Why people lurk in theme-oriented virtual communities: integrating achievement goal theory and cognitive dissonance theory

Introduction

Theme-oriented virtual communities are platforms where users with the same subject interest share and exchange their knowledge, experiences, opinions, and feelings (Ridings and Gefen, 2006). In fact, user-generated content on these platforms is a kind of valuable resource since it not only helps users broaden their knowledge, but also aids virtual communities to obtain competitive advantage. However, only contributing members (posters) impart knowledge, whereas non-contributing members (lurkers) are usually passive and seldom generate content on virtual communities (Honeychurch et al., 2017). Nonnecke et al. (2006) pointed out that lurking has a negative influence on virtual communities, since lurkers barely post, which causes a decrease in the development of virtual communities. Besides, when plenty of users seldom participate in sharing information and discussing ideas with others in virtual communities, not only the discussed topics become narrow, but also the volume of content is low. Moreover, it may also lead the shared knowledge and opinions to be less useful as the content on the platform just represents a small number of users’ experiences. Furthermore, the quality of the posted information may not be as good as users’ expectations. Therefore, with less participation, lurking creates an unattractive environment, and would not entice users to virtual communities. Hence, for managers of virtual communities, understanding the issue regarding lurking behaviour is vital.

Virtual community can be classified into theme-oriented virtual communities and social network websites (SNSs) (Rau et al., 2008), and their contextual features are different. While SNSs, such as Facebook, have become popular for maintaining social relationships (Shiau et al., 2018), theme-oriented virtual communities, like virtual communities of motorcycle modification, have also become important for solving problems, enhancing competence, obtaining knowledge, exchanging information, and contributing content (Hsu et al., 2007; Pai and Tsai, 2016; Wasko et al., 2009), primarily for the purpose of learning (Marett and Joshi, 2009).

Previous studies indicated that the factors influencing lurking behaviour on SNSs differ from those on theme-oriented virtual communities since the environmental features of them are distinct. For example, privacy concern is an important factor affecting lurkers on SNSs. This is because users in theme-oriented virtual communities communicate anonymously (Tsai and Bagozzi, 2014), whereas users on SNSs usually disclose their identities (Schau and Gilly, 2003). Besides, Ortiz et al. (2018) demonstrated that privacy risk belief has a significant and positive effect on lurking. Moreover, online community environmental features are regarded as key determinants when investigating the influences on lurking on the online cancer communities (e.g., Badreddine and Blount, 2021,). To explore the antecedents of lurking within theme-oriented virtual communities, the contextual feature, learning, is emphasised in this study.

In this study, virtual communities of motorcycle modification are focused on. A theme-oriented virtual community enables knowledge and information exchange for mutual learning and problem solving without physical constraints (Hsu et al., 2007; Wasko et al., 2009). Virtual communities of motorcycle modification can be considered as a kind of theme-oriented virtual community. Contents on motorcycle modification forums are various expertise associated with motorcycles, such as performance tuning and appearance enhancing. For example, riders discuss how to change tyres, exhausts, fairings and headlights for better appearance. In addition, they share knowledge about changing the silencer for sound, also about tuning carbs, or modifying cylinder head to improve the gas flow and performance. In this kind of online forum, people impart knowledge and skills about how to modify their motorcycles for better performance and appearance. Through learning information, knowledge and skills, many riders are able to modify their motorcycles based on their driving style and interests. Compared with other kinds of virtual communities, users on theme-oriented virtual communities, like online forums of motorcycle modification, focus on sharing, obtaining, and exchanging domain knowledge about motorcycle modification, instead of general information and news.

Based on the learning feature within virtual communities of motorcycle modification, this study draws on achievement goal theory to understand the determinants of lurking. A user on virtual communities of motorcycle modification can browse a theme, and share knowledge to a particular thread, and other users can respond to or comment on the post. They can expand their knowledge by browsing posts and solve problems by discussing with others. Participating in virtual communities of motorcycle modification can enhance abilities and skills in specific fields. Thus, engaging in virtual communities of motorcycle modification may be a behaviour associated with learning skills and improving abilities. In other words, whatever users behave on theme-oriented virtual communities can be regarded as a kind of competence-related behaviour. Elliot and McGregor (2001) proposed the 2x2 achievement goal framework and indicated that achievement goals are used to measure motivation for competence-related behaviour. However, there has been relatively little research into the role of achievement goals in behavioural engagement within theme-oriented virtual communities, although the influence of these goals has been recognised in key work in achievement-related activities. Therefore, this study adopts achievement goal theory to explore the influences of achievement goals on lurking behaviour in virtual communities of motorcycle modification.

Cognitive dissonance theory is another key theoretical foundation in this study. Cognitive dissonance has been defined as a negative state of uncomfortable arousal resulting from an inconsistency between two cognitive elements (Festinger, 1957) which are about their behaviours, beliefs, attitudes, and environments (Anderson, 1973). In other words, the basis of this theory is the imbalance in individuals’ cognitive systems caused by a discrepancy between expectations and outcomes (two cognitive elements). Cognitive dissonance also occurs in virtual communities of motorcycle modification. To elaborate, new users may expect that they will receive a lot of useful and friendly replies if they make a post. However, after they observe how other users interact with a new poster, they may realise that no one makes a reply to a new user most of the time. Then, cognitive dissonance happens. This is because their two cognitive elements (one is behavioural cognition and the other is cognition towards the interaction within the environment) are inconsistent.

Wicklund and Brehm (1976) pointed out that cognitive dissonance impels individuals to attempt to reduce and eliminate. Similarly, Elliot and Devine (1994) showed that cognitive dissonance motivates individuals toward its reduction. Previous studies indicated that people with cognitive dissonance may adopt behavioural strategies to relieve discomfort caused by it (e.g., Liang, 2016). Besides, Ng et al. (2011) expressed that people reduce this cognitive dissonance by changing their attitude, trivialising the importance of cognition, and executing behaviour. For the users who experience cognitive dissonance in the above case, they may tend to execute a non-contributing behaviour (a kind of behavioural strategy) instead of posting, so they can facilitate dissonance reduction. Although the relationship between behaviour and cognitive dissonance has been demonstrated in previous studies (e.g., Ng et al., 2011), few studies have explored the antecedents of lurking based on the cognitive dissonance theory. There is a need for research to examine the impact of lurking on theme-related virtual communities using the cognitive perspective. Accordingly, this study addresses the issue of lurking on virtual communities of motorcycle modification by drawing on achievement goal theory and cognitive dissonance theory.

This study extends previous work by responding to calls to (1) examine the influences of four achievement goals on lurking and (2) investigate the relationship between cognitive dissonance and lurking. Findings from this research have key implications for managers of theme-oriented virtual communities.

Literature

Lurking in online virtual communities

Lurkers are given plenty of names, like free riders, vicarious learners, browsers, read-only participants, non-public participants, observers, or invisible learners (Honeychurch et al., 2017). Actually, the definition of lurker is diverse. First, some scholars argued that lurkers refer to individuals who do not participate or post online (Schlosser, 2005; Edelmann, 2013; Phang et al., 2015). Secondly, others claim that lurkers are people who post online occasionally (Rafaeli et al., 2004). Thirdly, people who regularly visit the community but rarely post are called lurkers (Rafaeli et al., 2004,). Another definition is the act of seeking information online (Shang et al., 2006). Similarly, Leban et al. (2020) viewed lurking as the consumption of online content. A further definition is people who only read articles and search for information on the Internet, but never post (Takahashi et al., 2003). Lastly, Taylor (2002) showed that lurkers are community members who post fewer articles than the average number of articles that each member posts. According to the literature, in this study, if individuals seldom post in a virtual community, they are lurkers.

Since lurking is a common phenomenon on virtual communities, some theories have been applied. Firstly, scholars consider lurking as a personal trait based on personal trait theory (Panciera et al., 2010). Secondly, literature on engagement theory indicated not only lurking, but also contributing, is a type of engagement on online communities where some users engage in sharing activities and others engage to a low degree (Nonnecke et al., 2006; Nonnecke and Precce, 2000; Ridings et al., 2006). Thirdly, in terms of social learning theory, several studies argued that users usually act as lurkers but after a period, they obtain knowledge and confidence and then make contributions (Preece and Shneiderman, 2009). Finally, some researchers adopted the uses and gratification theory to explain the reasons why users lurk and they pointed out that lurkers are able to satisfy their information needs without posting (Preece et al. 2004). For example, Preece et al. (2004) claimed that there are five reasons for lurking: the necessity of posting; insufficient information about the group; being helpful by not posting; software usability; and community fit for personal needs. Kucuk (2010) offered five major reasons for students’ lurking behaviour: they feel no need to post; have no capability to use the software; do not like the forum; believe not posting any content can also help forum members; and think they need to gain more understanding of the forum.

Recently, with the popularity of social networking services, scholars also pay attention to the lurking topic on social networking services. The main task for users who use such services is to maintain their social relationship (Shiau et al., 2018) that is different from theme-oriented virtual communities, where users mainly focus on learning. In addition, scholars showed that individuals tend to express their identities more fully on social networking services (Schau and Gilly, 2003) compared with online forums. Therefore, the antecedents of lurking on SNSs differ from that of theme-oriented virtual communities. For instance, Rau et al. (2008) explained that verbal intimacy and affective intimacy are useful for discriminating posters/lurkers group of users.

Achievement goal theory

Achievement goals are defined as cognitive representations of competence-relevant possibilities that conduct behaviour to a competence-related end state that people are committed to either approach or avoid (Elliot and Trash, 2001; Hulleman et al., 2010). An achievement goal is related to the purposes of achievement behaviour (Ames, 1992). Dweck and Leggett (1988) expressed that achievement goals are cognitive frames that drive people to engage in achievement-related activities. Besides, many scholars argued that achievement goals affect how people allocate their attention and accomplish a task in achievement settings (Elliot and Church, 1997; Elliot and McGregor, 1999; Harackiewicz et al., 2000).

Ames (1992) indicated that achievement goal theory is a useful framework for understanding individuals’ motivation for achievement tasks. Achievement goals in the two-by-two achievement goal framework proposed by Elliot and McGregor (2001) suggested that both mastery and performance goals can be divided into approach and avoidance goals. Individuals with mastery achievement goals focus on acquiring new knowledge, improving their competence, and mastering their skills (Dweck and Leggett, 1988) whereas individuals with performance achievement goals prefer to demonstrate competence and to seek favourable judgments from others (Elliot and McGregor, 2001; VandeWalle, 1997). Elliot and Church (1997), and Elliot and McGregor (2001) showed that the latter approach frames highlight seeking gains whereas avoidance frames focus on preventing losses. Performance-approach goals are those that display one’s competence in relation to others while performance-avoidance goals are those that avert an individual from showing incompetence. Mastery-approach goals mean an endeavour to master tasks and the pursuit of success whilst mastery-avoidance goals refer to the avoidance of task-based or intrapersonal ineptitude.

Recently, a great number of studies explored the impacts of achievement goals on intention, behaviour, or performance, especially in a learning context. For instance, Duffy and Azevedo (2015) examined the influence of achievement goals and scaffolding on self-regulated learning and achievement. Besides, Wang et al. (2016) demonstrated that achievement goals are related to the perception of climate, physical activity intention, and participation. Moreover, Bernacki et al. (2014) explored whether strength or variability in achievement goals influences behaviour or achievement. Among others, and of particular relevance to the current study, Asterhan and Bouton (2017) explored the relationship between achievement goals and peer-to-peer knowledge sharing on social networking services in secondary schools. The results revealed that mastery goal orientation contributes to the predictions of knowledge sharing intensity and the use of shared knowledge.

Cognitive dissonance theory

Previous studies indicated that individuals have cognitive elements or knowledge about their past behaviour, beliefs, attitudes, and environments (Anderson, 1973). When there is an inconsistency between two cognitive elements, or between behaviour and cognitions, cognitive dissonance occurs (Festinger, 1957). Anderson (1973) demonstrated this concept. He indicated that individuals can receive product information from their experiences and salesmen, and these stimuli (information) may cause cognitive dissonance. In addition, cognitive dissonance, which leads to a psychological state of tension or discomfort (Elliot and Devine, 1994) within the individual, refers to an aversive motivational state (Harmon-Jones, 2002). Anderson (1973) contended that because cognitive dissonance causes psychological discomfort, it motivates the individual to eliminate it.

Cognitive dissonance can be alleviated by either cognitive or behavioural strategies (Liang, 2016). Cognitive strategies mean psychological changes associated with the cognitive elements, whereas behavioural strategies refer to individuals with cognitive dissonance who can adopt behaviour that reduces cognitive dissonance (Liang, 2016). To illustrate, individuals who possess negative attitudes towards a particular product obtain some positive information about this product from salesmen, experience cognitive dissonance, and then change their attitudes (cognitive strategy). On the contrary, when people who prefer this product receive plenty of bad news about it from online platforms, cognitive dissonance also occurs. This is because their attitudes are not consistent with the environmental information. Therefore, they may decide not to buy it (behavioural strategy). Not only cognitive strategy, but also behavioural strategy, could aid people to lessen cognitive dissonance.

Research model and hypotheses

In this study, theme-oriented virtual communities are the main contexts. This study focuses on motorcycle modification forums which are a kind of theme-oriented virtual community. The process whereby individuals mutually exchange and discuss information about modified motorcycles, by posting on and replying to motorcycle modification forums, is defined as information sharing. In contrast, individuals who only browse information about modified motorcycles and seldom participate in any discussion are defined as lurkers.

This study adopts achievement goal theory and cognitive dissonance theory as the theoretical foundation. The reasons are provided as the following. Firstly, learning is the main goal for online forum users (Marett and Joshi, 2009). Achievement goal theory has been used extensively to explain learning behaviour (e.g., Elliot and McGregor, 2001). Hence, based on the learning feature within motorcycle modification forums, achievement goal theory is a proper theory to be applied in this study. Secondly, online forum users may experience cognitive dissonance since they can easily access various kinds of information from theme-oriented virtual communities that may be inconsistent with cognition that they already own. To elaborate, to improve skills and gain new knowledge, users would like to participate in sharing, posting, and interacting with others on online forums frequently. However, if they perceive that the atmosphere of an online community is unfriendly and unpleasant, they will realise that even though they ask a question by posting, they might not receive feedback. In this case, their beliefs are inconsistent with environmental information, so cognitive dissonance occurs. Based on cognitive dissonance literature, people who experience cognitive dissonance may change their behaviour to lessen it (Liang, 2016). In this study, they may choose to lurk instead of posting. Thus, cognitive inconsistency could be an inhibitor of posting.

Accordingly, this study draws on achievement goal theory and cognitive dissonance theory to propose five factors: the mastery-approach goal, mastery-avoidance goal, performance-approach goal, performance-avoidance goal, and cognitive dissonance; moreover, this study explores the influences of these factors on lurking behaviour.

Achievement goal theory indicated that individuals with mastery-approach goals tend to gain knowledge and develop new skills (DeShon and Gillespie, 2005). In addition, previous studies showed that individuals with mastery-approach goals are interested in enlarging their knowledge base (Elliot and McGregor, 2001) and taking interactive behaviour such as seeking feedback from others (Janssen and Prins, 2007) and discussing with others (Gray and Meister, 2004). Besides, people with mastery-approach goals would like to enhance task-based intrapersonal competence and mastery tasks (Wang et al., 2016).

In this study, to master the knowledge of modifying a motorcycle, individuals with mastery-approach goals may want to acquire tuning performance information from online forums. For example, they browse engine tuning information which is posted by others previously, ask how a camshaft is replaced on a tuned engine by posting, and discuss with others by posting and replying. When they encounter tuning problems, help-seeking is considered as an adaptive approach to solving problems. Help-seeking is defined as an achievement behaviour that involves the search for and employment of a strategy to obtain success (Ames and Lau, 1982). For mastery-approach-focused individuals, seeking help is regarded as an adaptive way to improve their understanding by using available resources (Karabenick, 2006) because they can receive some feedback. VandeWalle (2003) showed that individuals with mastery-approach goals tend to consider feedback as an approach to master the task. Thus, on online forums, individuals with mastery-approach goals may seek help by posting their questions and discussing with others, then receive feedback. In other words, they might frequently engage in exchanging knowledge instead of lurking in motorcycle modification forums. Therefore, this study proposes hypothesis 1.

Hypothesis 1: Mastery-approach goal negatively affects lurking behaviour.

People with mastery-avoidance goals focus on averting from misunderstanding of knowledge and a loss of competence (Elliot and McGregor, 2001). They are intrinsically motivated to learn because they want to prevent making mistakes (Elliot and McGregor, 2001; Janssen and Prins, 2007). On forums, individuals with mastery-avoidance goals focus on avoiding task-based incompetence. For instance, to avoid not learning or not completing the tuning task, they may only browse modification information about upgrading the exhausts for adding more power to the engine that is posted by others previously. Karabenick (2004) indicated that individuals with mastery-avoidance goals regard help-seeking as a low ability signal, and do not want to seek help. Thus, in this study, they may seldom seek help such as how to improve the cooling system by posting. In other words, they may engage in lurking, such as browsing articles in motorcycle modification forums. Therefore, this study proposes hypothesis 2.

Hypothesis 2: Mastery-avoidance goal positively affects lurking behaviour.

Individuals with performance-approach goals are likely to demonstrate that their competence is better than others, and want to receive favourable judgments (DeShon and Gillespie, 2005; VandeWalle, 1997). In this study, users with performance-approach goals may want to show their competence on forums. To display their successes in modifying their motorcycles, they may exhibit their knowledge by posting. For example, they may share their knowledge about modifying the exhaust to increase the power of the machine and give a new look to a motorcycle by posting and replying. Therefore, they could demonstrate that their competence is better than others by these behaviours. That is, they may engage in posting instead of lurking. Therefore, this study proposes hypothesis 3.

Hypothesis 3: Performance-approach goal negatively affects lurking behaviour.

Individuals with performance-avoidance goals focus on averting situations in which they risk displaying incompetence or getting an unfavourable judgment (DeShon and Gillespie, 2005; VandeWalle, 1997). Moreover, Elliot and Church (1997) showed that individuals who focus on performance-avoidance goals tend to avoid information sharing because they are afraid of failure.

In this study, individuals with performance-avoidance goals might limit their activity to only browsing articles on forums to avoid showing incompetence. Moreover, they seldom ask for help by posting questions because they might consider feedback and replies as a judgment revealing a lack of competence or a threat to self-esteem (Karabenick, 2004). Thus, they may engage in lurking. Therefore, this study proposes hypothesis 4.

Hypothesis 4: Performance-avoidance goal positively affects lurking behaviour.

According to Festinger (1957) and Anderson (1973), individuals experience cognitive dissonance when their cognitive elements or knowledge about cognitions, behaviour, feelings, attitudes, and environments were inconsistent. In this study, individuals may experience inconsistent cognitions when they think that posting and discussing with other members can enhance their competence, and they observe environmental information showing that other members who posted articles can only get useless information at the same time. In other words, what they expect (the first cognitive element) is inconsistent with information that they receive from virtual communities (the second cognitive element). Hence, cognitive dissonance occurs. According to Festinger (1957), because feelings of dissonance are uncomfortable, individuals who experience dissonance are motivated to alleviate this uneasiness. Behavioural strategy is an approach to reduce cognitive dissonance (Liang, 2016). Therefore, lurking may be a way to decrease cognitive dissonance because individuals may change their behaviours from posting to lurking in response to aversive results discrepant information. Thus, this study proposes hypothesis 5.

Hypothesis 5: Cognitive dissonance positively affects lurking behaviour.

Research methodology

Context and subject

Taiwan is a country with a high motorcycle density: there are many modified motorcycles on the streets. Motorcycle modification is a way for riders to present their styles and it is also a specialised field. Riders who are concerned about motorcycle modification would like to browse or to be a participant on motorcycle forums. This study focuses on three theme-oriented virtual communities that are the Jorsindo Motor Club, the motorcycle forum of Mobile01, and the motorcycle forum of Gamer Club. These forums are chosen because of their popularity in the motorcycle modification field in Taiwan. Hence, this study surveys registered members of these three forums.

Instrument

This study adopted items of achievement need of Elliot and Murayama (2008) to develop items of mastery-approach goal, mastery-avoidance goal, performance-approach goal, and performance-avoidance goal. The items of cognitive dissonance proposed by O’Neill and Palmer (2004) were adopted in this study. For evaluating lurking behaviour, the items of knowledge sharing of Hsu et al. (2007) were adopted and modified to develop the items of lurking behaviour. All items were slightly modified to suit the context of motorcycle modification forums. Each item was measured on a five-point Likert scale. The items are shown in Table 1.

Investigation procedures

To collect data, an online survey was conducted. A pre-test was carried out involving twelve registered members of these three motorcycle modification forums who were invited to complete the questionnaire. These twelve responses were used to solve some problems, such as unclear wording and to verify the face validity of the questionnaire. Next, the formal questionnaire was created by a survey website. The link to this questionnaire was posted on these three motorcycle modification forums. After eliminating responses with duplicate information on IP addresses, e-mail addresses, and registered accounts of forums, 326 valid responses (not including the initial twelve responses) were collected for further analysis.

Data analysis and results

Sample characteristics

The online survey yielded 326 valid responses, including 320 males and 6 females. Their ages ranged between 19 and 26. Sixty-seven per cent of them have a bachelor’s degree. Over 95% spend more than two hours on the Internet every day; 44% attend these forums more than eight times per week. Fifty-six per cent have used the forums for more than two years. 53% posted fewer than six articles in the last three months.

Reliability and validity

Since all participants were asked to complete a self-reported questionnaire, a common method bias test was applied. Podsakoff and Organ (1986) pointed out that a common method bias can be tested through Harmon’s one-factor test. Six factors (mastery-approach goal, mastery-avoidance goal, performance-approach goal, performance-avoidance goal, cognitive dissonance, and lurking behaviour) with eigenvalues above 1 were extracted. No single factor emerged, and no single general factor accounted for most variance, providing no evidence of common method bias.

This study also tested non-response bias. According to Ma and Agarwal (2007), the analysis was to compare means for all variables for early and late respondents, and the difference was not statistically significant based on the t-test. Thus, in this study, the sample is free of non-respondent bias. Aside from that, reliabilities of these constructs range from 0.71 to 0.96, which exceed the acceptable value of 0.5 suggested by Hair et al. (1992). The correlation coefficients and alpha values of all constructs are shown in Table 2.

Hypotheses testing

This study adopts multiple regression analysis to test the research model. The analytical result is shown in Table 3. The analytical result shows that the overall model is significant (F = 16.02, p < 0.05), with adjusted R²=0.20. Furthermore, the variance inflation factor (VIF) of each independent variable is below 2, indicating no multi-collinearity problem. The analytical result points out that mastery-approach goal (β = -0.11, p < 0.1), performance-approach goal (β = -0.39, p < 0.05), performance-avoidance goal (β = 0.13, p < 0.05), and cognitive dissonance (β = 0.31, p < 0. 05) significantly affect lurking behaviour. As a result, hypotheses 1, 3, 4 and 5 are supported. However, the mastery-avoidance goal does not significantly affect lurking behaviour (hypothesis 2 is not supported).

Discussion

Mastery-approach goal and lurking behaviour

The analytical results find that the mastery-approach goal slightly and negatively affects lurking behaviour. According to the definition of the mastery-approach goal (Elliot and McGregor, 2001; Elliot and Murayama, 2008), in this study, when community members have high mastery-approach goals, they are interested in obtaining knowledge and developing skills of motorcycle modification. This study focuses on three online forums regarding motorcycle modification in Taiwan. These forums contain rich information about motorcycle modification. Members with mastery-approach goals can obtain information by browsing on forums which enable them to broaden their knowledge and ability of motorcycle modification. However, modifying motorcycles is not an easy task. When they encounter some problems with modifying motorcycles, they may seek help. For example, they may ask questions by posting and getting answers from others’ replies. Karabenick (2003) displayed that when mastery-approach-oriented people look for assistance, they tend to ask for instrumental help. Instrumental help means that key clues and hints are presented to help an individual to understand and clarify the fundamental problems, and then they can master knowledge and skills (Karabenick, 2003). In other words, when members with mastery-approach goals encounter questions of motorcycle modification, they seek clues and hints of answers or solutions instead of direct answers or solutions. They figure out answers or solutions based on these clues and hints by doing motorcycle modification and discussing with other members. Therefore, for members with mastery-approach goals, although they browse (lurking behaviour) on forums, they tend to spend a lot of time seeking instrumental help by posting. As a result, mastery-approach goal slightly and negatively affects lurking behaviour.

Mastery-avoidance goal and lurking behaviour

The analytical results indicate that mastery-avoidance goal is not significantly related to lurking behaviour. In this study, members with mastery-avoidance goals avoid not learning or not mastering knowledge. Therefore, the best way for them to gain information is browsing on forums. However, in the real world, sometimes, even though they browse all shared contents, their problems about motorcycle modification cannot be solved. For instance, users who follow all instructions to change the battery still cannot start their motorcycles. In this case, to avoid task-based incompetence, they may seek help. Karabenick (2003) argued that when mastery-avoidance-oriented people need assistance, they tend to ask for executive help. Executive help can be realised in many forms. For instance, on online teaching platforms, users may search online to find solutions, or they post questions on the platforms and then hope to receive answers from others (Cheng and Tsai, 2011). That is, they tend to seek executive help, and there are two ways of seeking executive help (Cheng and Tsai, 2011; Karabenick, 2003). One is asking for help from others. The other is searching for assistance. In this study, the act of asking for help is like posting articles to ask for other people’s answers. The movement of searching for assistance is similar to browsing on online forums. Because people with mastery-avoidance goals may perform both acts (posting and browsing), the mastery-avoidance goal is not significantly related to lurking behaviour.

The learning motivations of the two kinds of mastery-oriented people (people with mastery-approach goals and people with mastery-avoidance goals) are different. Users with mastery-approach goals tend to ask for instrumental help and get clues and hints of answers or solutions. They usually participate actively in forums by asking questions, posting their ideas, and discussing with other members to understand and clarify the fundamental problems and to master skills. On the contrary, people with mastery-avoidance goals tend to ask for executive help instead of instrumental help. They like to get answers directly and immediately by asking for help. Thus, for people with mastery-avoidance goals, the frequency of actively engaging in activities of virtual communities such as posting and discussing with others may be less than people with mastery-approach goals. Thus, hypothesis 1 is supported slightly whereas hypothesis 2 is not supported.

Performance-approach goal and lurking behaviour

As expected, there is a negative relationship between performance-approach goal and lurking behaviour. Maehr (1989) expressed that achievement goal is the purpose of task engagement. Performance-approach goals are defined by the goals of demonstrating superior abilities and outperforming other people. In this study, when users desire to outperform others and achieve better accomplishments than others, posting occurs frequently. This is because their needs could be satisfied by showing their better competence. For example, after they change their average headlights into high-performance LED headlights, they would like to share why this headlight is a perfect option for their motorcycles, how they plug and install them, and how riders’ visibility on the roads could be higher after installing them. They can demonstrate that they are full of knowledge on choosing a headlight, and they have skills in modifying their motorcycles by generating contents on these forums. Besides, if they post, they may receive approval and compliments from other users to prove that they have better performance. This result is in line with previous studies (Elliot, 1999, 2005; Elliot and Murayama, 2008).

Performance-avoidance goal and lurking behaviour

The result expectedly showed that performance-avoidance goal is positively associated with lurking behaviour. People with performance-avoidance goal concentrate on avoiding lack of skills in comparison with peers and being negatively judged by others, and their attention is on avoiding failure (Elliot and Church, 1997). Elliot (1999) indicated that performance-avoidance goal is an aversive motivation and people with this goal have feelings under threats. In this study, users with performance-avoidance goals are worried and concerned about others’ reactions if they make posts. Furthermore, they may regard posting as a threat since their posts might be judged as poor contents. To avoid these unfavourable judgements, lurking is a good behavioural strategy for them not to disclose their incompetence in public.

Cognitive dissonance and lurking behavior

The result of this study revealed that there is a positive relationship between cognitive dissonance and lurking behaviour. The explanation is provided as follows.

Studies on cognitive dissonance theory indicated that dissonance may arise when individuals are exposed to attitude-discrepant information, which may come from friends, family members, or colleagues who express distinct attitudes from theirs (Festinger, 1957; Elliot and Devin, 1994). Many scholars argued that the concept of dissonance is an arousal state which people are motivated to avoid or abolish (Kiesler and Pallak, 1976). In fact, their reactions to dissonance are various. Generally, three common solutions are adopted to reduce dissonance: (1) change the attitude towards the behaviour; (2) adjust behaviour in accordance with the attitude; (3) justify the behaviour (Hinojosa et al., 2017).

In this study, when users access attitude-discrepant information from online forums such as user replies or only negative feedback is offered, their coping strategy to this discomfort is changing their behaviour from posting to lurking. After, their behaviour and attitude could be consistent. According to Hinojosa et al. (2017), the dissonance reduction way depends on which one, cognition or behaviour, is easier to change. The result showed that on online forums, compared with the other two solutions mentioned by Hinojosa et al. (2017), it is effortless for users to adjust their behaviour.

Conclusions

Implications

This research makes several contributions. First of all, although some studies have explored the issues of theme-oriented virtual communities (Chen and Hung, 2010; Yan et al., 2016), few discussed the topic of them based on its environmental feature, learning. These communities, especially professional forums like virtual communities of motorcycle modification, attract many users because of various types of information, the high quality of shared content, and frequent and friendly interactions among users. Based on its learning feature, the result of this study enriches the literature on theme-oriented virtual communities.

Secondly, this study argues that achievement goals motivate people to be lurkers or posters. The two-by-two achievement goal theory is developed from the motivation theory (Stoeber et al., 2008). It suggests that people are motivated by different goals and reasons to achieve their goals. The results of this study show not only users who desire to master their skills (mastery-approach goal) tend to browse, ask questions, and interact with other users, but also individuals who want to display their better achievement (performance-approach goal) incline to post. However, people who are worried about their incompetence (performance-avoidance goal) choose to lurk. The results can contribute to understanding lurking behaviour based on an achievement goal perspective.

The third contribution is this study adopts mastery-approach goals and mastery-avoidance goals together with the concept of help-seeking in acquiring knowledge and applies them to explain lurking on forums. This study emphasizes that people with different mastery orientations actively seek various help, so they are less lurking. This study provides a new perspective to examine lurking behaviour based on help-seeking literature. Furthermore, managers should also pay attention to achievement goals of members. Theme-oriented virtual communities ought to be designed to meet the needs of members with different achievement goals.

Lastly, although many studies have applied the cognitive dissonance theory to explain consumer behaviour, such as re-patronage behaviour and product returns (e.g., Powers and Jack, 2013; Kim, 2011), little is known about how dissonance occurs to users on online forums and how it affects online user behaviour, like lurking. This study extends the cognitive dissonance theory to theme-oriented virtual communities and demonstrates dissonance as a determinant of online user behaviour (lurking). Further studies may adopt this study as a foundation to explore online user behaviour within various virtual communities, such as archived social media or ephemeral social media.

Limitations

Virtual communities can be divided into theme-oriented virtual communities and social networking services. This study primarily focuses on the issue of lurking behaviour on the former. Therefore, the present findings may not be over-generalisable. Additionally, 98% of respondents are male in this study. The sample characteristic is similar to studies regarding motorcycle riders. For example, Stephens et al. (2017) explored the issue of motorcycle rider behaviour, and the sample characteristics in Stephens et al. (2017) showed that 89% of respondents are male. In the motorcycle context, the number of males is more than females. Although the results of this study are dominated by men, the results can still explain why people lurk on theme-oriented forums.

About the authors

Hsiu-Hua Cheng is an assistant professor at National Yunlin University of Science & Technology, Yunlin, Taiwan. She can be contacted at: chenghh@yuntech.edu.tw
Chi-Wei Chen is a master'd degree student at Chaoyang University of Technology, Taichung, Taiwan.

How to cite this paper