Understanding Social Media Usage at Work from the Perspective of Social Capital Theory

: With the pervasive use of social media (SM) in organizations, it is regarded as a relevant driver that can inﬂuence an employee’s job performance. This study ﬁlls in the gap that extends the job performance concept by discovering the role of SM in innovative performance in introducing new ideas beyond standard speciﬁcations to produce novel and valuable organizational outcomes. By adopting the social capital theory (SCT), the present study investigates the roles of social media use at work in predicting social capital (network ties, shared vision, and trust) that might promote work engagement and subsequently affect employees’ innovative job performance. The data was collected through an online survey


Introduction
Social media (SM) have increasingly become a valuable platform for communication and facilitate knowledge sharing for personal and work-life (Ahmed et al. 2019).Organizations have widely adopted it for work-related purposes (Chu 2020).Apart from individual usage, SM has been progressively embedded in the work environment.Organizations have been implementing SM tools in new management practices, starting from creating innovative business models to facilitating knowledge sharing, communication, and collaboration (Cao and Yu 2019).In addition, companies have been deliberately using such social media tools to support their employees in enhancing team and employee performance, as well as improving their business activities (Song et al. 2019;Braojos et al. 2019).
Due to the digital era, scholars have argued that face-to-face interaction in the organization was replaced by online interaction, specifically through social media, resulting in new online social capital (Park et al. 2013).Thus, several studies have explored the consequences of SM use on social capital in employees' social network ties (Huang and Liu 2017;Park et al. 2013;Sheer and Rice 2017;Yen et al. 2020).Hence, social capital has become a key variable in understanding the use and implication of SM, especially in an organizational setting.
Social capital has provided a foundation for SM usage and employees' social relations in organizations by focusing on its unique benefits on job performance.This framework was extensively used and well established, and has been widely adapted in SM-related studies e.g., (Ali-Hassan et al. 2015;Cao et al. 2016;Ghorbanzadeh et al. 2021;Kwon 2014;Tijunaitis et al. 2019).Moreover, the three dimensions of social capital by Nahapiet and Ghoshal (1998) align well with the complexity of SM usage in organizations due to its multi-faceted conceptualization.Ali- Hassan et al. (2015) mentioned that the cognitive and relational dimensions describe a person's ability, while the structural dimension focuses on the availability of resources.In fact, these three dimensions are obtained through employees' relational networks (Yan and Guan 2018).
Generally, findings from the link between SM use and job performance have been somewhat inconclusive and even contradictory.For instance, some researchers found SM usage in the workplace to be significantly associated with positive outcomes in employees' job performance (Eid and Al-Jabri 2016;Jafar et al. 2019;Lee and Lee 2020;Song et al. 2019), while other studies state that SM use has led to a deterioration in employee performance (Brooks 2015;Yu et al. 2018;Zhang et al. 2015).Hence, the inconsistent and uncertain findings of the existing studies do not explain the real impact of SM use at the workplace, whether SM usage improves or harms the employees' innovative job performance.Furthermore, the relevance of SM usage in the workplace has been hotly debated by scholars; whether the use of SM during work time, either for personal or work-related purposes, should be allowed, disallowed, or tolerated (Chu 2020).
Central to the present study, Yen et al. (2020) stated that studies of the exact mechanism of social capital and job performance through SM use for interaction between workers and coworkers are still scarce.The understanding of social media as social capital is still constrained by some persisting gaps in the social media literature, especially in innovative job performance (Chen et al. 2019;Ali-Hassan et al. 2015).In addition, the scope of generalizing research findings based on previous research is limited due to sample bias, affordances of SM platforms (Sheer and Rice 2017;Yen et al. 2020), and cultural differences (Cao et al. 2016).Due to the limitations, the present study aimed to explore the benefits of SM use at work on employees' innovative performance by utilizing social capital theory (SCT), and the objective of the present study is suited to the purpose of SCT.
Therefore, this study extends the model of previous studies (Ali -Hassan et al. 2015) by incorporating work engagement in order to gain a deeper understanding of how SM use at work can foster a deep connection with social capital that directly enhances work engagement and consequently, affect employees' innovative job performance.As Gibbs (1990) stated, social capital theory aims to explain the influence of people's interaction in obtaining psychological and tangible benefits.In addition, social capital is relevant to collectivistic cultures due to the emphasis on social relationships in daily interactions, including during working hours (Sheer and Rice 2017).Furthermore, the addition in this study of a new variable, work engagement, can be a good predictor of an employee, team, and organizational outcomes (Bakker and Albrecht 2018).Clausen et al. (2019) stressed that social capital is an essential element in developing employee engagement, specifically work engagement.In addition, lack of studies on SM use and its influence on employee job performance in Malaysia (Radhakrishnan et al. 2018), social capital is one of the critical variables in understanding the impact and outcome of SM use at work (Sheer and Rice 2017).Consequently, the integration of SCT and work engagement will contribute to the existing body of literature in the area of SM usage and job performance from the perspective of Malaysian culture, as Malaysians in general are more collectivistic in nature (Sumaco et al. 2014;Jayasingam et al. 2021).

Social Media Use at Work
Generally, there are two major purposes of SM use by the employee while at work, which are personal or work-related reasons.Though researchers have addressed the essential use of SM for work-related reasons, the negative consequences of SM use on employee outcomes also have been identified in several studies.Initially, the use of SM for work purposes contributes to positive work morale, but excessive use of SM may lead to negative consequences on employee performance and morale (Demircioglu and Chen 2019;Yu et al. 2018).In addition, the high frequency of social media usage during work hours may unintentionally cause a pressured environment in the workplace (Bucher et al. 2013).Furthermore, employees who connect with SM for work, especially after regular working hours, could diminish the availability of time and energy of employees in fulfilling their life-role's demands, and thus eventually generate both time-based and strain-based conflicts between their work domain and life domain (Chu 2020).
Basically, the work-related use of SM mainly means for performing work, facilitating knowledge sharing (Ahmed et al. 2019;Jafar et al. 2019), work communication between employees and organization (Zhang et al. 2019), creating content for work or collaboration with colleagues (Ali- Hassan et al. 2015), developing and strengthening contact with stakeholders for work-related issues (van den Berg and Verhoeven 2017).Furthermore, a recent study by Ghorbanzadeh et al. (2021) discovered that SM usage has positively predicted social capital and subsequently enhanced employees' job performance.Hence, the purpose of SM usage for work-related reasons is essential to achieving organizational goals and performance enhancement.Therefore, we believe using SM in a workplace for work-related purposes has many benefits on employee performance.
In the context of this study, social media referred as a group of internet-based applications that allow users to communicate, build connections and facilitate ideas by creating, sharing, and exchanging information in other formats with multiple communities (Ahmed et al. 2019;Sheer and Rice 2017).Employees and organizations use various social media platforms for work, personal, or both purposes.This study focuses on using public social media accounts and Instant Messages (IM) to establish network ties, share information, and perform tasks or duties.
The public social accounts in this study are related to Facebook, Twitter, and LinkedIn.These platforms have been discovered to be the channel most frequently used by employees for work purposes and professional connections (Davis et al. 2020;Pekkala and van Zoonen 2022;Zoonen and Treem 2019).These platforms are typically individually owned accounts characterized as publicly available media that are free and easily accessible.In addition, Facebook, Twitter, and LinkedIn are more suited for work purposes due to the essential technical and textual features which help employees obtain or share work-related information and expand professional connections rather than using Instagram or Snapchat (Zoonen et al. 2017).
Instant Messaging (IM), particularly WhatsApp messenger and Telegram, has become an essential formal medium for work interaction in organizations.Recently, employees have used WhatsApp Messenger and Telegram extensively in Malaysia for work matters.These messengers enable employees to be quicker and more effective in sharing and discussing new ideas or products with their colleagues directly and help them engage with and manage remote teams.In addition, WhatsApp Messenger and Telegram features are user-friendly and easily adapted by employees from different age groups to accomplish individual or group tasks.
Therefore, this study measures SM use at work through Facebook, Twitter, Linked In, WhatsApp, and Telegram to discover the advantages SM use has employee's innovative job performance in Malaysia.

Underpinning Theory
According to Nahapiet and Ghoshal (1998), organizational social capital is a strategic resource and a multidimensional concept which can be divided into three dimensions, namely cognitive, relational and structural dimensions.The dimensions of social capital relate to the general nature of the relationship between organization members.The cognitive dimension refers to a common understanding and view among organization members that share similar goals, visions, and languages, thus enabling collective understanding (Nahapiet and Ghoshal 1998;Ali-Hassan et al. 2015).The relational dimension is described as resources such as trust, commitment, and reciprocity embedded in a personal relationship that focus on a particular relationship such as friendship and respect, in which individuals are willing to engage in exchange (Cao et al. 2016;Nahapiet and Ghoshal 1998).The structural dimension relates to the network ties between organization members and provides access to communication (Nahapiet and Ghoshal 1998).Each of these dimensions of social capital constitutes an aspect of the social structure that facilitates the exchange and combination of resources within an organization (Hauser et al. 2016).These three dimensions must be fulfilled in order to realize social capital in the workplace.Kwon and Adler (2014) stated that social capital must be recognized as a multidimensional concept to improve the understanding of social links because it has a value in both relations and cognition that go beyond the structural dimension.Thus, the present study conceptualized social capital developed by Nahapiet and Ghoshal (1998) that comprised three dimensions, namely the cognitive (shared vision), the structural (network ties) and the relational (trust).Hence, the three dimensions of social capital are essentially needed to examine the connection between SM use and job performance because, as Swanson et al. (2020) stated, social capital is realized when organizational members have common cognitive, relational and structural characteristics to attain the larger goals of the organization.
This study is essential to emphasize the role of SCT in the link between SM uses at work, social capital, and employee job performance in Malaysia.Moreover, several scholars stressed the importance of SM in creating social capital among employees (Men et al. 2020), as social capital has a vital role in influencing employees' job performance, especially in innovative performance (Cappiello et al. 2020;Yan and Guan 2018).Based on dimensions of social capital by Nahapiet and Ghoshal (1998), the present study employed network ties, shared vision, and trust to gain a comprehensive view of SM use at work in influencing employees' social capital and job performance.

Network Ties
With the pervasive use of technology in the digital era, the development of network ties in organizations can be accessed through SM use at work.Moreover, Cao et al. (2016) stated that social media can help employers connect with employees who share similar backgrounds and interests, which is helpful in strengthening ties with colleagues, maintaining a professional network, and discovering potential relationships.Unlike regular physical contact, SM usage can assist employees in establishing a relationship with colleagues, primarily when they work in a different place or remotely at any time (Sheer and Rice 2017).The resources included emotional, financial, information support, mutual trust, and others (Cao et al. 2016;Yen et al. 2020).Having network ties with various organization members is essential for employees to obtain resources at work.By establishing network ties in the workplace, SM enables them to reach out for support (e.g., emotional, informational, and social) from their colleagues when performing work-related tasks.

Shared Vision
The aspect of the cognitive dimension of interest in this study is a shared vision.A shared vision is an emergent state that develops in a team when members have access to the same information and share the same tools, work processes, and work cultures (Hinds and Mortensen 2005;Ali-Hassan et al. 2015).A shared vision facilitates amongst employees the sharing of common collective goals and aspirations of an organization that can quickly be achieved through collaboration (Chang et al. 2012;Cao et al. 2016) which may increase mutual understanding among them.Establishing a shared vision in organizations is essential to employees' motivation and a sense of purpose (Tijunaitis et al. 2019) and may raise spirits at work (Berraies et al. 2020).Thus, employees who share the same vision have a common understanding, views, and interpretation of organizational goals that lead to better work performance.

Trust
The third dimension of social capital in this study is trust.Trust has become a key component in the workplace environment, both at an employee level (Clausen et al. 2019) and an organizational level (Ranjay and Sytch 2008).Since employees widely use SM in their work, relationships and trust are fostered via SM platforms.SM can assist employees in getting to know their colleagues, leading to informal relationships that may enable employees to develop trusting relationships (Hauser et al. 2016).Social media capabilities in facilitating informal, social exchange with organizational members across time and over geographic boundaries have permitted employees to reach mutual understanding and lead to the formation of trust (Valenzuela et al. 2009;Cao et al. 2016).Deepening mutual understanding can enhance trust among employees in the workplace, and it is much easier to promote trusting relationships with colleagues through various media, including SM.

Work Engagement & Innovative Performance
In the digital age, employees widely adopt SM usage for work-related purposes.However, interaction through SM use at work can either act as a resource or a hindrance in the context of work engagement.Scholars discovered that SM use at work has significantly enhanced job resources, specifically social capital (Yen et al. 2020;Huang and Liu 2017;Sheer and Rice 2017), which can increase work engagement.In fact, SM allows employees to maintain social network ties and to support colleagues (Charoensukmongkol 2014), as it can be accessed anywhere, anytime, and across organizational boundaries.Meanwhile, the direct role of SM usage in organizations is also associated with work engagement.A few studies have discovered the effects of SM usage on employee engagement.Empirically, SM use at work has significantly enhanced work engagement (Oksa et al. 2020;Zoonen et al. 2017;Men et al. 2020).
Employees who experience social capital at the workplace have extensively displayed work engagement that results in better task and contextual performance (Bhatti et al. 2018;Clausen et al. 2019).Due to the resources provided by good interpersonal relationships in organizations, employees will be engrossed in work-related tasks as they work harder to complete their work and perform better.In addition, highly engaged employees are positive and enthusiastic about their work (Bakker and Albrecht 2018), which leads to the enhancement of their creativity and innovation (Cheng et al. 2020).Further, some scholars also discovered the positive association between work engagement and innovative performance (Zyl et al. 2019;Sharma and Nambudiri 2020;Waheed et al. 2017).Thus, social capital is viewed as a job resource that leads to work engagement and consequently affects employees' innovative job performance.
Most studies have focused on the association between SM use at work and job performance.However, Zoonen et al. (2017) stated that studies neglect the essential dimension of employees' well-being, such as engagement.In addition, Sharma and Nambudiri (2020) suggest that considering social capital in the relationship between work engagement and its outcome could have a crucial intervening role in this association.Ali- Hassan et al. (2015) also suggested a possible extension to the association between SM use, social capital, and job performance that focuses on other potential variables.Thus, work engagement is incorporated in this study.

SM Use and Network Ties
Employees are using SM to develop and maintain social relationships with their colleagues.As mentioned by Charoensukmongkol (2014), SM allows employees to create and maintain social relationships, besides acting as a mechanism for giving support or advice to colleagues.Moreover, network ties within an organization are easily expanded through SM due to the ability of SM to easily connect to employees from different places or times.In addition, several studies discovered that SM has significantly contributed to the development of social capital in the workplace represented by network ties (Babu et al. 2020;Huang and Liu 2017).Therefore, SM use at work can generate social capital by establishing network ties among employees.The first hypothesis is presented as: Hypothesis H1.Social media use at work positively influence employee's network ties.

SM Use and Shared Vision
Aside from face-to-face interaction, SM has been widely used in organizations to foster a shared vision.SM can leverage a flexible medium for employees to participate in work tasks.It potentially involves them actively through collaboration embedded within informal social interactions, making it easier for employees to develop and maintain a shared vision within an organization (Cao et al. 2016).In addition, scholars have found that a shared vision in the workplace was stronger when employees use SM at work (Babu et al. 2020;Tijunaitis et al. 2019;Ali-Hassan et al. 2015;Cao et al. 2016).These findings showed that SM use at work can promote a shared vision by facilitating employee communication and interaction.Therefore, a second hypothesis is proposed: Hypothesis H2.Social media use at work positively influence employee's shared vision.

SM Use and Trust
Recently, trust within an organization has evolved through SM usage.SM has become a platform for introducing organizational members to informal relationships that encourage the development of trust in the workplace (Tijunaitis et al. 2019).Besides, SM can help employees actively share work-related information that indirectly promotes mutual trust.Thus, well-informed employees can communicate more effectively with their colleagues and trust each other's work (King and Lee 2016).Moreover, prior studies have demonstrated that SM use has contributed significantly to developing workplace trust among employees (Kelton and Pennington 2020;Louati and Hadoussa 2021;Hauser et al. 2016).Therefore, SM use at work can generate social capital in workplaces by establishing trust among employees without relying on face-to-face in-person contact.The third hypothesis is as follows: Hypothesis H3.Social media use at work positively influence employee's trust.

Network Ties and Work Engagement
SM allows employees to create and maintain social relationships, besides giving support or advice to colleagues (Charoensukmongkol 2014).Cao et al. (2016) stated that social media can help employers connect with employees who share similar backgrounds and interests, which is helpful for strengthening ties with colleagues, maintaining a professional network, and discovering potential relationships.Adopting social media at work will positively enhance work engagement due to the ability of SM to make it easier for employees to obtain resources from colleagues who are in their network ties.In structural social capital, individuals' interactions with social networks provide them with various resources, such as advice, information and social support (Jutengren et al. 2020).Moreover, Oksa et al. (2020) discovered that employees who received social support from their network ties via SM interaction positively increased work engagement among the Finnish working population.Thus, employees are more engaged with their work when network ties among colleagues are forged through SM.The fourth hypothesis is: Hypothesis H4.Network ties positively influence employee's work engagement.

Shared Vision and Work Engagement
In an organization, shared vision helps create a sense of common responsibility and collective action (Mostafa 2019), which leads to an improvement in mood in the workplace (Berraies et al. 2020).It facilitates in employees the sharing of common collective goals and aspirations of an organization, which can easily be achieved through SM use at work.In addition, it directly supports workplace knowledge sharing (Chang et al. 2012).As Ali- Hassan et al. (2015) mentioned, to develop a shared vision, employees need to be familiar with their colleagues, aware of the social environment and understand their problems and solutions.Hence, a shared vision can enhance work engagement if employees have a common understanding within an organization and can achieve it through collaboration.The following hypothesis is denoted as: Hypothesis H5.Shared vision positively influence employee's work engagement.

Trust and Work Engagement
Work engagement within an organization can evolve through social support from colleagues (Bakker and Albrecht 2018).Employees might develop mutual trust with each other when they feel their relationships with colleagues are mutually supportive, which would result in high work engagement (Jutengren et al. 2020).In addition, trust is established through an excellent interpersonal relationship that develops over time through various interactions (Chiu et al. 2006).Moreover, previous studies discovered that mutual trust in relational social capital at co-worker levels was positively associated with work engagement (Strömgren et al. 2016;Mostafa 2019).Undoubtedly, trust is a significant dimension of social capital in developing employees' work engagement.Therefore, the following hypothesis is presented as: Hypothesis H6.Trust positively influences employee's work engagement.

Work Engagement and Job Performance
Innovativeness is critical for the growth, long-term survival, and sustainability of organizations (Agarwal 2014).Employees with high innovative performance are part of the success factors for organization success.Work engagement has significantly brought creativity to organizations (Waheed et al. 2017).Some scholars have discovered a positive association between employee work engagement and innovative performance (Sharma and Nambudiri 2020;Cheng et al. 2020).Employees with work engagement have more creative ideas that enhance innovative performance because of their openness to new experiences (Gawke et al. 2017;Orth and Volmer 2017).In addition, engaged employees seek support and feedback as they manifest innovativeness through openness to new experiences and challenges (Sharma and Nambudiri 2020).It is posited that employees who display a remarkable level of work engagement show innovative job performance.The hypothesis is as follows: Hypothesis H7.Work engagement positively influence employee's innovative job performance.
The following Figure 1 presents the research model of this study.

Sample
A quantitative approach was used by conducting a survey based on a selfadministered questionnaire.The samples are Malaysian public and private employees.

Sample
A quantitative approach was used by conducting a survey based on a self-administered questionnaire.The samples are Malaysian public and private employees.The Government's success largely depends on the employees' ability, high cognitive skills, and work performance in demonstrating their knowledge-based service.Since public sector employees are the backbone of the country in providing outstanding public services, implementing measures at the individual level may enhance an organization's overall performance (Amran et al. 2021;Hassan et al. 2021).Despite its significance, research on public sector employees' job performance has received little attention (Johari et al. 2019), especially on innovative performance.Furthermore, the private sector, specifically government-linked companies (GLC), also plays a significant component in Malaysia's economy in which Government holds a certain number of shares.Money returns and profits have always been the key elements in the private sector in expanding Malaysia's economy (Permarupan et al. 2013).Thus, this study focuses on employees from the public and private sectors.
A quota sampling frame was applied to ensure sample representation within the population.The employees were divided into government, statutory body, and GLC based on the ratio of 50:30:20, respectively.This study used the G*Power application to draw the sample size from the target population.Hair et al. (2014) recommended that with a power of 0.8 and an effect size of f2 = 0.15, plus the predictor of the variable with the highest value of 3, the minimum sample was determined as 77.

Data Collection Procedure
The online survey was utilized for data collection.The population target was employees who use SM for work-related purposes.The questionnaires were distributed through email to 1500 participants.Participants filled in a Google form attached within the mail.To increase response rate, a follow-up email was sent to the participants.Two hundred ninetyfour responses were obtained within three months (November 2021 to February 2022), with a response rate of 19.6%.Data were cleaned and screened for duplicate, incomplete and inappropriate responses, resulting in 291 respondents.

Measurement
In this study, the measures of all constructs were adapted from previously validated scales and items.The 3 items for SM use at work was adapted from Cao et al. (2016).The first component of social capital, network ties, was measured with four items adapted from Cao et al. (2016).The 4 items of shared vision are adapted from Cao et al. (2016), which measure whether employees have a shared vision within an organization through virtual communities based on SM use at work.In addition, this study measured the trust variable comprising five items adapted from Cao et al. (2016) to assess the development of trust among employees through SM use at work.The measurement of the work engagement variable with 5 items was adapted from Saks (2006).Lastly, the innovative job performance scale was adapted from Ali- Hassan et al. (2015) consisting of 6 items to measure the exact process of SM use at work on employees' innovative performance.All the items were rated on a 7-point Likert scale.The list of measurement items for the constructs is presented in Appendix A. The list of variables included with the number of items assessed in the present study is depicted in Table 1.Before the questionnaires were sent to potential respondents, a pre-test and pilot test were carried out, which served as a preliminary study.The first step was pre-testing the questionnaire by experts for content validity.To determine the clarity of the questionnaire, the items were reviewed by four experts from Malaysian public universities with different fields of expertise (such as organizational behavior, human resource management, and technology adoption).Amendments to the questionnaires were then made, based on reviews and comments by these expert panels.The next stage was the pilot study, which was conducted after the items were amended based on experts' comments.Finally, the questionnaires were distributed to employees working in the government and private sectors.Approximately 43 respondents participated in this pilot study.
For the measurement model of the pilot study, this study performed a factor analysis and reliability test using SPSS version 22.The findings discovered that for all items of this study, the construct validity of all the variables was sufficient, with all the values exceeding the threshold.Generally, the findings showed that the factor loading was more significant than 0.5 (Hulland 1999;Truong and McColl 2011), the KMO value was higher more than 0.5 (Hair et al. 2010;Kaiser 1974), total variance explained more than 60 percent (Peterson 2000), and Cronbach alpha's value exceeded 0.7 (Taber 2018).With these results, each construct's item remained and proceeded for the final survey.The summary of the pilot test results, including factor loading, KMO value, variance explained, and Cronbach alpha's value, is attached in Appendix B.

Data Analysis
Before proceeding with the inferential analysis, the data analysis began with descriptive statistics, and this study used SPSS version 22 to measure the frequency of background characteristics.The study applied the Partial Least Squares Structural Equation Modeling (PLS-SEM) approach in analyzing the data using SmartPLS 3.0 software.The PLS-SEM method was helpful in estimating complex structural models with many constructs, indicators, and/or model relationships, as well as its ability to adequately use non-normal data (Hair et al. 2019;Ringle et al. 2015).In addition, it was possible to perform exploratory research in developing theory and estimate model that commonly displays a high degree of statistical power compared to the CB-SEM (Hair et al. 2011;Law and Fong 2020;Sarstedt et al. 2014).Advanced model elements, such as hierarchical component models, moderator variables, or nonlinear relationships, were handled flexibly using PLS-SEM (Chin 2010;Henseler et al. 2016;Hair et al. 2017).Thus, in exploring the association of SM and job performance in this study, PLS-SEM was deemed appropriate.

Demographic Information
Table 2 presents the demographic information of the 291 respondents.In terms of gender, 40.5 percent (118) were male and 59.5 percent (173) were female.For the level of education, most respondents (59.5 percent) were degree holders and Masters's degrees (23.0 percent).28.2 percent of respondents had worked for more than 11-15 years, and 20.3 percent for 6-10 years.For SM platforms, WhatsApp (87.3 percent) and Facebook (7.6 percent) were frequently used by respondents for work-related purposes.

Common Method Bias
A statistical remedy was employed in this study to manage common method bias, which is common in behavioural research (Podsakoff et al. 2003;Podsakoff et al. 2012).In PLS-SEM, a full collinearity test was used to assess common method bias, and a variance inflation value (VIF) below 5.0 (Kock 2015) and below 3.3 (Kock and Lynn 2012) indicated the dataset did not suffer common method bias.From the test, there was no significant issue in the dataset, as the VIF values of all constructs were lower than 3.30, as shown in Table 3.

Measurement Model
The measurement model is the first stage of using PLS-SEM that specifies the relations between the latent variable (construct) and its indicator (manifest variable).The purpose of measurement model analysis is to ensure all the required relationships between the latent variables and their indicators are met by the model assessment (Hair et al. 2017).For construct reliability and validity, the convergent validity was evaluated by assessing the factor loadings and average variance extracted (AVE), and Cronbach alpha and composite reliability for the internal consistency reliability (Fauzi 2021;Chin 2010).Table 4 shows that all the factor loadings exceed the minimum of required value 0.6 for an exploratory study (Hair et al. 2017).The Cronbach alpha and composite reliability for all constructs were higher than the required value of 0.7 (Hair et al. 2017).Table 4 presents the construct validity of the measurement model for the reliability and validity analysis.Discriminant validity is essential to ensure that each variable is distinct and not supposed to be related to each other (Chin 2010).This study applied the Heterotraitmonotrait (HTMT) ratio of correlation, which has a better performance for measuring the discriminant validity in variance-based SEM than the cross loadings and Fornell Larcker criterion (Henseler et al. 2015).Franke and Sarstedt (2019) stated that to establish the discriminant validity that reliably distinguishes between those pairs of latent variables, a cut-off value of HTMT for conceptually dissimilar constructs should be less than 0.85, while for conceptually similar constructs, it should be less than 0.9, depending on the study context.Table 5 shows that none of the respective constructs violates the cut-off HTMT value of 0.85, suggesting that the variables of this study possess satisfactory discriminant validity.

Structural Model
Before proceeding with the structural model assessment, the normality of data was measured by implementing Mardia's multivariate kurtosis.The online tool available at http://webpower.psychstat.org/models/kurtosis/(accessed on 15 September 2022) was used to calculate univariate/multivariate skewness and kurtosis, as suggested by Cain et al. (2017).The results indicate that data was not multivariate normal, as shown by the skewness (β = 5.479, p < 0.01) and kurtosis (β = 49.476,p < 0.01).As suggested by Hair et al. (2019), this calls for a nonparametric analysis tool to perform bootstrapping.
A 5000 bootstrapping re-sampling technique was performed to assess the structural model based on the path coefficient and statistical significance (Banjanovic and Osborne 2016).The SmartPLS 3 Version 3.6.8bootstrapping function was used to explore the path coefficient (β-value) of exogenous to endogenous variables, t-values, squared multiple correlation (R 2 ) values of explained variance on the endogenous variable, and to assess the predictive relevance of the model.Table 6 shows the results of R 2 , f 2 and Q 2 , and Table 7 displays the structural analysis results and decision on hypotheses.Figure 2 illustrates β-value, t-value and R 2 of hypotheses in the structural path.model based on the path coefficient and statistical significance (Banjanovic and Osborne 2016).The SmartPLS 3 Version 3.6.8bootstrapping function was used to explore the path coefficient (β-value) of exogenous to endogenous variables, t-values, squared multiple correlation (R 2 ) values of explained variance on the endogenous variable, and to assess the predictive relevance of the model.Table 6 shows the results of R 2 , f 2 and Q 2 , and Table 7 displays the structural analysis results and decision on hypotheses.Figure 2 illustrates β-value, t-value and R 2 of hypotheses in the structural path.
The result shows that five of the seven proposed hypotheses were supported.As hypothesized, H1, H2, and H3 were supported, as SM use at work has a positive influence on network ties (β = 0.557, t = 12.923), shared vision (β = 0.485, t = 9.696) and trust (β = 0.412, t = 7.494).The R 2 of the three variables are 0.310, 0.235 and 0.170, denoting that SM use at work explained 31%, 23.5% and 17% of the variance, respectively.Next, H4 and H6 were accepted, which posited that network ties (β = 0.203, t = 2.511) and trust (β = 0.205, t = 1.843) showed a significant positive effect on employees' work engagement, but the effect of shared vision on employees' work engagement was not significant (β = −0.023,t = 0.234), thus rejecting H5.The R 2 was 0.128, indicating that the three predictors explained 12.8% of the variance in work engagement.Lastly, H7 (β = 0.467, t = 8.546) demonstrated that work engagement was found to be significant for innovative job performance among employees, with an R 2 of 0.218 indicating 21.8% of the variance in innovative job performance.
A guideline by Cohen (1988) has provided benchmarks in reporting effect size (f 2 ), which are small = 0.02, medium = 0.15 and large = 0.35.Table 6 presents that the effect size varies from 0.000 to 0.450.This study discovered that the path of H1 has a large effect of social media use at work on network ties; meanwhile, the effect size of the path of H2, H3 and H7 are medium.However, the study discovered the insignificant path of H5, which indicated no effect of a shared vision on work engagement.The result shows that five of the seven proposed hypotheses were supported.As hypothesized, H1, H2, and H3 were supported, as SM use at work has a positive influence on network ties (β = 0.557, t = 12.923), shared vision (β = 0.485, t = 9.696) and trust (β = 0.412, t = 7.494).The R 2 of the three variables are 0.310, 0.235 and 0.170, denoting that SM use at work explained 31%, 23.5% and 17% of the variance, respectively.Next, H4 and H6 were accepted, which posited that network ties (β = 0.203, t = 2.511) and trust (β = 0.205, t = 1.843) showed a significant positive effect on employees' work engagement, but the effect of shared vision on employees' work engagement was not significant (β = −0.023,t = 0.234), thus rejecting H5.The R 2 was 0.128, indicating that the three predictors explained 12.8% of the variance in work engagement.Lastly, H7 (β = 0.467, t = 8.546) demonstrated that work engagement was found to be significant for innovative job performance among employees, with an R 2 of 0.218 indicating 21.8% of the variance in innovative job performance.
A guideline by Cohen (1988) has provided benchmarks in reporting effect size (f 2 ), which are small = 0.02, medium = 0.15 and large = 0.35.Table 6 presents that the effect size varies from 0.000 to 0.450.This study discovered that the path of H1 has a large effect of social media use at work on network ties; meanwhile, the effect size of the path of H2, H3 and H7 are medium.However, the study discovered the insignificant path of H5, which indicated no effect of a shared vision on work engagement.

Discussion
The study utilized the social capital theory to explore how SM use at work predicts social capital (network ties, shared vision & trust), which affects work engagement and, subsequently, relationships on innovative job performance.Firstly, the study discovered the new outcomes of SM use and social capital in the workplace and found SM use at work predicts network ties, shared vision, and trust among Malaysian employees.Based on the findings, H1 (t = 12.923), H2 (t = 9.696), and H3 (t = 7.494) showed a t-value of more than 1.65, and these hypotheses were supported.These findings are congruent with previous studies (Babu et al. 2020;Huang and Liu 2017;Cao et al. 2016;Tijunaitis et al. 2019;Louati and Hadoussa 2021;Ali-Hassan et al. 2015), SM has become a platform to create network ties within an organization.The integration of SM into work-life allows employees to develop and maintain social relationships, besides becoming a supportive tool in giving support or advice to colleagues.Furthermore, the deployment of SM in the workplace helps employees exchange work-related knowledge that enhances their shared vision in executing work tasks.In addition, SM use at work can assist in spreading social information with other members of organizations, which can help develop and reinforce network ties and trust among them.
Second, the findings of this study revealed that network ties and trust promoted work engagement among Malaysian employees.The H4 (t = 2.511) and H6 (t = 1.843) were accepted as a t-value of more than 1.65, positing network ties and trust encouraged employees' work engagement.These findings align with previous studies and emphasize that employees foster high work engagement and more satisfaction with their work when network ties and trust among colleagues are connected through SM (Oksa et al. 2020;Hauser et al. 2016).In structural social capital, individuals' interactions with social networks provide them with various resources, such as advice, information, social support and other supports (Jutengren et al. 2020).SM's ability to provide employees with a range of benefits and resources from colleagues might enhance network ties and mutual trust with each other when they feel their relationships with colleagues are mutually supportive, resulting in high work engagement.
Meanwhile, this study found that a shared vision on employees' work engagement was not significant as t = 0.234, t-value < 1.65, thus rejecting H5.The results indicate that a shared vision did not promote the employees' work engagement.This finding is inconsistent with past studies (Babu et al. 2020;Tijunaitis et al. 2019;Ali-Hassan et al. 2015;Cao et al. 2016).Basically, the main components in work engagement consist of behavioral, emotional and cognitive elements, which are strongly directed toward individual work performance (Saks 2006;Zhang et al. 2019).Employees with high work engagement tend to immerse themselves in their work tasks cognitively, emotionally, and physically.In addition, a shared vision facilitates employees to share common collective goals and aspirations of an organization that can quickly be achieved through collaboration (Berraies et al. 2020).In this study, employees might feel that a shared vision could not promote work engagement due to the specific job scope already assigned to them.Also, the employees understand and share a common goal and aspiration; however, they might not feel enthusiastic and inattentive with their work activities due to years of work experience.Based on the demographic information, most of the employees had long-term work experience, between 11-20 years in their field, thus influencing their motivation and level of energy to engage with their work.Furthermore, long-term work experience may lead to boredom and a repetitive routine in completing work tasks subsequently, decreasing work engagement (van Hooff and van Hooft 2017).
Lastly, this study ascertained that work engagement was significantly associated with employees' innovative job performance, indicating H7 (t = 8.546) t-value > 1.65.As expected, this result is in line with previous studies (Gawke et al. 2017;Orth and Volmer 2017).This result suggests that work engagement is a significant force that leads employees to perform with a strong focus on work.They can show their creative and critical thinking to produce, adopt, promote, and implement novel ideas.Engaged employees are associated with better task performance, high levels of creativity, client satisfaction, and organizational citizenship behavior (Bakker et al. 2014).

Implication
By employing the social capital theory, this study contributes to the theoretical understanding of the role of SM use among employees in government departments.The present study provides an essential extension of current research by offering a detailed theoretical understanding of the social relationship as resources underlying employees' work behavior, specifically in innovative job performance.In addition, this study replicates and extends SCT and research on the area of SM within the organizational context, which is essential to understand in today's technology-driven workplaces.The outcome provides a meaningful theoretical contribution to the literature on social relations within the organization and the effects of SM use on employee outcomes at the workplace.
Employees who use SM at work may be unaware of the consequences of SM usage.They view SM as an integral part of their daily life.The importance of understanding communication needs is essential for ensuring that employees literally engage with the information.Considering the fact that SM use has become an integral part of employees' work life, they need to have greater insight into the role of SM usage in executing the work tasks at the workplace.In addition, this study is ground-breaking in causing employees to consider SM's usefulness as an innovative tool that can enhance their work engagement through information delivery, knowledge sharing, and facilitating relationships.
Last, the findings of this study offer management a guide to adopting an emerging and popular technology, SM, as a medium for facilitating employees' communication, work engagement, and job performance.Such measures can effectively provide practical insights for management to create new strategies for establishing and maintaining social interaction ties that are likely to result in even more successful job performance and better decision-making.In addition, management can strengthen existing guidelines or policies regarding the terms of SM use, which may further ensure the productive and consistent use of communication channels to generate better routines and innovative performance.

Limitation and Future Work
Although this study offers valuable insights, certain limitations should also be acknowledged and addressed in future studies.Given the pervasiveness of SM use at work and its effects on employees' innovative job performance, our findings imply that future research should pay closer attention to social capital, its antecedents, and its outcomes in the other type of organizational settings (i.e., private, large corporate organization) or specific industry such as manufacturing, services, and education.Different contexts would provide a detailed and diverse understanding of adapting to SM usage that may influence employees' job performance.
Second, this study highlighted only the positive consequences of SM use in the workplace.Future studies should explore the harmful and benefits of SM use at work by integrating two theories/models (e.g., the SSO model and social capital theory).We believe that researchers in other disciplines can improve the understanding of SM use at work and employees' outcomes from another perspective through a combination of theories.
Third, this study collected data from a single source by obtaining the responses from an internet survey only.Although the statistical result showed no issues in common method bias, respondents might be unable to inform the actual situation or condition in answering the sensitive questions related to personal relation, trust, and vision.Future scholars should consider applying a mixed-method design by adding interview sessions or observation to measure employees' social ties and innovative job performance accurately.

Conclusions
As employees have increased connectivity and diverse social groups on SM platforms across personal and professional settings, SM usage at work has facilitated employees to maintain social relations with co-workers.Several scholars have accepted that social capital through SM platforms has significantly played a critical role in maximizing work engagement and employee performance.The present study adopts the SCT theory to explain how SM use at work predicts social capital that influences work engagement and, subsequently, employees' innovative job performance in the workplace.The findings of this study provide an essential extension of prior knowledge for the conceptual relationships for social capital that were empirically validated in terms of work engagement and employees' outcomes.Moreover, the outcome of this study will be of immense benefit to employees and employers in adapting SM use at work to promote employees' innovative job performance.
BCI LL = Bias confidence interval lower limit, BCI UL = Bias confidence interval upper limit.

Table 1 .
The list of variables measured in this study.

Table 4 .
Reliability and convergent validity analysis.

Table 6 .
The result of the coefficient determination (R Square), Effect Size (f Square), and predictive relevance (Q Square).

Table 6 .
The result of the coefficient determination (R Square), Effect Size (f Square), and predictive relevance (Q Square).

Table A1 .
Cont. members in the virtual community created by social media would go out of their ways to make sure I was not damaged or harmed.TR3I felt like members in the virtual community created by social media cared what happened to me.TR4I believed that members in the virtual community created by social media approached their jobs with professionalism and dedication.TR5Given members in the virtual community created by social media track record, I saw no reason to doubt their competence and preparation.sointomyjobthatIlosetrack of time.WE3This job is all consuming, I am totally into it.WE4Mymindoftenwanders and I think of other things when doing my job (R).My mind often focuses and concentrates on my job.WE5I am highly engaged in this job.