consumers loyalty can also be increased when ones self-image is improved by the brand. Therefore, we proposed the following two hypotheses: H5a. The consumers cognitive involvement in a brand will have a positive effect on his/her loyalty to that brand. H5b. The consumers affective involvement in a brand will have a positive effect on his/her loyalty to that brand. Iwasaki and Havitz (1998) suggested that consumers commitment and loyalty could be developed while engaged in involving activities. While virtual community participation can be seen as an involving activity on the Internet, a consumers involvement can increase participation in the community and lead to increased loyalty to the brand. In order to examine whether consumer participation in virtual communities truly benets companies, we test whether participation could contribute to an explanation of loyalty that goes beyond the explanation of involvement. Therefore, we proposed the following hypothesis: H6. Participation in the virtual community will have additional explanatory power in predicting consumer loyalty that beyond consumer involvement alone. Method We conducted an online survey of the members of Frostyplace (www.frostyplace.com) to test the hypotheses we proposed. Frostyplace is a well-known virtual community of Apple computer users in Taiwan. Founded in 1999, it now draws more than 7,000 amateur Apple computer users. Apple is famous for its devoted customers and is iz often used to illustrate the value of brand communities (Mun and OGuinn, 2001; Muniz and Schau, 2005). Inspection of the messages in Frostyplace did not show that physical interactions in the real world among members were popular. Scales to measure the constructs are shown in the Appendix. Involvement refers to product involvement in this study and is operationally dened as ones perceived personal relevance to Apple computers. The measurements of both cognitive and affective involvement were adopted from the Personal Involvement Inventory (PII) (Zaichkowsky, 1994); ten items of a 7-point semantic differential scale were used. Researchers have suggested that attitudinal loyalty is a better indicator of brand loyalty than behavioral. Therefore, loyalty refers only to attitude in this study. The attitudinal loyalty was measured using scales developed by Quester and Lim (2003); two items were deleted due to semantic problems during translation, resulting in 13, 7-point Likert-type items.

404

Participation is dened as an individuals lurking time per week and posting time per month. Questions were designed for the most active time period and the present time period for both posting and lurking. The perceived attitude of the messages means the subjects perception of the positive or negative attitudes toward the brand in the messages in the community. Three items were used to measure the perceived attitude of the messages within Frostyplace. The rst item concerns members impressions about the attitude of the messages towards Apple computers. The next two items use a 5-point Likert-type scale to measure the perceived frequency of positive and negative messages in the community. Trust refers to the members trust in other members of Frostyplace and was measured by the items developed by Wong and Sohal (2002). Some items were deleted due to a translation problem, resulting in eight 7-point Likert-type scale items. One more item that asks about members willingness to provide personal information to others in the community was added to include the specic feature of trust in the online community (Schoenbachler and Gordon, 2002). Finally, a dummy variable for product possession, meaning whether the member currently owned an Apple product, was used as the control variable because product possession might affect the other variables in the model. Two stages of pretests were performed to develop instruments. First, ten subjects were asked to examine the meaning and correct the wording of items. Then we invited 33 Frostyplace members to ll out an online questionnaire and discussed with them via MSN (Microsoft Messenger) in order to test the system and procedures. Members who had posted e-mail addresses or MSN accounts on the Frostyplace member list were invited to participate in this study. It should be noted that the subjects were not randomly selected. The sample of members willing to provide personal contact information could be more active or more trusting of others in the community than the community at large. The survey was conducted via an online questionnaire lasting about ve weeks. 430 MSN accounts were obtained and 157 MSN users were contacted. In the meantime, 1,718 e-mails were sent and 233 invalid e-mails were returned. Two pairs of questions on participation (P1/P2, P3/P4) served as the criteria for a valid response. Answers regarding the frequency of lurking and posting in the most active period should not be lower than those in the present period. Finally, 350 responses were obtained and 316 of them were valid. The demographics of the respondents and their experiences with Apple computers and Frostyplace are shown in Table I. Results A principal components analysis with orthogonal rotation by varimax method was conducted to examine the measurement model. Nine factors with eigenvalues larger than one were extracted and the factor structure is shown in Table II. The items regarding trust were separated into two factors. All the positive questions of trust were loaded in factor 3 and the reversed questions were loaded in factor 5. This result suggests that the subjects responses to the questions of trust and distrust differed. We dened factor 3 as trust and the reversed score of factor 5 as reversed distrust; both served as variables that measure trust. Items I1, I5, I8, and A3 were dropped because those items were not loaded in predicted constructs. Item I10 was deleted because the difference between the largest score of factor loading and the second largest was lower

The value of participation

405

INTR 16,4
Gender Male Female Age 0-20 21-30 31-40 41-50 Over 50 Product possession Yes No Community experience (years) 0-0.5 0.5-1 1-2 2-3 Over 3 Product experience (years) No 0-0.5 0.5-1 1-3 Over 3

Sample 262 54 10 166 123 16 1 261 55 40 67 90 56 63 13 19 43 72 169

Percent 82.9 17.1 3.2 52.5 38.9 5.1 0.3 82.6 17.4 12.7 21.2 28.5 17.7 19.9 4.1 6.0 13.6 22.8 53.5

406

Table I. Demographics of respondents and respondents experiences

than 0.1. Item T3 was deleted to increase the reliability. The Cronbachs a of variables are also shown in Table II. All the Cronbachs a were larger than 0.7, except for the reliability of the perceived attitude of the messages. Hierarchical regressions were applied to test the research model. Tables III and IV shows the impacts of involvement and trust on lurking and posting, respectively. Model 1 shows that cognitive involvement is positively related to lurking, but the impact of affective involvement is not signicant. The results of Model 5 show that the impacts of cognitive and affective involvement are not signicant. Therefore, H1a is supported but H1b, H1c and H1d are not. Among the four hypotheses regarding the impacts of involvement on participation, only cognitive involvement was found to relate to lurking. Cognitive and affective involvement do not cause posting. All the R 2 of the four models in Table IV are small, indicating that the independent variables cannot explain posting well. Besides, the fact that there is no relationship between affective involvement and lurking indicates that the value of virtual community may primarily be providing information related to functions or performance of the products to members, instead of facilitating emotional interaction among members. This result is consistent with previous arguments that state participation in a virtual community is primarily motivated by the need for information (Romm et al., 1997; Chan et al., 2004). The direct effects of trust and reversed distrust were added in Models 2 and 6. The moderate effects of trust and reversed distrust were separately tested using the moderated multiple regression (MMR) procedure in Models 3 and 4 for lurking, and in

Items

Component 4 5

The value of participation

Predicted constructs Cognitive involvement I1a 0.62 0.50 I2 0.72 c I3 0.45 0.75 I6 0.75 0.65 0.48 I8a Affective involvement I4 I5a 0.67 I7 I9 0.50 0.50 I10a Loyalty L1 0.78 L2 0.74 L3 0.84 L4 0.81 L5 0.86 L6 0.69 L7 0.86 L8 0.89 L9 0.62 L10 0.73 L11 0.77 L12 0.80 L13 0.84 Perceived attitude of the messages A1 A2 A3a Trust T1 T2 T3b T4 T5 T6 T7 T8 T9 Lurking P1 P2 Posting P3 P4 Reliabilityd 0.76 0.82
a

407

0.61 0.72 0.62 0.53

0.41

0.78 0.75 0.68 0.71 0.71 0.61 0.75 0.41 0.73 0.74 0.63 0.70 0.70 0.79 0.85 0.88 0.89 0.77

0.82

0.96
b

0.74

0.65

0.73

Notes: Suppress absolute values , 0.4; item was deleted; item was deleted in succeeding analysis to increase the reliability; citems selected for the construct are shown in italic; dthe Cronbachs a of the measures composed of the selected items for the corresponding constructs

Table II. Factor structure of principal components of variables

408

INTR 16,4

Variable

Product possession Cognitive involvement (CI) Affective involvement (AI) Trust (T) Reversed distrust (TR) CI0 T0 AI0 T0 CI00 TR0 AI0 TR0 Condition index R2 F (df1/df2) DR 2 F (df1/df2) 0.18 0.16 20.03 0.15 0.08 0.18 0.17 20.03 0.15 0.08 0.04 20.02 24.20 0.140 7.16 (7/308) * * 0.001a 0.23 (2/308) 17.23 0.097 11.20 (3/312) * * 23.69 0.139 9.98 (5/310) * *

Notes: *p , 0.05; * *p , 0.01; * * *p , 0.1, N 316; 0 : the centered mean; acompared with Model 2

Table III. The impact of involvement and trust on lurking Model 1 Model 2 T Model 3 t Model 4 t 3.02 * * 2.64 * * 2 0.51 2.46 * 1.31 0.68 2 0.32 0.18 0.16 20.04 0.15 0.09 20.01 20.03 24.32 0.140 7.15 (7/308) * * 0.001a 0.19 (2/308)

b
0.19 0.18 0.01 3.16 * * 2.74 * * 0.24 3.08 * * 2.60 * * 2 0.45 2.47 * 1.30

t 3.11 * * 2.59 * * 2 0.57 2.43 * 1.38 2 0.17 2 0.46

Model 6 t

Model 7 t

Model 8 t 0.32 20.10 20.84 3.03 * * 0.88 0.95 20.77 0.02 20.01 20.06 0.20 0.06

Variable

Model 5 b

b
0.02 2 0.01 2 0.04 0.20 0.06 0.30 2 0.18 2 0.66 3.02 * * 0.86 0.02 20.01 20.05 0.20 0.06 0.06 20.05 24.20 0.056 2.60 (7/308) * 0.003a 0.53 (2/308)

t 0.37 2 0.16 2 0.92 3.02 * * 0.93 1.68 * * * 2 1.49

0.03 0.00 0.00

0.44 0.01 0.05

Product possession Cognitive involvement (CI) Affective involvement (AI) Trust (T) Reversed distrust (TR) CI0 T0 AI0 T0 CI0 TR0 AI0 TR0 Condition index R2 F (df1/df2) DR 2 F (df1/df2) 17.23 0.001 0.084 (3/312) 23.69 0.053 3.44 (5/310) * *

0.10 20.09 24.32 0.063 2.97 (7/308) * * 0.011a 1.77 (2/308)

Notes: *p , 0.05; * *p , 0.01; * * *p , 0.1, N 316; 0 the centered mean; acompared with Model 6

The value of participation

409

Table IV. The impact of involvement and trust on posting

INTR 16,4

410

Models 7 and 8 for posting to avoid multicollinearity. The problem of multicollinearity is virtually guaranteed in MMR and so the centering mean that subtracted the mean from each score prior to transforming into the multiplicative term was used (Aguinis, 1995; Cronbach, 1987; Morris et al., 1986). Table III shows that trust affects lurking directly, but both the F-tests for the difference between the R 2 of Models 3 and 2, and of Models 4 and 2, are not signicant. Trust is the only variable related to posting in Table IV. Although the impact of CI00 TR0 on posting shows a t-value with p , 0.1, the F-tests for the DR 2 in both models are not signicant. The results of Models 3, 4, 7, and 8 indicate that H3 is not supported. As shown in previous studies (Ba and Pavlou, 2002), the results also suggest that the impacts of trust and distrust differ. Trust was found to directly affect lurking and posting but the moderating effect of trust is not signicant. This means the impacts of trust and cognitive involvement on lurking should be independent. Therefore, increasing consumers trust in virtual communities could be an effective way to increase participation. Table V presents the impacts of participation, the perceived attitude of the messages, and involvement on brand loyalty. First, the direct effects of participation were tested in Model 9. Next, the direct effects of participation, the perceived attitude, and the moderating effects of the perceived attitude were tested in Models 10 and 11. Results from Model 9 indicate that lurking will affect brand loyalty, but the impact of posting is not signicant. Therefore, H2a is supported and H2b is not. This nding again shows the difference between lurking and posting. Since, the F-tests for the DR 2 in Model 11 are not signicant, H4 is not supported. The direct effect of the perceived attitude of messages on brand loyalty is signicant, but smaller than the impact of lurking. Anyway, this result supports the suggestion that negative messages in consumer communities could threaten a brands image (Maclaran and Catterall, 2002). Since, the impact of posting on loyalty is not signicant, we only examined the explanatory power of lurking for H6. Both cognitive and affective involvements were tested in Model 12 to explain brand loyalty. Lurking was then added in Model 13 to examine its additional explanatory power. Results from Model 12 show that both H5a and H5b are supported. That the F-test for the DR 2 in Model 13 is signicant indicates that lurking does contribute to an explanation of brand loyalty that goes beyond the explanation of involvement, showing that H6 is supported. On the other hand, consumers posting behaviors do not increase their loyalty towards the brand. Discussions Some hypotheses were not support in the study. H1b argues that highly affectively involved consumers will focus more on symbolic and image dimensions of a brand and may care more about issues regarding that brand or product. However, researchers may view CMC as an information channel with limited capacity that primarily facilitates task-oriented work as opposed to socio-emotional interaction (Daft and Lengel, 1984, 1986). Messages in the virtual community may not be able to evoke the emotional states and satisfy the needs of affectively involved consumers. Therefore, affective involvement was found not to relate to lurking behavior. Both cognitive and affective involvement do not cause posting. Previous studies showed that anonymity in the virtual community can decrease the social presence and social cues conveyed in interactions (Kiesler et al., 1984; Short et al., 1976; Rutter, 1984), leading to de-individuation and decreasing self-awareness (Rice and Love, 1987).

Model 11 t

Variable t

Model 9 b

Model 10 b

b
0.42 0.27 0.41 9.52 * * 5.79 * * 0.24 0.41 0.16 8.55 * * 0.27 0.24 6.36 * *

Model 12 b t

Model 13 b

t 5.71 * * 5.22 * * 9.69 * * 3.99 * *

8.61 * *

4.05 * * 0.97 0.21 0.05 2.42 *

4.01 * * 0.97

Product possession 0.42 8.46 * * 0.43 Cognitive involvement Affective involvement Lurking (PL) 0.22 4.14 * * 0.21 Posting (PP) 0.05 0.95 0.05 Perceived attitude (PA) 0.16 PL0 PA0 PP0 PA0 Condition index 7.54 23.47 R2 0.28 0.30 F (df1/df2) 40.86 (3/312) * * 25.71 (6/309) * * DR 2 F (df1/df2) 0.12 2.50 * 20.02 20.37 0.07 1.36 23.62 0.30 32.59 (4/311) * * 0.004a 0.93 (2/309) 17.23 0.53 116.06 (3/312) * *

19.15 0.55 95.19 (4/311) * * 0.023b 15.92 (1/310) * *

Notes: *p , 0.05; * *p , 0.01; * * *p , 0.1; N 316; 0 the centered mean; acompared with Model 10; bcompared with Model 12

Table V. The impact of participation, trust, and attitude on brand loyalty

The value of participation

411

INTR 16,4

412

Impact of the social norm of reciprocity in the CMC environment may decrease. Therefore, although highly cognitively involved consumers may look for more information in the virtual community, they may not be more likely to feel the obligations and willing to repay others in the community. Besides, anonymity in a virtual community may also cause posting not an effective way for affectively involved consumers to present themselves and build self-image. Result of the test of H2b shows a consumers posting in the virtual community cannot increase the loyalty toward the brand. Although posting requires consumers to expend more effort than lurking, more effort may not lead to a more positive attitude towards the brand. This nding suggests that members of a community may discriminate between loyalty towards the community and loyalty towards the brand (Muniz and OGuinn, 2001; Muniz and Schau, 2005), and posting probably only relates to the loyalty towards the community. The moderating effect of the attitude of messages was found to be insignicant and the direct effect was small. This nding may also relate to the lack of emotional and social bonds in the virtual community. Studies on WOM suggest that WOM messages can inuence consumers attitudes because personal sources, that usually are familiar friends, are viewed as more trustworthy than the information provided by the companies (Buttle, 1998; Murray, 1991). However, the sources of messages in the community may not be so trustworthy. Relationships built in the CMC environment have often been shown as weak-tie relationships (Cummings et al., 2002; Constant et al., 1996; Pickering and King, 1995). Information that traverses a weak tie has the opportunity to reach more people, and people can encounter more diverse information sources in the community. However, the impact of this information is weaker than the impact of information that traverses a strong tie. Conclusion This study explored the impact of consumers participation in virtual consumer communities on brand loyalty. Since, posting and lurking behaviors were used as the indicators of active and passive participation, respectively, the results of this study show that the causes and effects of active and passive participation differ. Results of this study show the model we proposed cant explain the causes and effects of active participation in a virtual community well. Posting might be motivated by social factors, such as pro-social reasons, rather than by personal considerations (Constant et al., 1994; Dholakiaa et al., 2004; Hennig-Thurau et al., 2004). Anyway, this issue is worth being studied in the future. On the other hand, affected by the special characteristics in the CMC environment, virtual communities may provide information to satisfy the cognitive needs of members, but information in the virtual community cannot satisfy their emotional bonds with the brand. Findings from this study may suggest that online consumer communities should be better understood as a forum for strangers to meet and exchange weak-tie WOM messages, rather than as a form of brand communities. McAlexander et al. (2002) emphasized the importance of physical brandfest in developing a brand community. Face-to-face contact and nonverbal cues may be necessary for building intimate interpersonal relationships (Bugoon et al., 1984; Knapp and Vangelisti, 1992). Accordingly, rms can try to integrate some physical activities or artifacts into the community to increase social interactions among members and the

value of the community to the company. The identication of members within the community should also be increased so the community can further satisfy the emotional needs of affectively involved consumers, resulting in higher loyalty. This study found that lurking behavior in the virtual consumer community does contribute to an explanation of brand loyalty that goes beyond the explanation of involvement. Although the causal connection between lurking and loyalty is arguable, this result implies that promoting consumers participation in consumer communities may increase brand loyalty. Since the impact of lurking and the impact of the perceived attitude of the messages on loyalty are independent, companies should encourage the establishment of communities about their products and try to attract consumers to participate in virtual communities, even if someone else operates it. Although negative messages in the community could hurt brand image, it is good for a company when more people are interested in, talking about, and taking notice of their products. Furthermore, the impact of lurking on brand loyalty was found to be larger than the impact of the perceived attitude of the messages. Because increasing consumers trust in virtual communities is one way to increase participation, a trustworthy community that can encourage people to participate may be better for a company than a community that is full of positive messages toward the company. Finally, limitations of this study should be noted. First, the sampling procedure was not random. The subjects included only members who left personal contact information. Those people are more active or more trusting than the community at large. Although we separated lurkers and posters in this study, people who simply browse and hide in the community will not provide any personal information, and are not included in this study. Second, this study examined only a community of Apple computer users. According to the FCB involvement grid, products can be classied by whether their involvement is high or low and whether the motive is utilitarian or emotional (Vaughn, 1986; Ratchford, 1987). Computers belong to the quadrant with high involvement and utilitarian motives. Therefore, satisfying information needs may be the major function of a community for computers. The motivations for and impact of participating in a virtual community may differ for products in other categories. The impact of product type is an issue that should be further investigated.
References Aguinis, H. (1995), Statistical power problems with moderated multiple regression in management research, Journal of Management, Vol. 6, pp. 1141-58. Armstrong, A. and Hagel, J. (1996), The real value of online communities, Harvard Business Review, Vol. 74 No. 3, pp. 134-41. Ba, S. and Pavlou, P.A. (2002), Evidence of the effect of trust building technology in electronic markets: price premiums and buyer behavior, MIS Quarterly, Vol. 26 No. 3, pp. 243-68. Barki, H. and Hartwick, J. (1989), Rethinking the concept of user involvement, MIS Quarterly, Vol. 13 No. 1, pp. 53-63. Barki, H. and Hartwick, J. (1994), Measuring user participation, user involvement, and user attitude, MIS Quarterly, Vol. 18 No. 1, pp. 59-82. Beatty, S.E. and Smith, S.M. (1987), External search effort: an investigation across several product categories, Journal of Consumer Research, Vol. 14, pp. 83-95. Bickart, B. and Schindler, R.M. (2001), Internet forum as inuential sources of consumer information, Journal of Interactive Marketing, Vol. 15 No. 3, pp. 31-40.

The value of participation

413

INTR 16,4

414

Blanchard, A.L. and Markus, M.L. (2003), The experienced sense of a virtual community: characteristics and processes, Database for Advances in Information Systems, Vol. 35 No. 1, pp. 65-79. Bloch, P.H. and Richins, M.L. (1983), A theoretical model for the study of product importance perceptions, Journal of Marketing, Vol. 47, pp. 69-81. Bloch, P.H., Sherrell, D.L. and Ridgway, N.M. (1986), Consumer search: an extended framework, Journal of Consumer Research, Vol. 13, pp. 119-26. Bugoon, J.K., Buller, D.B., Hale, J.L. and DeTurck, M.A. (1984), Relational messages associated with nonverbal behaviors, Human Communication Research, Vol. 10 No. 3, pp. 351-78. Buttle, F.A. (1998), Word of mouth: understanding and managing referral marketing, Journal of Strategic Marketing, Vol. 6, pp. 241-54. Chan, M.L., Bhandar, M., Oh, L.B. and Chan, H.C. (2004), Recognition and participation in a virtual community, Proceedings of the 37th Hawaii International Conference on System Sciences, Hawaii, USA. Chaudhuri, A. and Holbrook, M.B. (2001), The chain of effects from brand trust and brand affect to brand performance: the role of brand loyalty, Journal of Marketing, Vol. 65 No. 2, pp. 81-93. Cheung, M.Y., Shek, P.W. and Sia, C.L. (2004), Virtual community of consumers: why people are willing to contribution, Proceedings of the Eighth Pacic-Asia Conference on Information Systems, Shanghai, China, pp. 2100-7. Constant, D., Kiesler, S. and Sproull, L. (1994), Whats mine is ours, or is it? A study of attitude about information sharing, Information Systems Research, Vol. 5 No. 4, pp. 400-21. Constant, D., Sproull, L. and Kiesler, S. (1996), The kindness of strangers: the usefulness of electronic weak ties for technical advice, Organization Science, Vol. 7 No. 2, pp. 119-35. Cronbach, L.J. (1987), Statistical tests for moderator variables: aws in analyses recently proposed, Psychological Bulletin, Vol. 12, pp. 414-7. Cummings, J.N., Butler, B. and Kraut, R. (2002), The quality of online social relationships, Communications of the ACM, Vol. 45 No. 7, pp. 103-8. Daft, R.L. and Lengel, R.H. (1984), Information richness: a new approach to managerial behavior and organizational design, Research in Organization Behavior, Vol. 6, pp. 191-233. Daft, R.L. and Lengel, R.H. (1986), Organizational information requirements, media richness and structural design, Management Science, Vol. 32, pp. 554-71. Dholakiaa, U.M., Bagozzia, R.P. and Pearob, L.K. (2004), A social inuence model of consumer participation in network- and small-group-based virtual communities, International Journal of Research in Marketing, Vol. 21, pp. 241-63. Engel, J.F., Blackwell, R.D. and Miniard, P.W. (1993), Consumer Behaviour, 7th ed., Dryden Press, New York, NY. Gounaris, S. and Stathakopoulos, V. (2004), Antecedents and consequences of brand loyalty: an empirical study, Journal of Brand Management, Vol. 11 No. 4, pp. 283-306. Hennig-Thurau, T., Gwinner, K.P., Walsh, G. and Gremler, D.D. (2004), Electronic word-of-mouth via consumer-opinion platforms: what motivates consumers to articulate themselves on the Internet, Journal of Interactive Marketing, Vol. 18 No. 1, pp. 38-52. Herring, S.C. (1996), Computer-mediated Communication: Linguistic, Social, and Cross-cultural Perspectives, John Benjamins Publishing Co., Amsterdam. Howard, J.A. and Sheth, J.N. (1969), The Theory of Buyer Behavior, Wiley, New York, NY.

Iwasaki, Y. and Havitz, M.E. (1998), A path-analytic model of the relationship between involvement, psychological commitment, and loyalty, Journal of Leisure Research, Vol. 30 No. 2, pp. 337-47. Ives, B. and Olson, M.H. (1984), User involvement and MIS success: a review of research, Management Science, Vol. 30 No. 5, pp. 586-603. Jacoby, J. and Kyner, D.B. (1973), Brand loyalty vs. repeat purchasing behavior, Journal of Marketing Research, Vol. 10, pp. 1-9. Kiesler, S., Siegel, J. and McGuire, T.W. (1984), Social psychological aspects of computer mediated communication, American Psychologist, Vol. 39, pp. 1123-34. Knapp, M. and Vangelisti, A.L. (1992), Interpersonal Communication and Human Relationships, 2nd ed., Allyn and Bacon, Boston, MA. Krugman, H.E. (1967), The measurement of advertising involvement, Public Opinion Quarterly, Vol. 30, pp. 583-96. Leimeister, J.M., Sidiras, P. and Krcmar, H. (2004), Success factors of virtual communities from the perspective of members and operators: an empirical study, Proceedings of the 37th Hawaii International Conference on System Sciences, Hawaii, USA. McAlexander, J.H., Schouten, J.W. and Koening, H.F. (2002), Building brand community, Journal of Marketing, Vol. 66, pp. 38-54. McGuire, W.J. (1974), Psychological motives and communication gratication, in Blumer, J.G. and Katz, E. (Eds), The Uses of Mass Communications: Current Perspectives on Gratication Research, Sage, Beverly Hills, CA, pp. 167-96. Maclaran, P. and Catterall, M. (2002), Researching the social web: marketing information from virtual communities, Marketing Intelligence & Planning, Vol. 20 No. 6, pp. 319-26. Mathwick, C. (2002), Understanding the online consumer: a typology of online relational norms, Journal of Interactive Marketing, Vol. 16 No. 1, pp. 40-55. Mitchell, A.A. (1979), Involvement: a potentially important mediator of consumer behavior, Advances in Consumer Research, Vol. 6, pp. 191-6. Mittal, B. (1987), A framework for relating consumer involvement to lateral brain functioning, in Wallendorf, M. and Anderson, P. (Eds), Advances in Consumer Research,Vol. 14, Association for Consumer Research, Provo, UT, pp. 41-5. Morris, J.H., Sherman, J.D. and Manseld, E.R. (1986), Failures to detect moderating effects with ordinary least squares-moderated multiple regression: some reasons and a remedy, Psychological Bulletin, Vol. 99, pp. 282-8. Muniz, A.M. Jr and OGuinn, T.C. (2001), Brand community, Journal of Consumer Research, Vol. 27 No. 4, pp. 412-31. Muniz, A.M. Jr and Schau, H.J. (2005), Religiosity in the abandoned Apple Newton brand community, Journal of Consumer Research, Vol. 31 No. 4, pp. 737-47. Murray, K.B. (1991), A test of services marketing theory: consumer information acquisition activities, Journal of Marketing, Vol. 55, pp. 10-25. Nonnecke, B. and Preece, J. (2000), Persistence and lurkers: a pilot study, Proceedings of the 33rd Hawaii International Conference on System Sciences, Hawaii, USA. Nonnecke, B. and Preece, J. (2004), What lurkers and posters think of each other, Proceedings of the 37th Hawaii International Conference on System Sciences, Hawaii, USA. Oliver, R.L. (1999), Whence consumer loyalty?, Journal of Marketing, Vol. 63, pp. 33-44.

The value of participation

415

INTR 16,4

416

Park, C.W. and Young, S.M. (1983), Types and levels of involvement and brand attitude formation, Advances in Consumer Research, Vol. 10, pp. 320-4. Pickering, J.M. and King, J.L. (1995), Hardwiring weak ties: interorganizational computer-mediated communication, occupational communities, and organizational change, Organization Science, Vol. 6 No. 4, pp. 479-86. Quester, P. and Lim, A.L. (2003), Product involvement/brand loyalty: is there a link?, The Journal of Product & Brand Management, Vol. 12 No. 1, pp. 22-38. Ratchford, B.T. (1987), New insights about FCB grid, Journal of Advertising Research, Vol. 27, pp. 24-38. Rheingold, H. (1993), Virtual Community: Homesteading on The Electronic Frontier, The MIT Press, Cambridge, MA. Rice, R.E. and Love, G. (1987), Electronic emotion: socioemotional content in a computer-mediated communication network, Communication Research, Vol. 14 No. 1, pp. 85-108. Romm, C., Pliskin, N. and Clarke, R. (1997), Virtual communities and society: toward an integrative three phase model, Academy of Journal of Information Management, Vol. 17 No. 4, pp. 261-70. Rutter, D.R. (1984), Looking and Seeing: The Role of Visual Communication in Social Interaction, Wiley, Chichester. Schau, H.J. and Gilly, M.C. (2003), We are what we post? Self-presentation in personal webspace, Journal of Consumer Research, Vol. 30, pp. 385-404. Schoenbachler, D.D. and Gordon, G.L. (2002), Trust and customer willingness to provide information in database-driven relationship marketing, Journal of Interactive Marketing, Vol. 16 No. 3, pp. 2-16. Short, J.A., William, E. and Christie, B. (1976), The Social Psychology of Telecommunication, Wiley, Chichester. Thorbjrnsen, H., Supphellen, M., Nysveen, H. and Pedersen, P.E. (2002), Building brand relationships online: a comparison of two interactive applications, Journal of Interactive Marketing, Vol. 16 No. 3, pp. 17-34. Tyebjee, T.T. (1977), Renement of the involvement concept: an advertising planning point of view, paper presented at the Eighth Annual Attitude Research Conference of American Marketing Association. Vaughn, R. (1986), How advertising works: a planning model revisited, Journal of Advertising Research, Vol. 26, pp. 57-66. Vroom, V. and Jago, A.G. (1988), The New Leadership: Managing Participation in Organizations, Prentice-Hall, Englewood Cliffs, NJ. Wong, A. and Sohal, A. (2002), An examination of the relationship between trust, commitment and relationship quality, Journal of Retail & Distribution Management, Vol. 30 No. 1, pp. 34-50. Yoon, S.J. (2002), The antecedents and consequences of trust in online-purchase decisions, Journal of Interactive Marketing, Vol. 16 No. 2, pp. 47-63. Zaichkowsky, J.L. (1986), Conceptualizing involvement, Journal of Advertising, Vol. 15 No. 2, pp. 4-14. Zaichkowsky, J.L. (1994), Research notes: the personal involvement inventory: reduction, revision, and application to advertising, Journal of Advertising, Vol. 23 No. 4, pp. 59-70.

Appendix. Measures for the variables

The value of participation

417

Figure AI.

INTR 16,4

418
Figure AI.

Corresponding author Rong-An Shang can be contacted at: rashang@scu.edu.tw

To purchase reprints of this article please e-mail: reprints@emeraldinsight.com Or visit our web site for further details: www.emeraldinsight.com/reprints