Paradoxical effects of social media use on workplace interpersonal conflicts

Abstract The prevalence of social media use urged researchers to identify its antecedents and consequences. However, research related to social media use in occupational settings has been substantially ignored. Therefore, this study aims to examine the effect of job stress on social media use (social media addiction and work-related social media activities) and its effect on interpersonal conflict at the workplace. The study was conducted using a cross-sectional design. An online questionnaire survey was used to collect data from 253 office employees working in the oil and gas industry in Malaysia. Structural equation modelling was employed for statistical analysis using smart PLS. Results demonstrate a significant positive effect of job stress on social media use (social media addiction and work-related activities on social media) and its significant positive effect on interpersonal conflict in the workplace. Social media behaviour mediates the relationship between job stress and workplace interpersonal conflicts. The study findings show that employees use social media to cope with job stress, which in turn increases interpersonal conflicts at the workplace. Further, the inclination towards social media use indicates the absence of active social support at the workplace. The study findings support the efforts of organisational stakeholders to manage workplace stress by reducing employee’s dependency on online social media and by encouraging “offline active social support” in the offices.


Introduction
Job stress is a key concern in occupational health and safety because it endangers employees' health and well-being. Job stress poses a serious threat to employees' health (O'Connor et al., 2021). According to a report by the National Institute of Occupational Health and Safety NIOSH (2017), occupational stress is on the rise, particularly among office workers. Job stress creates mental problems in addition to physiological and musculoskeletal disorders (Cooper & Baglioni, 2018). Depression, sleeplessness, headaches, and cardiovascular illnesses are all symptoms of chronic stress (Kivimäki & Kawachi, 2015). Eventually, it leads to suicide (Milner et al., 2018). Besides health issues, productivity loss, high employee turnover, and increased absenteeism and presenteeism are attributes of job stress (Ramlawati et al., 2021;Schmidt et al., 2019;Yazdanirad et al., 2021).
In this regard, a literature review revealed three research gaps related to stress coping strategies using social media and their outcomes. First, previous studies largely focused on social, behavioural, technological, and mental health factors as antecedents of social media addiction (Aksoy, 2018;Al-Samarraie et al., 2021;Aygul & Akbay, 2019;Nguyen et al., 2020). All these investigations were conducted on the population of students and adults. The factor of job stress and the population of working individuals were ignored in previous research, despite the fact that more than 60% of the population consisted of employees and they spent more than 60% of their time at work (Thorp et al., 2012). Second, the factors of performance, health, and technological conflict were investigated as outcomes of social media addiction. However, the phenomena of social media addiction attenuate or exacerbate workplace interpersonal relationships remain unaddressed. Specifically, the role of interpersonal conflict in the workplace was largely ignored. According to Cargill (2019) social media addiction causes relationship conflicts due to disregarding surrounding social contacts. This study filled the research gap by investigating the effect of social media addiction on workplace interpersonal conflicts.
Third, previous research focused on social support and emotion regulation by diverting attention through online connections and watching social media stuff as a stress copying style using social media (Abdul Aziz et al., 2015;Coates et al., 2019;Kanwal et al., 2021). Another way to cope with stress is by sharing about problems, people, and places through social media posts or activities (Valkenburg & Peter, 2007;van Ingen & Wright, 2016;Wolfers & Schneider, 2021). This coping strategy and its effect on increasing or decreasing interpersonal conflicts at work have not been addressed yet. Thus, this study raises the question, does employees' social media behaviour resolve their workplace problems or lead to intensification? Therefore, this study aims to examine the effect of employees' social media addiction and work-related activities in response to job stress on workplace interpersonal conflicts.
The research model of the study (Figure 1) is theorised based on the conservation of resources theory (COR) (Hobfoll, 1989), which postulates that employees enthusiastically seek to preserve existing resources and acquire new resources. However, their act of resource conservation can gain or deplete resources (Hobfoll, 2001), such as the outcome of social media use. In this way, employees use social media to cope with and avoid job stress through their social media behaviour, which includes building online connections, following, liking, and actively responding to superiors and bosses on social media forums. This study contributes to existing research and has practical implications in three ways. First, the study has contributed an empirical relationship between stress and social media addiction and interpersonal conflicts in occupational settings to the existing research. Related to this, the study findings can be helpful for organisational management to improve performance, the health, and wellbeing of the employees and to manage interpersonal conflicts in the organisation. Second, this study has added an empirical relationship between job stress and social media activities (related to work) of the employees and interpersonal conflicts. In this regard, the study findings help organisational stakeholders manage job stress, employee social media behaviour, and the resulting interpersonal conflicts at the workplace. Third, the study contributes to the theory of conservation of resources by identifying employees workrelated social media activities as a resource factor.
The rest of the study is structured as the second section that presents the theoretical framework and research hypotheses of the study. The third section explains the research method, data, and variables. The statistical results of the study are explained in the fourth section, followed by a discussion and conclusion in the fifth section. The sixth section presents the practical implications of the study. Limitations of the study and future research directions are given at the end.

Theoretical framework and research hypothesis
Stress is a dysfunctional perception of physical and psychological reactions and stimuli experienced in the process of performing a job. As a result of this dysfunction, employees' emotions are threatened, and psychological dissatisfaction increases (Shin et al., 2021). Employees adopt different strategies to cope with this situation of stress. Stress coping strategies can be positive (meditation, yoga, mindfulness, etc.) and negative (smoking, aberrant working behaviour, drug addiction, etc.) as well. Employees try to divert their attention from stress using different tactics. Accessibility of social media at any time anywhere due to the internet and androids provided a platform for the employees. Therefore, they consider it a resource and engage in social media activities. Such as by interacting with online friends, watching entertainment videos, sharing, and liking contents of interest. In this way, in the beginning, individuals spend their time on social media to avoid feeling stressed. Later on, gradually, it becomes an individual's habit. Consequently, spending time on social media increases and leads to addiction (Blackwell et al., 2017;Bowlby, 2008;Sun & Zhang, 2021). Social media addiction further creates distance between individuals who are online engaged and those who are physically present (Alvarez-Rodriguez et al., 2021;Greijdanus et al., 2020). Individuals with social media addiction ignore physically present individuals and pay less attention to physically present things, which portrays their inattentive behaviour. This behaviour separates social media addicts from others and leads to further disagreements. Based on the above discussion, we have postulated the first and second hypotheses of this study.
H 1 : Job stress has a positive effect on social media addiction H 2 : Social media addiction increases interpersonal conflicts at the workplace According to the COR theory (Hobfoll, 1989), individuals seek to conserve previous resources and acquire new resources in order to use them at the time of need. Employees use social media as a resource in two ways. First, to reduce the effect of job stress on their health and wellbeing, it can be done by sharing on social media the negative stuff in their heads related to jobs. Second, through social media, employees strengthen online connections with workmates and superiors by getting positive reactions to their posts. Through this method, individuals try to manipulate superiors' behaviour in their favour. Therefore, job stress stimulates individuals to perform work-related activities on social media websites. Further, actively speaking about jobs, job problems, and peers (problem creators) on social media creates gaps between the voice raiser and others at the workplace. Individuals try to avoid voice raisers, as they perceive relationships with such persons can endanger their reputation online over little disputes. Therefore, coping with job stress through social media activities related to work can create further interpersonal problems at work. Hence, we have theorised the second and third hypotheses of the study and constructed a research model for this study based on the conservation of resources theory (Hobfoll, 1989).
H 3 : Job stress positively affects employees' work-related activities on social media H 4 : Work-related activities on social media increase interpersonal conflicts at the workplace Stress is one of the most common mental disorders (Weigel & Shrout, 2021), which reduce individuals' capabilities to manage relationships. It intensifies the behaviour of blaming others. In turn, this causes interpersonal conflicts (Weigel & Shrout, 2021). Therefore, we have hypothesised: H 5 : Job stress causes interpersonal conflicts at the workplace Two further hypotheses, H 6 and H 7 , are suggested based on the discussion above and the development of the aforementioned four hypotheses (H 1 to H 4 ).
H 6 : Social media addiction mediates the relationship between job stress and workplace interpersonal conflicts H 7 : Work-related social media activities mediate the relationship between job stress and workplace interpersonal conflicts

Research design
An online survey was conducted to collect data from office employees working on administrative posts such as CEOs, finance managers, HR managers, business development and commercial managers, legal counsel, and other officials related to corporate affairs in the oil and gas industry in Malaysia. However, employees related to the marketing field (marketing managers and associates) were excluded from the study population due to the contents of their jobs and their work on social media websites. 300 questionnaires were sent to the respondents through email and the LinkedIn network. Among those, 260 filled-out questionnaires were received. Seven responses with inadequate information and the same answer for all the questions were removed. The remaining 253 (n = 253) responses were considered complete and used for statistical analysis.

Work-related social media activities (WSMA)
Items related to work-related social media activities were extracted from the exploratory study of Connor, Schmidt, and Drouin (2016) and modified according to the context of this study. All the items were measured on a 5-point Likert scale: totally positive (5), positive (4), neutral (3), negative (2), and totally negative (1). The included items were as follows: (1) "My talk about work on social media tends to be about co-workers." (2) "My talk about jobs on social media tends to be about the job itself." (3) "My talk about jobs on social media tends to be about customers." (4) "My talk about work on social media tends to be about injustice at work." (5) "My talk about jobs on social media tends to be about working conditions." (6) "My talk about jobs on social media tends to be about employers." The reliability and validity of the scale were measured through Cronbach's alpha. The value of 0.87 confirmed its reliability and validity.

Social media addiction (SMA)
The scale of social media addiction was adopted from the study by Hawi and Samaha (2016), which contained 8 items, "1" strongly disagree to "5" strongly agree. The included items were as follows: (1) "I often think about social media when I am not using it." (2) "I often use social media for no particular reason." (3) "Arguments have arisen with others because of my social media use." (4) "I interrupt whatever else I am doing when I feel the need to access social media." (5) "I feel connected to others when I use social media." (6) "I lose track of how much I am using social media." (7) "The thought of not being able to access social media makes me feel distressed." (8) "I have been unable to reduce my social media use." Cronbach's alpha is used to measure the consistency and reliability of the scale, and the value is 0.933, which indicates good internal consistency.

Job stress
Job stress was measured by six items on a 5-point Likert scale ranging from 1 "strongly disagree" to 5 "strongly agree". The scale items were adopted from the study of Dodanwala and Santoso (2022). The included items were as follows: (1) "I feel emotionally drained by my job." (2) "I feel burned out by my job." (3) "I feel frustrated at my job." (4) "I feel tense at my job." (5) "I feel appetite because of my job-related problems." (6) "Job-related problems make my heart beat faster than usual." This job stress scale has good psychometric properties, as confirmed by Firth et al. (2004). Cronbach's alpha is used to measure the consistency and reliability of the scale, and the value is 0.813.

Workplace interpersonal conflict (WIC)
The construct of workplace interpersonal conflict was adopted from the study of Wright et al. (2017) and adapted within this study context. The items were measured on a 5-point Likert scale ranging from 1 "never" to 5 "very often". The included items were as (1) "I feel unfair treatment by others at work." (2) "I disagree with others over the work I do." (3) "I feel a lack of respect or underappreciated by others at work." (4) "My co-workers treat me with hostile or rude behaviour." (5) "My co-workers yell at me." (6) "My co-workers criticise me for things I am not responsible for." Reliability and validity of the scale were ensured through Cronbach's alpha (0.757).

Data analysis
Structural Equation Modelling (SEM) was employed to conduct statistical analysis. However, before proceeding with SEM, the data were tested for the common method bias. The problem with the common method can occur when the collected responses vary due to the instrument rather than the predisposition of the respondents (Kock, 2020). It can affect the robustness of statistical results. Therefore, data were tested for common method bias using Harman's single-factor analysis. Results revealed that one component loaded 18.2% of the variance, which is lower than 50% (Aguirre-Urreta & Hu, 2019), which confirmed that the data are free from common method bias and ready to proceed for further analysis.
Smart PLS is used for the main analysis of the data. In the beginning, the reliability and validity of the relationship between indicators and associated latent variables were assessed through measurement modelling. Study hypotheses were tested through structural equation modelling.

Descriptive analysis
The descriptive analysis is presented in Table 1. Men and women accounted for 69.6% and 30.4% of respondents, respectively; among them, most of the respondents were aged 31-35 years. As concerned about education background, most (46.6%) of the employees were graduated and 36% were with Master's degree. A first-large number (63.2%) of respondents were working on a contract basis; a second-large number (19.4%) of respondents were working on a permanent basis; and 17.4% were part-timers. Based on ethnicity, 73.5% of the respondents were Malay, whereas the percentages for Chinese and Indians were 9.5% each. 7.5% of the respondents were foreigners with varied ethnicity and classified as others.

Reliability and validity statistics
Initially, the reliability and validity of the model were assessed through measurement modelling. Internal consistency was measured through composite reliability (Kalkbrenner, 2021). Table 2 presents composite reliability and shows that the internal consistency of all the constructs exceeded the recommended threshold value of 0.7 (J. F. Hair et al., 2017). Item loadings (ranging from 0.6 to 0.9) of all the constructs show that the items are consistent with what they intend to measure. The exclusion of the items below loading 0.7 did not increase the average variance extracted or the composite reliability, so the items were retained in the model (J. J. F. Hair et al.,  Table 2, which confirm the convergent validity of the constructs and indicate that items are theoretically related to constructs (J. F. Hair et al., 2021).
Discriminant validity is assessed through the Fornell-Larcker criterion and the Heterotrait-Monotrait (HTMT) Ratio (Hamid et al., 2017). The method of the Fornell-Larcker criterion for discriminant validity was criticised due to its accuracy issues, whereas the HTMT ratio is considered reliable to measure discriminant validity (Franke & Sarstedt, 2019;N. Kanwal, A. S. N. B. Isha, et al., 2022). However, in this study, we have used both methods to measure discriminant validity. For the Fornell-Larcker criterion, the square root of AVE is presented in bold and diagonal form, and the values are greater than the correlation among the constructs (Table 3), which indicates that the results of the study satisfy the Fornell-Larcker criterion and establish discriminant validity. The results of the HTMT ratio provided statistical support to confirm discriminant validity. As the results show in Table 3, all the values are less than the HTMT threshold value .85 (Rasoolimanesh, 2022).

Structural modructural model
Study hypotheses were tested through a structural model after ensuring the reliability and validity of the data through measurement modelling. We proceeded with the structural model. It was identified that job stress has 47.9% (β) positive effect on social media addiction with ±.042 standard deviation. It accounts for 23% (R 2 ) variation in social media addiction. f 2 denotes the effect size, the result shows that 29.8% (f 2 ) change in R 2 is due to job stress, which is categorised  The effect of job stress on social media activities related to work is identified as 37% (β) positive with ±.062 standard deviation. The strength of association between job stress and social media activities related to work is 13.7% (R 2 ). The effect size is 15.9% (f 2 ), which is categorized as moderate to substantial (Goulet-Pelletier & Cousineau, 2018). Social media activities related to work increase workplace interpersonal conflict by 38% (β) with standard deviation ±13. Social media addiction and work-related activities on social media accounts for 32.7% variation in the dependent variable workplace interpersonal conflict. The effect size (f 2 ) for both the variables social media addiction (f 2 = 0.137) and social media activities related to work (f 2 = 0.174) is found to be significant. The direct effect of job stress on workplace interpersonal conflict is found to be insignificant. However, job stress significantly positively affects workplace interpersonal conflict through social media addiction and work-related activities on social media. Table 4 presents results of the structural model.
Results of the indirect effect of job stress on workplace interpersonal conflicts are presented in Table 5, which illustrates that job stress has a positive effect (β = 13.3%) on workplace interpersonal conflicts with a standard deviation of ±0.036 at the 1% level of significance through social media addiction. Likewise, the indirect effect of job stress on workplace interpersonal conflict through work-related social media activities is also positive, with the β coefficient "16.2%" and standard deviation ±0.047 at the 1% level of significance.

Discussion and conclusion
We conceptualise the framework of stress coping through social media use and its consequences. We have investigated job stress, which stimulates social media use (social media addiction and social media activities related to the workplace) and leads to interpersonal conflict at the workplace. The results provide statistical support to confirm our study hypotheses (1-4). The results reveal that stressed employees suffer workplace interpersonal conflicts through addiction and  The study findings related to the first hypothesis show job stress has the greatest effect on employee behaviours related to social media addiction. Previous studies (Zhao & Zhou, 2021) support the findings of this study. However, the relationship between stress and social media addiction was conceptualised in a social and pandemic context in earlier studies. Furthermore, the second hypothesis-that social media addiction causes interpersonal conflicts at work-has been confirmed through empirical results, which extends the findings of previous research by taking interpersonal conflict at work as a dependent variable. Earlier research in this regard demonstrated social media addiction's effect on performance, health and well-being, technostress, and technological conflicts Hou et al., 2019;Saleem et al., 2021;Sun & Zhang, 2021).
Results of the third hypothesis reveal employees' work-related activities on social media as a stress coping strategy. We have identified that job stress stimulates employees to talk about workplace problems and people on social media. However, their activities on social media about work and workers create further problems for them. For example, the fourth hypothesis establishes a positive relationship between work-related social media activities and interpersonal conflicts at work. Previous research in this regard supports the findings of this study (Valkenburg & Peter, 2007;van Ingen & Wright, 2016;Wolfers & Schneider, 2021). The fifth hypothesis of the study was about the relationship between job stress and interpersonal conflicts at work. The results did not provide statistical support to accept this relationship. However, job stress causes interpersonal conflicts at the workplace through social media addiction and work-related activities on social media. The results related to hypotheses 6 and 7 evidenced the indirect effect of job stress on workplace interpersonal conflicts through social media addiction and work-related activities on social media, which reaffirms the results of hypotheses 1 to 4.
Overall, it was concluded that employees use social media as a result of job stress. They cope with job stress through social media addiction and sharing work-related problems on social media, which in turn create interpersonal conflicts at the workplace. The findings of the study indicate that a lack of social support at the workplace stimulates employees to seek social support online through social media. Social media behaviours like addiction and sharing about workplace problems expand detachment among employees. When individuals engage in social media, they may become so focused on their virtual connections that they neglect the people around them. This situation disconnects individuals from those physically present and results in missing out on important social cues and opportunities for bonding and building relationships in person. Physical presence with an intellectual absence leads to conflicts of opinion and interpersonal conflicts.

Theoretical and practical implications
The study has enriched previous literature by identifying social media as a resource depletion in an occupational setting and measuring the mediating effect of social media addiction and workrelated social media activities between job stress and interpersonal conflict at the workplace. The study findings have three practical implications for organisational stakeholders. First, given the augmented effect of job stress on social media use (social media addiction and work-related activities on social media), measures should be taken on an administrative level to manage job stress in the organisation. Second, given the increased effect of social media use (social media addiction and work-related activities on social media) on interpersonal conflicts in the workplace, discourage social media use as a stress coping strategy in the workplace. Social media activities related to work indicate forthcoming stress for the employee and for the organisation. Therefore, careful behaviour on social media is required to avoid future workplace conflicts (N. . Third, the behaviour of coping with stress through social media use indicates a lack of social support at the workplace. Therefore, active social support at the workplace is indeed necessary to eliminate social media use as a stress coping strategy.

Study limitations and future research directions
This study has some limitations and future research directions. First, related to the conceptual framework, the behaviour of social media use can vary based on an individual's personality. A social media user's personality can be taken into account as a moderating factor in a future research model. Second, related to the study sample, this study is restricted to the population of office employees, excluding marketing and other technical workers. In the future, investigations can include a diverse sample for further generalisation of the results.