SOCIAL MEDIA USAGE ALLURE JOB PERFORMANCE: MEDIATING ROLE OF SOCIAL CAPITAL AND KNOWLEDGE SHARING

Social Capital Theory

The social capital theory is a useful framework for understanding how people interact with one another in social networks to gain psychological and material benefits (Yen et al., 2020). Previous research has shown that social capital is an important component of comprehending the applications and implications of modern media (Ali-Hassan et al., 2015; Bharati et al., 2015; Yen et al., 2020). The current study focuses on how employees interact with coworkers in organizations using social media. As a result, we define social capital as a worker's social relationships that make available the resources embedded in those relationships (Ali-Hassan et al., 2015; Bharati et al., 2015).

There are two reasons why this framework was chosen. First and foremost, this framework is the most widely used and accepted social capital framework in social media research (Ali-Hassan et al., 2015; Bharati et al., 2015; Chua et al., 2012; Kamboj et al., 2017; Sun et al., 2014). Second, the multidimensional conceptualization of this social capital framework is well suited to the complexities of social media use in organizations. According to Koroleva et al. (2011), the structural aspect addresses resource availability, whereas the cognitive and relational aspects address an individual's ability to access these resources. Both dimensions, we believe, are critical for assessing the links between social media use and job performance.

Social exchange theory

Knowledge sharing is the process by which an organization gains access to its own and other organizations’ knowledge. (Nelson & Rosenberg 1993). Successful knowledge sharing results in firms mastering and implementing novel product designs, manufacturing processes, and organizational structures (Nelson K., 1993). Social exchange theory is concerned with the process of knowledge exchange between individuals, groups (Kankanhalli et al., 2005), and online communities (Wasko and Faraj, 2005). Individuals’ behaviors that maximize their benefits from social interaction are defined by social exchange theory (David Gefen, 2002). The ability of an individual to form and maintain social connections with others is required for social exchange (He et al., 2009). Furthermore, when attempting to establish a professional identity and reputation in relevant communities, individuals prefer to share their knowledge (Hsu & Lin, 2007).

Social Media Usage and the Structural Dimension of Social Capital

In a social network, the structural dimension is primarily concerned with the “patterns of relationships between different participants” (Nahapiet & Ghoshal, 1998). It is also understood as the “general tendency for interaction and integration between individuals” and is defined by the quantity and quality of pre-existing interpersonal relationships as well as the structure of the network (Ali-Hassan et al., 2015). The strength, density, and organization of network ties are discussed along with the ways in which these network ties affect social mobility, competitive advantage, and information flow within social networks. This study examines how the strength, density, and organization of network ties influence social dynamics within social networks (Burt, 1992). The extent of relationships, amount of time spent together, and frequency of communication are all reflected in social interaction links (also known as network relationships) (Chiu et al., 2006). As a result, the use of social media in the workplace encourages employees to share both social and professional information with other employees, which can strengthen and build employee social networks (Ellison et al., 2014; Cao et al., 2016). We formulated a hypothesis based on the information discussed in prior studies regarding social media usage and the structural dimension of social capital.

Social Media Usage and the Relational Dimension of Social Capital

The trust and reciprocity of interpersonal relationships that people develop during network interactions are key components of the relational dimension of social capital (Nahapiet & Ghoshal, 1998). Much of the research attention has focused on the core element of belief in the relational dimension of social capital (Steinmo & Rasmussen, 2018). Social media has a significant positive effect on trust-building in the workplace through improved social bonding, communication, collaboration, access to information and expertise, employee engagement, involvement, etc. (Cao et al., 2015; Cao et al., 2016; Kelton & Pennington, 2019; Son et al., 2016). According to Hau and Kim (2011), social interactions within social networks lead to a favorable attitude and a sense of belonging to others, resulting in social trust. According to Cao et al. (2012), social media use in the workplace enhances social trust by maintaining and promoting workplace relationships and communication. Additionally, the fundamental reciprocity that governs interactions between peers is described as relational social capital (Kamboj et al., 2017). Social exchange theory serves as the foundation for the idea of reciprocity (Louati and Hadoussa, 2021). According to social exchange theory, people who participate in network communities seek reciprocity so that they can invest time and effort in teaching others about their expertise. Online reciprocity norms may arise because people are constantly interacting with other Internet users, leading to the development of expectations and norms (Pai & Tsai, 2016). Unlike face-to-face interactions, where direct reciprocity is expected, many studies argue that reciprocity can be generalized in online networks (Wasko et al., 2009). Therefore, the above studies clearly mention how social media usage leads toward social trust to ultimately affect the relational dimension of social capital. Hence, the following hypothesis has been framed.

Social Media Usage and the Cognitive Dimension of Social Capital

The shared vision of network members, which fosters a sense of mutually shared ideals and standards of action in a social environment, is linked to the cognitive dimension (Aslam et al., 2013; Tsai & Ghoshal, 1998). The shared goals and objectives of the members within the organization that can be achieved through collaboration are reflected in the shared vision (Wagner, 1995). The collective aims and aspirations of the members of an organization are embodied in its shared vision (Tsai & Ghoshal, 1998). Social media use facilitates the development and maintenance of a common goal (Ali-Hassan et al., 2015; Cao et al., 2016; Yen et al., 2020). Social media platforms provide open channels for communication within organizations, fostering informal networks of connections that are essential for employee collaboration across skill sets (Awolusi, 2012). Social media enables users to actively participate in casual social interactions through integrated collaboration, which helps team members develop a common vision (Cao et al., 2016). Therefore, it is clear that social media usage would encourage an organization’s members to work in a team to achieve common organizational goals. Here, the following hypothesis has been developed.

Social Capital and Knowledge Sharing

Knowledge sharing, which is a part of knowledge management, encompasses the sharing of knowledge both within and between organizations, and it describes “how individuals communicate with other co-workers about their professional experience, knowledge, and skills within the context of the organization”(Lin, 2007). When people communicate tacit and explicit knowledge and create new information, they are transferring or disseminating knowledge among themselves, groups, or organizations (Nonaka & Takeuchi, 1995). In this process, individuals interact and exchange knowledge with each other, resulting in the formation of new knowledge. In other words, it includes both “donating” and “collecting” knowledge. Knowledge “collection” is an effort to persuade other organizational members to contribute their knowledge, whereas knowledge “donation” is “communication between individuals that depends on their willful transfer of intellectual capital to an individual (Louati & Hadoussa, 2021).” It has been shown by Lin and Haung (2023) that the exchange of information is positively mediated by a favorable emotional tone in interpersonal relationships, shared values, and trust.

Three dimensions of social capital can influence knowledge donation and collection. In fact, social capital functions by providing people with access to relevant information as well as fostering a sense of community, shared values, and trust between individuals (Van Den Hooff & Huysman, 2009). Several studies (Koranteng & Wiafe, 2019; Akhavan and Mahdi Hosseini, 2016; Chung et al., 2016; Hu & Randel, 2014; Louati and Hadoussa, 2021; Sargis Roussel & Deltour, 2012; Chow & Chan, 2008) have suggested that social capital, as represented by its structural, relational, and cognitive dimensions, is important for promoting knowledge sharing in various organizational and social contexts. Building and using social capital can encourage a collaborative, trusting, and mutually understood culture, which in turn fosters knowledge sharing and creativity.

Additionally, social interactions facilitate the sharing of information between people and influence how everyone can access knowledge (Chiu et al., 2006; Chow & Chan, 2008; Cao et al., 2016). Since network participants are more willing to share knowledge when they are familiar, social interactions can be viewed as networks between individuals that act as channels for knowledge transmission (Koranteng et al., 2018). Additionally, social interactions make it possible to affordably access a variety of knowledge sources and can promote the blending and sharing of knowledge (Razak et al., 2016). We have proposed the following hypotheses based on the above mentioned studies.

Knowledge sharing also results in a belief in the benefits of sharing, including rewards, interpersonal relationships, and mutual benefits (Chen & Hsieh, 2015). Relationships based on mutual trust have a big impact on how employees feel about sharing their knowledge (Bock et al., 2005). Members of online communities will view the exchange of knowledge favorably if they believe that by doing so, their peers will benefit (Wasko & Faraj, 2005). Hence, team members will contribute more innovative and beneficial ideas if they anticipate altruism (Moghavvemi et al., 2018). In light of this, mutual trust is an important factor in knowledge sharing (Louati and Hadoussa, 2021). As a result, the following hypotheses are made on the basis of the study discussed above.

The cognitive dimension of social capital is the next social capital dimension to have an impact on knowledge sharing. A shared vision is an expression of cognitive social capital that is made up of people in a social network who share common goals and aspirations (Koranteg, 2019). According to many researchers, the presence of a common purpose among social network users is said to encourage resource sharing (Chow & Chan, 2008; Omotayo & Babalola, 2016; Tsai & Ghoshal, 1998). When network participants have the same objectives and concerns, miscommunication can be avoided, which expands the possibilities for resource sharing (Koranteng et al., 2018). The exchange of ideas and perceptions is facilitated by common interests. Thus, it is possible to see common goals as the factor that brings individuals together and motivates them to share their knowledge (Chow & Chan, 2008). Understanding the meaning of knowledge increases the quantity and quality of knowledge that team members share because of a shared goal and vision (Chiu et al., 2006). Conversely, the lack of a common goal among team members leads to conflict that may inhibit the transfer of knowledge (Lefebvre et al., 2016). Based on the given review of the literature, the following hypotheses are suggested.

Knowledge Sharing and Job Performance

Individual job performance is the degree to which an individual is capable of performing tasks or the level of performance that results in the fulfillment of organizational goals (Hadoussa, 2020). Aslam et al., (2013), Cao et al., (2015), and Cao et al., (2016) have shown that knowledge sharing has a significant impact on employees’ job performance in various ways, such as social media-based knowledge transfers improving employees’ problem-solving abilities, which ultimately enhance their work performance. Knowledge sharing on a regular basis enhances employee collaboration and understanding (Lee, 2018) and motivates them to support each other and reduce errors in work, thereby improving the overall job performance of the employees (Nguyen & Prentice, 2022; Kang et al., 2008). Additionally, knowledge sharing fosters the potential for team productivity and effectiveness on the job while preserving intellectual capital (Aksoy et al., 2016) and can enhance creativity by expanding on already existing knowledge, abilities, and skills (Lee & Hong, 2014; Hadoussa, 2020; Hu & Zhao, 2016). Knowledge sharing activities that took place in the context of social media increased job performance (Kwahk and Park, 2016). The findings of this study indicate that technology-based knowledge sharing has direct and indirect effects on employees’ job performance as well as their mental health. Deng et al. (2023) showed that the improved coordination and communication brought about by digital technologies has a major impact on knowledge sharing. According to Nguyen et al., (2023), both donating and collecting knowledge have significant and beneficial effects on the job performance of the employees. Based on the above arguments, the following hypothesis is framed.

Measures

We based our survey questions on scales that have already been used in the literature to increase validity. Social media usage has been employed as the independent variable, and it is taken from Ven Den Hooff et al. (2016) and Chiu et al., (2006). Five items have been taken from Ya-Ling Wu et al., (2016) and Leana and Pil’s (2006) Structural Social Capital Scale. Four items of relational social capital have been selected and adapted for this study from the scale of Ali-Hassan et al., (2015) as it measures the perception of trust among all teaching faculties. On the basis of developed items, the cognitive dimension of social capital has been measured by Leana and Pil (2006) and Wing et al., (2008). The measurement of the two parts of knowledge transfer, namely, knowledge donation and collection, has been adapted from a study by Ven Den Hooff et al. (2004). Job performance is measured by six items adopted from Yen et al., (2020). This study has been adapted from earlier studies to measure various constructs on a 7-point Likert-type scale, ranging from “strongly disagree” to “strongly agree”. The questionnaire is included in Appendix 1 along with a list of all constructs, their respective sources, and the number of items employed for each construct.

Data Collection

We employed a paper-based survey to collect data from the teaching faculties of various departments in Indian public universities. A survey tool was designed and distributed to 660 teaching faculties in public universities in India. Respondents to this study are those who use social media in the workplace for work-related activities. A final sample of 608 usable responses (92.1% response rate) from a range of male (45.5%) and female (54.5%) employees was retained after invalid responses were removed. Table 1 shows the demographic profiles of the respondents. The average respondent is between 31–40. In terms of marital status, 57.8% of the respondents are unmarried. 67.4% of the respondents were assistant professors, and 70.7% of the respondents are employed on regular basis.

TABLE 1 Demographic Profile of Respondents

View Fullscreen

Data Analysis and Results

The Statistical Package for the Social Sciences (SPSS-20) and the Analysis of Moment Structures (AMOS-20) were used to examine the data for this study. A structural equation modeling approach has been used for the data analysis in this study. This approach has been found to be effective for the study of many correlations between different variables (Hoyle, 1995). Composite-based models should opt for partial least squares (PLS), whereas factor-based models should use covariance based or PLSc. Following the conceptual model (based on theory), the constructs can be connected (Dash and Paul, 2021). The first approach is based on covariance, and the second is based on variance (partial least squares). We used Amos SEM in order to analyze the data for this study.

Measurement Model

The reliability of the scales was assessed using a confirmatory factor analysis (CFA). The CFA results show that there is a satisfactory relationship between the measurement model and the dataset: CMIN = 3.296, RMSEA = 0.062, CFI = 0.937, IFI = 0.937, NFI = 0.913, TLI = 0.930, and RFI = 0.903. All metrics are within acceptable limits, indicating a reasonable overall fit to the model (Fornell, 1987; Wold, 1989). To evaluate a measurement model in terms of item reliability, convergent validity (individual item reliability), and discriminant validity, it is first necessary to assess the reliability and validity of the research tool (Chin, 1998). Item loadings were used to assess convergent validity. According to Hair et al. (1998), a factor loading of 0.45 and above is considered appropriate when evaluating rotational factor patterns. Survey tools have been suitable for measuring each variable separately, as evidenced by the fact that, as shown in Tables 2 and 3, all items in the improved model exceeded the threshold. The number of factors was determined using an eigenvalue criterion of greater than 1 (Hair et al., 1998). Each factor is listed in Table 2 for a simple explanation.

TABLE 2 Reliability and Validity of Constructs

View Fullscreen

TABLE 3 Discriminant Validity

View Fullscreen

TABLE 4 Path Analysis and Hypothesis Testing

View Fullscreen

The internal consistency reliability of the items representing each factor were evaluated using Cronbach’s alpha. The findings show that after purification, the Cronbach’s coefficient alpha of each scale exceeds the required minimum of 0.70, which is the threshold for the acceptable range (Nunnally, 1979). According to the findings, Cronbach’s alpha is above the minimum threshold (0.852 to 0.954), indicating that the internal stability is sufficient. The findings on the validity and reliability of the items are presented in Table 2.

To determine the degree to which a particular construct differs from other constructs, the average variance extracted (AVE) has been evaluated. For a construct to be valid, the variance obtained by the indicators with respect to the measurement error must be greater than 0.5, as measured by the AVE (Chin, 1998). For all values greater than 0.5, the construct successfully captured at least 50% of the measurement variance, as shown in Table 3. The values along the diagonal of the table represent the square root of AVE and are greater than the off-diagonal correlations between variables, providing additional support for discriminant validity. The structural models and hypotheses can be confidently evaluated because the instrument has demonstrated reasonable levels of validity and reliability.

Discriminant Validity

The construct validity of the scales has also been assessed via an exploratory factor analysis (Moore and Benbasat, 1991). A principal components analysis with the varimax rotation resulted in the seven factors for the seven constructions presented in Table 2. A typical unidimensional structure of an instrument is indicated by the fact that most items are strongly loaded on their respective factors (Hair et al., 1998). To further evaluate the discriminant validity of all construct measures, a confirmatory factor analysis (CFA) was performed in the second phase of measurement validation (Gefen and Straub, 2005; Lewis et al., 2005). The results presented in Table 2 indicate that all items are significantly loaded on their respective constructs and that no items are cross-loaded, thereby supporting the validity and reliability of the measurement model. According to Hair et al., (1998), all item loadings are greater than 0.50 (Table 2), and Table 3 indicates that there are sufficient levels of discriminant validity among the constructs. The bootstrap method was used to perform the average variance extracted (AVE) analysis in the SEM analysis. The square root of the AVE of each construct should, as a general rule, be greater than the correlation of that construct with other constructs and should surpass 0.50 (Chin, 1998b; Gefen and Straub, 2005). These prerequisites have been met (Table 3), which supports the conclusion that all constructs have sufficient discriminant validity, as there is weak correlation among the constructs, meaning that they are different from each other. These tests conclude that the scale is reliable, valid, and suitable for hypothesis testing.

Structural Equation Model Evaluation

The structural model depicted in Figure 2 was used to establish the relationship between the independent and dependent variables and test the hypotheses after the measurement model was fit. The goodness-of-fit is almost sufficient for the model; the goodness-of-fit statistics and values according to the structural equation model’s test results are CMIN = 3.151, NFI = 0.916, CFI = 0.941, IFI = 0.941, TLI = 0.934, RFI = 0.907, and RMSEA = 0.060. This shows that our structural model can be accepted to test the hypotheses and gives good results. Initially, the mediation hypothesis was tested using the structural equation model. The test findings in Table 3 show the positive effects of social media usage on the structural, relational, and cognitive dimensions of social capital. In fact, social media usage appears to have a positive effect on structural social capital (SSC) (β = 0.203, t-value = 4.441, p < 0.001), relational social capital (RSC) (β = 0.156, t-value = 3.469, p < 0.001), and cognitive social capital (CSC) (β = 0.255, t-value = 5.407, p < 0.001). Hence the H1, H2, and H3 hypotheses are supported. The results of this study are concordant with earlier results (Cao et al., 2016; Ali Hassan et al. 2015; N. B. Ellison et al., 2014; Louati and Hadoussa, 2021; Ghorbanzadeh et al., 2021). Furthermore, the results demonstrated that knowledge donation by employees is strongly correlated with all dimensions of social capital (β = 0.081, t-value = 1.996; β = 0.083, t-value = 2.039; β = 0.112, t-value = 2.740). Additionally, the findings demonstrate that all dimensions of social capital, including structural, relational, and cognitive, have a positive impact on knowledge collecting (β = 0.178, t-value = 3.991; β = 0.187, t-value = 4.148; β = 0.097, t-value = 2.230). Thus, the H4a, H4b, H5a, H5b, H6a, and H6b hypotheses are all supported (Berraies et al., 2020; Cao et al., 2016; Chiu et al., 2006; Louati and Hadoussa, 2021). We can also see that knowledge donating and knowledge collecting improve job performance (β = 0.105, t-value = 2.698; β = 0.120, t-value = 2.799). Finally, the test results show that knowledge donating and collecting have positive and significant impacts on job performance, which supports hypotheses H7a and H7b (Chiu et al., 2006; Cao et al., 2016; Berraies et al., 2020).

View Figure

• Download as Powerpoint
• Download Figure

Mediating Effects

The mediating effects (Table 5) of social capital and knowledge sharing have been investigated in the model of the present study. According to the results of earlier studies, bootstrapping is a more effective method than the Sobel test for examining mediation effects (Hayes, 2009).

TABLE 5 Mediation Analysis

View Fullscreen

APPENDIX Survey Statements

View Fullscreen

Thus, to investigate the mediation effects in our proposed models (Figure 1), we used a bootstrapping bias-corrected confidence interval approach. We employed 5,000 samples to calculate the confidence interval, and the bias-corrected percentile approach produced a 95% confidence interval. According to the findings in Table 5, there is a significant indirect effect of social media usage on job performance (Estimate = 0.017, 95% CI = 0.007 to 0.030, p = 0.013). Social capital and knowledge sharing appear to mediate between social media usage and job performance because the confidence interval does not include zero. Hence, all the hypotheses of this study are supported. There has been a partial mediation between social media usage and job performance, with significant associations between social media use and job performance controlled by social capital and knowledge sharing (mediators).