Sharing travel related experiences on social media – Integrating social capital and face orientation

Tourists can share their travel experiences or related information freely through social media platforms. The purpose of this research is to explore the factors associated with sharing travel related experiences on social media by travellers in the context of social capital theory and person-context interaction theory. Data were collected among consumers in Malaysia using a structured survey; 423 responses were generated. The findings show that the social capital dimensions (social ties, trust, reciprocity and shared vision) positively impact sharing intention on social media while identification doesn’t impact the sharing intention. The results also partially support the hypothesis that protective face orientation negatively influences social capital dimensions and sharing intention while acquisitive face orientation positively influences them. The authors shed light on the construction of social capital dimensions and face orientation and the relative importance of the sharing intention on social media.


Introduction
The recent emergence of social media has transformed the way we communicate with each other, how we discover product-related information, and more importantly, the way we express our product experiences and opinions with others. Previously, consumers relied on marketer-generated sources such as advertisements or sought advice from family and friends or looked for third-party certificates in their decision-making. Social media is now included among these sources and has become the most preferred mode for searching and exchanging product-related information in some instances (Kapoor et al., 2018).
Social media is considered a tool for facilitating communication mechanisms and bringing people together through sharing content, which is known as user-generated content (Balaji et al., 2015;Michaelidou et al., 2011). The Internet Users Survey (IUS) conducted by the Malaysian Social engagement continues to be the most frequent activity amongst Internet users, with 85.6% of Malaysian Internet users reportedly having access to social networking sites (MCMC, 2019).
The survey also stated that the majority of Internet users relied on the Internet for information seeking and had shared content online via social media. Among those who have shared content online, it was claimed that it was beneficial to the recipients or to raise awareness about an issue (MCMC, 2019). Upon realising both the practicality and opportunity along with recent finding, Tourism Malaysia is initiating the incorporation of social media and utilising digital marketing in promoting the Visit Malaysia 2020 (VM 2020) campaign.
The development of Internet and social media has not only changed the way people communicate, but also the way people plan their trip and share their travel experience with family and others.
Specifically, the proliferation of social media has significantly impacted tourism industry in the way travellers search, access, and utilize tourism information (Law et al., 2014;Oliveria, Araujo, & Tam, 2020;Ritchie et al., 2011). Social media allows travellers to produce, consume, and distribute travel-related information as well as collaborate with tourism service providers.
Furthermore, travellers may share public related travel information as well as private individual travel information on social media. As experiences shared by travellers on social media (traveller UGC) can influence many prospective travellers, it has gained increased interest in recent time from both tourism service providers as well as travellers. While past literature had focused on the impact of social media on tourism industry (Conrady, 2007;Gretzel & Yoo, 2008), recent studies have focused on understanding the motivations for the creation of traveller UGC, the impact of traveller UGC on prospective travellers attitude and behaviours, and tourism outcomes such as image and choice. The present study seeks to contribute to the later by investigating the motivations for sharing travel experience on social media.
As motivations are key determinants of behaviours (Deci & Ryan, 1985), an understanding of the factors that determine sharing of travel experience is important. The current literature, however, does not fully explain the factors that motivate travellers to share their travel experience on social media (Wang, Huang, Li, & Peng, 2016). In particular, given that experience sharing is a complex process that is influenced by many personal and social factors (Kang & Schuett, 2013;Leung & Chan, 2003), it is critical to understand the factors that determine intentions to share travel experience on social media platforms. Prior studies suggest that sharing travel experiences on social media platforms supports the development and maintenance of social capital as it helps connect with other network members and build social bonds (Belk & Llamas, 2012). However, prior studies indicate that personal factors may determine the social interactions or how much information is shared with network members (Kim, Kang, & Lee, 2018;Tsiakali, 2018). Thus, sharing travel experience involves an interplay between personal and social factors. Despite this scant attention is paid in literature on how personal factors determine the access to social resources on social media platforms from sharing travel experience.
The present study addresses this research gap by examining the role of travellers face orientation in the relationship between social capital and travel experience sharing intentions. Face is defined as the travellers cognitive response to social evaluations. While face was originally developed in the Chinese Confucian culture, an individual's concern for image in public is universal. The literature suggests that travellers can engage in behaviours aimed at gaining face (acquisitive) or saving face (protective) (Wang et al., 2016). Protective orientation travellers are vigilant, shy, and keep low profile, who believe that miscommunication or mistakes can cause them lose face. On the contrary, travellers with acquisitive orientation are attention seekers who have a high desire to establish a positive image through self-promotion and ingratiation (Wang et al., 2016). Thus, the decision to share travel experience on social media is a social dilemma. By not sharing travel experience, a traveller can save face from miscommunicating wrong information. On the other hand, a traveller can build his public image by sharing the travel experience with others and build social capital. Given this, an understanding of how a travellers face orientation influences their social capital and resources from sharing travel experience on social media platforms is critical for both practitioners and researchers.
The present study has two aims: first, to explore how social capital dimensions (structural, relational, and cognitive) influence travellers intentions to share travel experience on social networking sites, and second, to examine the moderating role of face orientation (protective vs. acquisitive). The contributions of the present study to the literature is threefold. First, the present study contributes to literature on motivations for sharing travel experience on social media (Oliveria, Araujo, & Tam, 2020). Second, the present study contributes to the literature on social capital and experience sharing by examining the role of social capital dimensions on sharing travel experience (Kim & Shim, 2018). Third, by examining the role of face orientation, the present study contributes to the literature on factors that motivate or hinder sharing travel experience on social media (Chang et al., 2020;Hung, Xiao, & Yang, 2013).
The following study develops and tests an integrated model of travellers intention to share their travel experiences on social media. Specifically, it integrates social capital and face orientation concepts in developing a research framework for sharing intentions on social media. Information system literature argues that the network of relationships among members on social media could provide the necessary resources to facilitate sharing (Amblee and Bui, 2011;Chu and Choi, 2011;Kasabov, 2015). Thus, the notion of social capital can be utilised to explain why customers might engage in sharing communication on social media. Social capital refers to the sum of assets or resources with a collective value of all networks of relationships (Nahapiet and Ghoshal, 1998). Chang andChuang (2011), andJin et al. (2015), have applied social capital dimensions as a means of investigating how social factors, including social ties, trust and others, influence a user's decisions or actions in connection with social media use for sharing activities.

Studies conducted by
Although previous studies have examined sharing intention on social media for travel related experiences (Bilgihan et al., 2016;Kim et al., 2017;Yuan et al., 2016), the effect of social capital dimensions on sharing travel-related experiences on social media has received scant attention.
Secondly this study focuses on the role of face orientation among the three dimensions of social capital and their impact on sharing intention for travel-related experiences on social media. The literature suggests that individuals can have behavioural goals aimed at protecting themselves from losing face, and there are two types of face orientation: protective face orientation (PFO) and acquisitive face orientation (AFO) (Chou, 1996;Ho, 1976). Face is a personal and situational factor that relates to the way individuals engage in interpersonal interaction (Leung and Chan, 2003). Huang et al. (2010) indicate that face is a two-edged sword whereby AFO has a negative impact on the intention to share while PFO has a positive impact. However, Ding (2011) and Wang et al. (2016) indicated that AFO has a negative impact on sharing while PFO does not play a positive role. Face orientation has received less attention, despite its importance for communication, negotiation and social interaction (Ding et al., 2017). Thus, unlike existing studies, the present work integrates social capital theory and person-context interaction theory to explore the effect on individual tourism-sharing intention on social media. This study enhances the understanding of travellers intentions to share their travel experiences on social media by focusing on the technological factor of social capital and individual factor of face-orientation.
The next section of this paper presents a critical examination of the literature and development of the study's hypotheses. This is followed by a description of the research design, data collection methods, and data analysis methods and techniques. Finally, the study results are presented, and theoretical and practical implications are provided. The paper concludes by highlighting emerging conclusions from the study and providing suggestions for future research.

Social Capital Theory
Social capital theory, introduced by Coleman (1988), refers to the resources gathered through relationships among people. This theory has been widely used in the social, organisation, and management domains. Social capital refers to the sum of assets or resources with a collective value of all networks of relationships (Nahapiet and Ghoshal, 1998). The idea is that knowledge and resources are exchanged, work gets done, and value is created through social relationships (Rottman, 2008). Social capital is embedded in relationships between individuals and their connections with their communities. Gu and Wang (2014) stated that social interaction is a necessary feature for the social network system in terms of sharing information and word of mouth.
Past literature suggests that social capital consists of three dimensions, namely, structural, cognitive, and relational. The structural dimension deals with the overall pattern of connection or relationships between individuals. It describes the impersonal configuration of linkages between individuals and the extent to which individuals are connected with one another (Nahapiet and Ghoshal, 1998). The structural dimension includes network ties, network configuration, and appropriable organisation (Nahapiet and Ghoshal, 1998). An individual with specific network ties may access valuable information before others do, obtain referrals, and enhance information value (Zheng et al., 2014).
The relational dimension refers to personal relationships between people that arise from interpersonal contact made through a history of interactions (Nahapiet and Ghoshal, 1998). The relational dimension captures the concepts of trust, norms, obligations, and identification (Nahapiet and Ghoshal, 1998). The relational dimension can affect individual behaviour by decreasing communication barriers, enhancing the intention to share and encouraging members to take action and achieve group goals by helping each other (Chang & Chuang, 2011;Pai & Tsai, 2016).
The cognitive dimension refers to the shared paradigm and refers to those resources providing shared representations, interpretations, and systems of meaning among parties (Nahapiet and Ghoshal, 1998). The cognitive dimension includes shared codes and language, and shared narratives (Nahapiet and Ghoshal, 1998). The cognitive dimension is articulated as a shared vision predicated on a common language, shared meanings, and an understanding of values and goals among the members of the network as they interact to share information (Chang & Chuang, 2011).

Face Orientation
According to person-context interaction theory, individuals and the environment or situation are not two separate entities but represent a continuous two-way interaction (Magnusson & Stattin, 1998). Based on this theory, knowledge sharing on social media is influenced by personal and situational factors. Face is a socio-cultural factor that was initially developed in Chinese Confucian society (Lin, 1936); however, the concept of face is universal in nature (Ho, 1976). Zhang (2011) mentioned that in order to maintain smooth and effective social interactions, individuals involved in the interaction are required to convey minimal acceptable public images of them as well to assist others in maintaining their social identities. According to Ho (1976), the face of an individual continually changes as it can be enhanced or diminished during social interaction. Researchers have realised the importance of face during social interaction, but prior research has only examined the effect of face in knowledge sharing in the Chinese context (Huang et al., 2008;2010;Ding et al., 2011;Wang et al., 2016). Face orientation consists of two dimensions, namely protective face orientation (PFO) and acquisitive face orientation (AFO), which are described according to positive or negative evaluations (Ho, 1976;Wang et al., 2016). PFO is also known as face-gaining and refers to a behavioural orientation that maintains and promotes an individual's positive image and social-esteem from others. By contrast, AFO, also known as face-saving, refers to a behavioural orientation to avoid a negative evaluation and retain social standing in all potentially damaging situations.

Structural Dimension: Social Interaction Ties
Social interaction ties refer to the network-related social relationships that describe the pattern of linkages in terms of density, hierarchy, and connectivity. Social interaction ties are channels for information and resource flow. Structural dimension reflects the impersonal properties of network relations while social interaction ties represent the basic element of communication networks.
Sharing intention is achieved when networks have strong connections and direct ties between their members. Studies have shown that social interaction ties have a positive relationship with sharing intention in social networking sites (Tan, 2013 andPrasetio, 2014) and positively influence the intention of consumers to engage in electronic word-of-mouth (eWOM) on social networks (Wang et al., 2016).
When consumers perceive social interaction ties as strong, they may be more willing to share information. They feel strong ties are closer and important for each other to share information that can influence each other's decisions (Chu and Kim, 2011). Prasetio (2014) also stated that the level of social interaction ties affects willingness to share information such that the more consumers feel tied to each other, the more likely they are to share information. Studies by Chai and Kim (2011) and Ho et al. (2012) indicate that initiative to share information increases when a close real-life friendship exists on social networking sites. In contrast, when consumers perceive social interaction ties as weak, they tend to share less information. They evaluate that information shared with weak ties will have less influence on their decisions. Chances of exchanging and sharing information exist when there is cooperation between consumers, while in weak ties, cooperation is low, which leads to less information sharing (Tohidinia and Mosakhani, 2010). Further, consumers with strong ties share credible information with others more frequently, while consumers with weak ties are likely to share narrative information (Wang et al., 2016). It is expected that sharing intention is high when social interaction ties are strong and vice-versa. Thus, we hypothesise that: H1: Social interaction ties positively impact sharing intentions on social media.

Trust
Trust is known as a set of specific beliefs dealing primarily with integrity, which refers to an individual's expectation that members in a group will follow a generally accepted set of values, norms, and principles. Trust is an enabler for social exchange and cooperation, and it opens up people for knowledge sharing. Previous studies have found that trust is significantly related to information exchange and knowledge sharing (Chai et al., 2011;Gu and Wang, 2014) and electronic word of mouth (Choi and Scott, 2013;Chu and Kim, 2011).
When relationships are trustworthy, members of the social network are more willing to share their information since they have no fear of being exploited by the other members (Chang and Chuang, 2010). Meanwhile, when relationships are low on trust, members in the social network tend to share less information. Cao et al. (2013) explained that trust is a stronger predictor of information sharing, because social media includes loosely connected individuals with different backgrounds and fewer interactions and thus, removing communication barriers is important to further information sharing. According to Cao et al. (2013), trust allows consumers to evaluate the source and value of information and justify their decision to share their information. Chu and Kim (2011) stated that a consumer's willingness to rely on their social connections is enhanced due to perceived reliability and trustworthiness. Besides, trust also promotes the use of information since it enhances the credibility of the information. Chang and Chuang (2010) found that trust has a positive effect on the quality of knowledge sharing and does not have an effect on the quantity of knowledge sharing. It is expected that consumers share more information when trust is high and vice-versa. Thus, we hypothesise that: H2: Trust positively impacts sharing intentions on social media.

Identification
Identification is one's conception of self in terms of the defining feature of the self-inclusive social category, the process whereby individuals see themselves belonging to a set of people or group. It also includes a member's sense of belongingness and positive feelings towards a social network and explains the readiness to remain an active member of the network. It reflects individuals' efforts to express and present themselves to others, and it enhances the degree of knowledge contribution (sharing of information, knowledge creation, and learning).
Choi and Scott (2013) found a positive relationship between identification and electronic word-ofmouth (eWOM) quality and information sharing. When consumers perceive high identification, they may be more willing to share information. Consumers perceive opportunities for exchanging information with others as well enhancing the actual frequency of cooperation when identification is high. Another study by Chang and Chuang (2010) also showed that identification had a significant and positive effect on knowledge sharing. They also stated that users on social media are based on existing networks or profiles with similar characteristics between them, which in turn increases identification. Reliability of a consumer's opinions or information increases when identification increases.
In contrast, when consumers perceive low identification, they tend to share less information.
Valuable knowledge is embedded in individuals and people would not contribute their knowledge unless another person is recognised as their group-mate and the contribution is conducive to their welfare. The perception of social unity and togetherness of the community will elevate people's activeness to share and increase the depth and breadth of shared knowledge. It is expected that consumers share more information when identification is high and vice-versa. Thus, we hypothesise that: H3: Identification positively impacts sharing intention on social media.

Norm of reciprocity
Reciprocity norm refers to knowledge exchange that is mutual and perceived by both parties as fair. It is expected that knowledge sharing in the social network by a member is induced by the expectation that others would reciprocate their requests when required. If the invested efforts in knowledge sharing can be reciprocated, members are motivated to contribute more. Prior studies have shown that the reciprocity norm has a significant impact on consumer information-sharing behaviour (Chang and Chuang, 2011;Liu, Cheung and Lee, 2016).
When consumers perceive the reciprocity norm as high, they may be more willing to share information. Consumers who receive appreciation from others for their contribution become even more motivated to continue contributing (Liu, Cheung and Lee, 2016). When consumers perceive a low reciprocity norm, they tend to share less information. Consumer motivation to share information decreases when they receive a lower response from others.
When a consumer provides his/her opinions or information to others, consumers who receive it are indebted to give worthy feedback to the information giver (Hung, Lai and Chang, 2011). As mentioned by Pai and Tsai (2016), this enhances mutual interaction as consumers who receive responses are willing to interact even more with others and share more information. Furthermore, information feedback and mutual benefits in the future exist when consumers reciprocate comments from other consumers (Yang and . It is expected that consumers share more information when the reciprocity norm is high. Thus, we hypothesise that: H4: Norm of reciprocity positively impacts sharing intentions on social media.

Cognitive Dimension: Shared Language
Shared language is a tool for the evaluation of the benefits of knowledge exchange. Mutual knowledge can increase the likelihood of understanding between members because it allows them to formulate their knowledge as a result of knowing what the receiver does and does not know (Chang and Chuang, 2010). Studies have shown that shared language has a significant positive relationship with knowledge sharing (Cao et al., 2013;Prasetio, 2014) and eWOM (Choi and Scott, 2013).
When consumers perceive shared language as high, they may be more willing to share information.
Consumers are willing to share their knowledge when they have the same vision as others (Prasetio, 2014). Consumers feel that shared language enhances their ability to access information as well as to share information as it reduces both encoding and decoding efforts (Cao et al., 2013).
Besides, consumers also think communication efficiency and mutual understanding is enhanced by shared set of vocabulary and terms.
In contrast, when consumers perceive shared language as low, they tend to share less information.
Consumers feel individuals who do not share a common language understand each other less, and this might result in misunderstandings in communication (Liu, Cheung, and Lee, 2016). Further, Chang and Chuang (2010) stated that shared language motivates the network members to indulge in knowledge exchange activities actively and enhances the quality of knowledge shared in the network. It is therefore expected that consumers share more information when shared language is high and vice-versa. Thus, we hypothesise that: H5: Shared language positively impacts sharing intentions on social media.

Revisit Intention
Revisit intention refers to the individual's judgment about visiting a designated service from the same company again, considering their current situation and likely circumstances (Hellier et al., 2003). According to Wang et al. (2012) tourists′ revisit intention has been important in the determination of tourist destination. Higher revisit intention was discovered when travellers were attached to an organisation social media page (Leung and Bai, 2013). Wang and Wu (2011) found that higher involvement of travellers induced higher intention to revisit a museum. An individual's higher involvement on the Internet brought about a higher level of intention to purchase travel products through the Internet (Moital et al., 2009). It is expected that consumer revisit intention depends on sharing intention on social media. Thus, we hypothesise that: H6: Sharing intentions on social media have a positive impact on revisit intention.

Well-being
Subjective well-being refers to hedonic perspectives, such as positive affect (happiness), lack of negative affect, and life satisfaction (Ryan and Deci, 2001). In short, well-being is gained when some state, such as a goal, has been achieved. Prior research has examined the relationship between the use of SNSs and individuals' well-being (Ellison, Steinfield and Lampe, 2007;Valenzuela, Park & Kee, 2009). In this study, subjective well-being is defined as an individual's own assessment of their happiness and satisfaction with their social life on social media (Chiu et al., 2013). According to self-determination theory, individuals are self-determining and intrinsically motivated to continue using social media and saying positive words about it when they feel happy about doing so and vice-versa. Thus, we hypothesise that: H7: Sharing intentions on social media have a positive impact on well-being.

Protective face orientation (PFO)
Protective face orientation (PFO) refers to the fact that some individuals with behavioural goals aim to protect themselves from losing face (Wang et al., 2016). Individuals with high PFO are modest, cautious, focused on getting along with others, and inclined to keep a low profile (Chou, 1996). PFO can act as a barrier preventing tourists from sharing knowledge or information about their travel experiences with others on social media. According to Ardichvili et al. (2006), individuals with high PFO will never publicly disclose their weaknesses or misunderstandings because it could cause them to lose face and be embarrassed or even disrespected by others. Wang et al. (2016) showed that PFO weakens the impact on knowledge sharing, especially among those who have higher PFO. Besides, Ding et al.'s (2017) study also showed that PFO is a negative social orientation. The authors further state that PFO is negatively motivated by psychological factors because of the desire to avoid negative evaluation as well as to retain social standing in all potentially damaging situations. Thus, we hypothesise that: H8: Protective face orientation negatively moderates the relationship between social capital dimensions and sharing intention on social media.
H8a: The relationship between social interaction ties and sharing intention on social media is weaker for tourists with high PFO.
H8b: The relationship between trust and sharing intention on social media is weaker for tourists with high PFO.
H8c: The relationship between identification and sharing intention on social media is weaker for tourists with high PFO.
H8d: The relationship between reciprocity and sharing intention on social media is weaker for tourists with high PFO.
H8e: The relationship between shared language and sharing intention on social media is weaker for tourists with high PFO.

Acquisitive face orientation (AFO)
Acquisitive face orientation (AFO) is characterised by attention-seeking through ingratiation, selfpromotion, and exemplification (Wang et al., 2016). According to Chou (1996) H9d: The relationship between reciprocity and sharing intention on social media is stronger for tourists with high AFO.
H9e: The relationship between shared language and sharing intention on social media is stronger for tourists with high AFO.
Based on the review of the literature and hypotheses developed, the following model is proposed (see Fig. 1).

Methodology
The data were collected through a survey using a structured questionnaire from respondents in Malaysia. The reason is that, in recent years, Malaysia has experienced an increase in Internet usage, especially on social media. As a result, Tourism Malaysia has focused on a social media campaign to promote Malaysia 2020. It is important to understand Malaysian tourists' sharing behaviour of their travel experiences on social media, as such understanding will further help Tourism Malaysia to develop effective social media strategies. The respondents were chosen from a Malaysian population using a purposive sampling technique. This technique was used to select respondents who had travelled recently and were active on a social media platform. Table 1 provides demographic statistics of the sample data. A total of 423 respondents were collected, 207 were for a positive travel experience and 216 were for a negative travel experience. This sample size was adequate to represent the population in Malaysia at a 95% confidence level, according to Krejcie & Morgan's (1970) sample table.
The questionnaire encompassed six demographic (gender, age, education level, occupation) items, seven SNS usage items, five tourism-related activity participation items, and a total of 42 items for nine factors. Items used to measure study constructs were based on indicators from prior research studies. Social ties, trust, reciprocity, identification, and shared vision and language, which are sub-dimensions of structural, relational, and cognitive dimensions of social capital, were The variable revisit intention has been measured using three items developed by Kim and Cho (2013) and Liu and Lee (2016). Four items for the well-being were adapted from Chiu et al. (2013) study. Lastly, 12 items for PFO and AFO, which are sub-dimensions of face orientation were adopted from Wang et al. (2016). All of the items were rated on a 7-point Likert-type scale in which 1 represented 'strongly disagree' and 7 represented 'strongly agree'. To reduce potentially confounding effects, this study controlled for several common background variables known to impact user's intention to share travel-related experiences. In the present study, we controlled for respondents' gender, age, education, and occupation.
From the sample, 63.6% of the respondents (n=269) were females and the majority of the respondents (75.9%) were aged between 19 and 29 years. In terms of educational background, 81.3% had a bachelor's degree. Regarding occupation, 64.8.5 % (n=274) were students, 33.1 % (n=140) were employed in private and public firms. As per the data sample, 68.3% of the respondents chose Facebook as the most used SNS while Instagram was the second most used SNS with 23.6%. Table 1 reports the description of the final sample.
[ Table 1 near here] In order to analyse the data, a partial least squares (PLS) approach to structural equation modelling (SEM) was used to test the various relationships. Specifically, the SmartPLS 3.0 program was used to estimate the model's parameters (Ringle et al., 2015). SEM is a 'quasi-standard' in marketing research, which enables researchers to test complete theories and concepts (inner or structural model and outer or measurement model) (Hair et al., 2011).
In addition, PLS path modelling becomes an essential method in empirical research Sarstedt et al., 2011). PLS-SEM is a 'causal modelling approach aimed at maximizing the explained variance of the endogenous latent constructs' (Hair et al., 2011, p. 139  First, confirmatory factor analysis (CFA) was carried out to assess the measurement model.

Structural Equation Modelling (SEM) was performed to assess the hypothesised relationships.
Common method variance (CMV) threatens the reliability of the findings and it occurs when the data are obtained from a single source survey (Reio, 2010). We addressed this issue (CMV) by using Harman's single factor test (Podsakoff et al., 2003) and Lindell and Whitney's (2001) Marker Variable method. According to Harman's single factor test, an exploratory factor analysis was performed. Results showed the first factor explained 40.22% of the total variance, lower than the suggested 50% threshold. These tests did not indicate a single factor structure that accounted for the majority of the variances, suggesting common method bias is not a major issue in this study.
Furthermore, based on the method proposed by Lindell and Whitney (2001), a marker variable was introduced in the PLS model to assess the common method bias by determining the correlation between the marker variable and the latent variables. The correlations between the latent variables and the marker variable was less than 0.3. This suggests that common method bias is not a problem in this study.

Results
The two-step modelling approach recommended by Anderson and Gerbing (1988) was used to assess the constructs' measurement properties (the outer model) and the structural (the inner) model. Measurement properties were examined first to assess reliability and validity, after which the structural model was estimated. The structural model was used to test hypotheses H1-H9.

Confirmatory factor analysis (CFA) & constructs' measurement properties
At first, confirmatory factor analysis (CFA) was conducted to test the validity of the model. Initial analysis revealed that five loadings, namely social interaction ties (ST3), identification (ID4), revisit intention (RV1), protective face orientation (PFO3) and acquisitive face orientation (AF01) were not satisfactory (less than 0.6) and were dropped from further analysis. Following this, the CFA shows that the item loadings on their respective latent factor were satisfactory (greater than 0.7) and statistically significant (p < 0.01) (see Table 2).
Cronbach's alpha and composite reliability test were used to test the reliability of study constructs. Table 3, the Cronbach's alpha and composite reliability values were more than the recommended threshold level of 0.7. Thus, the measurement scales were internally consistent and reliable.

As shown in
An evaluation of the factor item loadings and average variance extracted (AVE) was undertaken to check the convergent validity of the construct measures. As shown in Table 3, the AVE values greater than the 0.50 level, confirming the convergent validity of the measures (Fornell and Larcker 1981;Hair et al., 2017). Further, discriminant validity was inspected by comparing the shared variance between factors with the square root of the AVE (Fornell and Larcker, 1981). Table 4 confirms that the first was lower than the latter. Thus, discriminant validity was achieved. Further, multicollinearity was not present in the findings as variance inflation factors (VIFs) were all beneath the cut-off value of 5.00. [

The Structural Model
A bootstrapping procedure with 5000 resamples was used to estimate the paths' significance (Hair et al., 2011) in the structural model. The estimated model was evaluated using a number of indices, including R2 values, path coefficients, bootstrapping critical ratios, and a redundancy analysis.
The usefulness of the model was established by combining predictive relevance and the strength of the path coefficients. The variance explained by the model in terms of R2 was 0.589 for intention to share, 0.039 for revisit intention and 0.599 for wellbeing. Hence, the R2 for intention to share and well-being can be classified as moderate while R2 for revisit intention was weak. For the completed model, we obtained a SRMR value of 0.071 and NFI value of 0.658. SRMR values less than 0.10 or of 0.08 (Hu and Bentler, 1999) are considered a good fit. The results suggest the model was a good fit to the data.
The individual path coefficients of the PLS structural model can be interpreted as standardised beta coefficients of OLS regressions. The results of the structural model analysis are shown in Figure 2 and presented in Table 5 as well an overview of results in Table 5.
[ Figure 2 here] [  predicted that identification exerts a positive direct influence on sharing intention on social media. This was not supported (β = 0.003, p = 0.995). H5 proposed that shared value exerts a positive direct influence on sharing intentions for travel experiences on social media. The results support the positive significant relationship between shared value and sharing intentions (β = 0.209, p < 0.05). The results for H6, which suggest a positive relationship between sharing intention on social media and revisit intention, was supported (β = 0.167, p < 0.05). Lastly, H7 that predicted a positive relationship between sharing intention on social media and well-being and this was also supported (β = 0.773, p < 0.05). In summary, Hypotheses 1, 2, 3, 5, 6 and 7 were accepted.

Moderation Analysis
H8 predicted a moderating relationship between protective face orientation and social capital dimensions on intention to share travel experiences on social media. As shown in Table 5, the moderating effect of social interaction ties and PFO had a negative significant impact on sharing intention on social media (β = -0.347, p < 0.05). The results for Hypotheses 8b to 8e were not supported as the moderation effect of PFO between the social capital dimensions and sharing intentions on social media were not significant. H9, which proposed the moderating relationship between AFO and sharing intention on social media, was partially supported. H9a was supported as the relationship between social interaction ties and sharing intention on social media was significant for AFO (β = 0.353, p < 0.05). However, Hypotheses 9b to 9e were not supported as the moderation effect of AFO between social capital dimensions and sharing intention on social media were not significant.

Discussion
This study integrates social capital theory and person-context interaction theory to explore the effect on individual tourism-sharing intention on social media. More specifically, the research clarified the influence of social capital dimensions (social ties, trust, reciprocity, identification, and shared value) and face orientation (protective face orientation and acquisitive face orientation) on the intention to share travel-related experiences on social media. The effect of consumer intention to share travel-related experience on social media was examined via nine hypotheses. The findings of this study provide valuable insights about customers' sharing intention on social media.
The present study indicates that structural, relational, and cognitive factors are dimensions of social capital. Social ties contributed positively to the intention to share travel-related experiences on social media. This result support previous findings that suggest a positive effect of social ties and sharing information (Chai and Kim, 2011;Ho et al., 2012;Wang et al., 2016). Such findings indicate that their social interaction ties drive the customer's knowledge or information sharing on social media. This further explains that customers feel closer and more important to each other in terms of sharing information when social ties are strong. Trust and reciprocity enhance the intention to share travel-related experiences on social media. The results suggest that consumers form intentions to share their travel experiences based on trust with social members and reciprocity they receive from members. These results support previous findings that suggest a positive effect of trust (Choi and Scott, 2013;Chu and Kim, 2011;Gu and Wang, 2014)  On the other hand, customers are more willing and motivated to share information when reciprocity norm is high because they receive appreciation from others for their contribution.
Shared value also enhances the intention to share travel-related experiences on social media. These results support previous findings that suggest a positive effect of shared value and information sharing (Cao et al., 2013;Choi and Scott, 2013;Prasetio, 2014). Shared value or language enhances their ability to access information as well to share information as it reduces both encoding and decoding efforts (Cao et al., 2013). Customers feel members who share common language and value, understand each other more, and might not end up in misunderstandings of communication.
Although all factors from three dimensions showed significant results, identification was not statistically significant. This finding contradicts research findings within the existing literature, where identification is a necessary foundation for knowledge sharing (Chang and Chuang, 2010;Choi and Scott, 2013). However, a study by Yu et al. (2010) showed that identification was not a significant factor that influences knowledge-sharing behaviour. This result is further explained by the fact that members in social media primarily viewed their community as an informational and social universe, rather than as a commercial entity. Identification is weak within this community because of alterable online personalities and social roles. The results extend the current knowledge to the fundamental social capital dimensions that influence the intention to share travel-related experiences on social media.
Prior evidence suggests that higher revisit intention Bai, 2013) andwell-being (Chiu et al., 2013) were discovered as travellers became attached to social media experience sharing.
Thus, this study also indicates that intention to share travel-related experiences on social media influences revisits intention as well as an individual's well-being. This finding is consistent with the previous literature by showing that in a consumer behaviour context, the higher consumers' intention to share information is, the higher the revisit intention (Leung and Bai, 2013;Moital et al., 2009;Wu 2011) andlevel of well-being (Chiu et al., 2013;Ellison et al., 2007;Valenzuela et al., 2009).
Finally, this study indicates that face orientation moderates the relationship between social capital dimensions and intention to share travel-related experiences on social media. The result shows that PFO and AFO are only significant for the structural dimensionsocial ties. These results are supported by a previous study done by Ding et al. (2017) that shows that PFO is a negative social orientation while AFO is the positive social orientation (Leung et al., 2014). This result is explained by the fact that opinion leaders would feel confident in their ability to make any individual product choice. Since opinion leaders have high levels of product expertise, they conduct little external search for information. AFO did not have a significant effect on the relational and cognitive dimensions. This corroborates the findings from Ardichvili et al. (2006), Ding et al. (2017 and Wang et al. (2016). It also further explains that even with high AFO, individuals are conscious of gaining face but they are not willing to share their travel experiences.

Theoretical and Practical Implications
From a theoretical perspective, this study enriches the current literature on influential factors that determine the tourists' intentions to share their travel experiences on social media. We do this by adopting social capital theory and person-context interaction theory. Notably, the theoretical relationships were empirically supported. The present study successfully confirmed the factors that influenced tourists' intention to share travel-related experiences on social media. Most importantly, the findings emphasise the dimensions of social capital and face orientation on attitudinal and behavioural intention. The results of the present study might be considered one of the emerging attempts to empirically investigate tourists' intention to share travel-related experiences on social media, especially in the Malaysian context.
Regarding the practical contribution, the findings of the present study provide insights on tourists' behaviour of sharing their travel experiences on social media. Social media is essential to effectively reach consumers who favour the use of technology-based and internet-based tools.
Travel companies need to understand the influential role of social media in tourism as they relate to driving tourists' decisions by informing and inspiring others. Travelers use social media to plan their vacations, and they trust the recommendations and suggestions of their virtual network.
According to Law et al. (2015), users communicate the messages to others in social media marketing, which is different from traditional marketing where the company directly sends marketing messages to their customers. Thus, it is important to understand information-sharing behaviours, which are a powerful asset to a business. Industry players must also be aware of the power of eWOM. When there is a positive feedback/ comment regarding the tourism destination, it will spread rapidly, and vice versa for negative feedback.
Thus, eWOM is of utmost importance as the feedback may be constructive as well as destructive to the business. Players, who use social media as a communication tool with customers, need to persuade customers to share their experiences. By encouraging their customers to share their tourism-related opinions, reviews, and experiences, it helps the firm to improve its service quality as well as to attract new customers and retain their existing customers. Besides that, the findings also help managers to understand the process by which social capital contributes to information sharing. Firms must engage in social interaction with their customers in order to receive feedback.

Limitations and Future Research
The present study has several limitations. It should be noted that this research was conducted in the Malaysian setting, and hence, the findings of the study may not represent respondents from other geographical areas. Therefore, there is a need to replicate this kind of research in other geographical settings in order to enrich the literature. Future study may be expanded to a broader geographical area to investigate the differences between the findings established in other geographical locations. Besides, a cross-sectional approach was employed in the present study, in which the data were collected at a single point of time (i.e., in a month). Consequently, it is also recommended to employ a longitudinal study in future research. Apart from that, this study did not consider the role of personality and psychological traits as a moderator for intention to share travelrelated experiences on social media. Therefore, future research looking into other possible moderators to obtain a comprehensive understanding of the drivers of behavioural intention is deemed necessary.

Conclusion
Overall, the proposed model of the present study shed some light on enhancing the sustainability of the Malaysian tourism industry. From the outcomes of this study, it is evident that the study has provided several significant and useful contributions to consumer behaviour literature, especially in the context of tourism. This study contributes both theoretical and practical implications by identifying the importance of social capital dimensions in influencing sharing travel-related experiences on social media.