Effects of Social Media Usage on Job Crafting for Female Employees during the COVID-19 Pandemic: Evidence from China

Abstract

During the COVID-19 pandemic, the use of social media platforms for working online has become a global phenomenon. For female employees, social media usage has made it easier to balance work and family, but the pressures have also multiplied. Especially in China, the COVID-19 prevention policies led to dramatic changes in working patterns, which significantly affected the emotions, cognitions, and behaviors of female employees. Based on the job demands resource theory, this study explores the double-edged effect of social media usage on job crafting for female employees. Survey data were collected from 563 female employees in Chinese enterprises. Hierarchical regression analysis and the bootstrap method by SPSS and AMOS software were used to test the hypotheses. The results show that work-related social media usage has a negative effect on job crafting, while social-related social media usage has a positive effect. Moreover, job autonomy and workplace friendship mediate the relationship between them. This study extends the research on the double-edge effect of female employees’ social media usage and enriches the antecedents and influencing mechanisms of job crafting. It also provides theoretical and practical guidance for managers on how to promote the sustainability of human capital during the COVID-19 pandemic.

1. Introduction

With the development of artificial intelligence and information technology, the working pattern of employees, such as working hours, office location, and communication methods, has been greatly changed [1]. Especially under the COVID-19 pandemic, employees’ reliance on online office platforms, such as WeChat, ZOOM, and the OA system, has become a global phenomenon. At this time, the concept of social media usage has attracted the attention of scholars [2,3]. In China, influenced by special COVID-19 prevention policies (e.g., staying at home, quarantining), the unnecessary personnel movements and aggregation activities are suspended. The resurgence of COVID-19 has led to the need for employees to be prepared to work from home or quarantine. Moreover, a large number of commercial activities such as production, sales, communication, and negotiation are restricted by China’s local lockdown policy. More and more individuals have to use social media platforms within their organizations as a new means to handle their job tasks, promote communication, and enhance collaboration [4]. Undoubtedly, the COVID-19 pandemic accelerates the transformation of working patterns and personnel structure [5,6], which brings great challenges to the sustainable development of human capital [7]. There is an urgency for organizations to give full play to the opportunities and challenges brought by work reform and to take corresponding measures to promote the sustainability of human resources [8]. Enterprise managers should implement a series of managerial strategies and working methods that are conducive to the organization’s sustainable competitiveness, such as creating a harmonious working environment and formulating suitable management policies to ensure the physical and mental health of employees [9]. Only by fully realizing the sustainable development of “Human–Job–Environment” can enterprises effectively mitigate the impact of the COVID-19 pandemic and achieve their economic, social, and environmental performance goals.

Theoretical researchers and management practitioners are also increasingly concerned about how social media usage affects the sustainability development of organizations and individuals [10,11,12]. Previous studies showed that this online working pattern based on social media not only breaks the limitation of time and space, but also greatly reduces the communication cost, makes the work more convenient, and is conducive to improving the job satisfaction and performance of employees [13,14,15,16]. However, the COVID-19 pandemic has also given rise to working models such as “fingertip working” that require quick response from employees. This directly aggravates the phenomenon of “potential overtime” in the organization, which distracts employees’ attention and reduces their work performance [17,18]. In addition, social media usage all the time invisibly increases the work intensity of employees in the family area and breaks the balance between work and family [19]. It can be seen that although social media usage has increased the work flexibility in the organization, it still extends the work to the family area, which increases work stress, anxiety, burnout, and other negative emotions of employees [20,21,22,23].

At the same time, compared with male employees, female employees still take more responsibility for housework, so they need more time and energy to meet the needs of the family [24]. Fierce workplace competition during the COVID-19 pandemic also requires female employees to devote themselves to the work to meet the job demands. Even with frequent video calls, female employees often feel like they are missing out on important events in organizations. Social media usage has made it easier for some female employees to balance work and housework, but their work and family responsibilities have also multiplied. This pressure increases job uncertainty and poses greater challenges to the career development of female employees. Some scholars proposed that employees may take a series of positive activities when confronted with an uncertain environment, such as job crafting, to cope with the current situation [25,26]. Especially in the face of great changes in working patterns caused by COVID-19, job crafting has received wide attention by academia and industry [27,28,29]. As an organizational behavior that helps individuals to match their jobs, job crafting refers to the behaviors that employees use to voluntarily and actively redesign their job specifications, change their work tasks, and rebuild interpersonal relationships [30]. Existing studies mainly focus on the outcomes of job crafting, such as individual performance, work engagement, and well-being [31,32,33,34], while there is a significant lack of research on the antecedents and influence mechanism of job crafting. Through a literature review, it was found that the current research of scholars mainly discusses the influence of demographics, individual personalities, and job characteristics on employees’ job crafting [35,36,37]. For example, Clement and Noxolo (2016) investigated the relationship between Big Five personalities and job crafting [38]. Some scholars proposed that job crafting is not only related to intrinsic motivation [39,40], but also closely related to the external work situation [41]. At this time, as a typical work situation during COVID-19, how does the female employees’ social media usage affect their job crafting? Is the effect positive or negative? How can organizations increase the competitiveness of female employees through job crafting, thereby further enhancing the sustainability of human capital?

To answer the above questions, this study introduces the two-dimensional model of social media usage proposed by Gonzalez et al. (2012) to analyze the double-edged effect of female employees’ social media usage on job crafting [42]. Scholars proposed that social media usage is divided into two dimensions based on different purposes: work-related social media usage and social-related social media usage. The former refers to the use of social media by employees to engage in work-related activities, such as information transmission, work arrangement, and knowledge sharing. It facilitates resource generation, team collaboration, and job engagement [43]. The latter refers to the use of social media by employees to build harmonious interpersonal relations, realize the exchange of personal information, and provide emotional support. In studies on social media usage, scholars pay more attention to the influence mechanism of job characteristics. For example, a large number of studies showed that social media usage had an impact on employee behavior and performance through job states, such as job satisfaction, job engagement, and so on [44,45,46]. Only a few studies focused on the field of “interpersonal characteristics” and ignored the influence of gender and situational factors [47]. In fact, for female employees, they focus more on “interpersonal relations” and “emotional support”. That requires us to pay more attention to how social media usage affects the behavior of female employees through interpersonal mechanisms. Therefore, based on the job demands resources theory (JD-R theory) [48], we take female employees’ social media usage as job demands and job resources, respectively, to estimate its impact on job crafting. Moreover, job autonomy and workplace friendship from job characteristics and interpersonal characteristics is used to uncover the double-mediation mechanism of female employees’ social media usage and job crafting.

In sum, this study mainly explores how the changes in working patterns affect the sustainability of human capital during the COVID-19 pandemic. We try to discuss how companies could properly use social media to ensure that female employees can effectively balance their job demands and resources, so as to achieve better “Human–Job–Environment” sustainability. This study attempts to make contributions in the following three aspects. Firstly, since the COIVD-19 pandemic has a greater impact on the work and life of female employees than male employees, this study focuses on the impact of female employees’ social media usage on job crafting under the COVID-19 pandemic in China. Secondly, different from previous research, this study uses two dimensions of social media usage, namely, work-related and social-related, to analyze their double-edged effects on job crafting, which is helpful to expand the relevant research in this field. Thirdly, this study investigates the mediating roles of job autonomy and workplace friendships, which enable us to better understand the double-mediation mechanism between social media usage and job crafting.

2. Theory and Hypotheses

2.1. Social Media Usage and Job Crafting

Based on the previous studies, Chen and Wen (2019) proposed that the differences between work-related and social-related social media usage were mainly reflected in motivations, behaviors, information, and socialization. Employees use work-related social media to handle work-related matters, such as information transmission, work arrangement, and knowledge sharing [45,49]. Social-related social media usage is used to exchange personal information and maintain friendly social relationships. Employees can use these media platforms to listen to, encourage, and support others [50].

Job crafting was first proposed by Wrzesniewski and Dutton (2001). It is a “bottom-up” way of work design in the organization, in which employees can take the initiative to adopt a series of positive behaviors to change the job tasks, optimize the working patterns, and promote the working relationship [51]. With the surge of complexity and uncertainty in the external environment, especially under the COVID-19 pandemic, enterprises increasingly need to enable employees to promote organizational changes and improve environmental adaptability through job crafting [28]. At the same time, employees can also achieve higher work efficiency and job satisfaction through job crafting [52,53]. Therefore, job crafting is of great significance to realize the integration between organizational sustainable development and individual performance [54,55].

This study attempts to analyze the impact of social media usage on job crafting by the JD-R theory. According to Demerouti et al. (2001), any job characteristics can be divided into job demands and job resources [48]. Job demands require employees to continuously make physical and mental efforts for work-related tasks, which are negative factors that consume individual energy in the workplace, such as work overload and role conflict. Job resources are positive factors that help individuals achieve job goals, reduce the cost losses, and stimulate personal growth. In fact, social media usage can be seen as a job demand that requires female employees to continuously invest resources, or as a platform to provide them with job resources [13,56]. Therefore, this study proposes that social media usage has a double-edged effect on job crafting for female employees.

On the one hand, the work-related social media usage will increase the frequency of female employees’ use of social media in non-working hours, which will increase the job demands, occupy the family time, and easily lead to burnout, stress, and other negative emotions [20,23,57]. Compared with male employees, female employees need to spend more time to balance work and family. Especially in the context of the COVID-19 pandemic, work-related social media usage requires female employees to use family time for work, resulting in higher work demands and stress for them. At this time, as a proactive behavior of female employees to change work tasks, cognition, and relationships, this “bottom-up” job crafting is often accompanied by errors and risks [54]. Therefore, under the uncertain changes of the external environment such as COVID-19, when female employees face higher work demands and pressures, they are more inclined to adopt a step-by-step work mode to avoid errors and risks caused by the changes in task, cognition, and relationships, and thus have lower willingness to job crafting. On the other hand, the social-related social media usage can help female employees obtain interpersonal information, help them form social networks, and bring emotional benefits [58]. Different from the traditional interactions, the social-related social media usage breaks the boundary of managerial level, helps to build social interaction and interpersonal relationships between superiors, subordinates, and colleagues, and makes female employees feel supported and trusted [59,60]. Female employees can gain sufficient energy through effective social interaction to improve individual psychological perception and positive emotional state, so that they can obtain higher job autonomy and psychological security [61]. Therefore, in the face of environmental changes brought about by the COVID-19 pandemic, female employees with high psychological security are more inclined to view uncertainties and changes as opportunities to revise their working patterns, so as to carry out job crafting behaviors. Thus, we propose the following hypotheses:

Hypothesis 1a (H1a).

Work-related social media usage is negatively associated with job crafting for female employees.

Hypothesis 1b (H1b).

Social-related social media usage is positively associated with job crafting for female employees.

2.2. Mediating Effect of Job Autonomy

Job autonomy originated from job design in organizational management practice. Turnner and Lawrence (1965) first proposed the concept of job autonomy [62], which was defined as “the degree of self-determination of employees in the process of finishing job tasks”. It represents the individual’s ability to control the environment and the strength of their ability. Subsequently, scholars defined job autonomy as the degree that employees perceive in their work that they can independently arrange work, decide work processes, and take individual initiative [63,64]. High job autonomy not only contributes to the growth, learning, and development of individuals, but also helps them avoid the possible task conflicts among colleagues in the organizational environment, which is an important job resource [65,66].

Based on the JD-R theory, this study indicates that work-related social media usage is negatively associated with female employees’ job autonomy. First, the widespread use of social media will increase the frequency of communication between female employees in work and non-work hours. Therefore, long-term use of social media for work can disrupt work–life balance and increase stress [67]. These excessive work pressures increase employees’ perception of “not knowing what to do”, reduce their self-control over career development, and thus reduces job autonomy [68]. Second, the COVID-19 pandemic in China has led to a shift in working mode from offline to online. The lack of work participation is also detrimental to female employee engagement. High levels of social media usage can also interfere with concentration and lead to frequent interruptions in tasks [69,70]. At this time, female employees need to spend more family resources to make up for the loss of resources [12,71], thus reducing job autonomy. Finally, work-related social media usage can also lead to excessive job demands, which can lead to job burnout, role conflict, and other negative emotions [23,57]. Previous studies have confirmed that these negative emotions can also reduce employees’ job autonomy [72,73].

However, social-related social media usage can provide social resources for female employees. When communicating with colleagues, female employees share epidemic information and discuss news through social-related social media, which can help them form a harmonious interpersonal atmosphere. A previous study found that a good interpersonal climate had positive impacts on employees’ attitudes and emotions, especially the impact on emotional support and psychological security [59]. These will further influence individual behaviors. Despite the dramatic changes in working patterns caused by the COVID-19 pandemic, female employees can meet their basic psychological needs through social-related social media usage [74,75]. This enables them to deal with the uncertainty and enhance their job autonomy. Moreover, frequent interactions between superiors and female employees through social media can promote the emotional communication with each other and thus enhance their perception of insider status [76]. At this time, the quality of the leader–members relationship will also be improved [77]. Research showed that the better the relationship between the leader and the subordinates, the higher the subordinates’ work enthusiasm, and the higher the job autonomy. Thus, we propose that:

Hypothesis 2a (H2a).

Work-related social media usage is negatively associated with job autonomy for female employees.

Hypothesis 2b (H2b).

Social-related social media usage is positively associated with job autonomy for female employees.

Employees with high job autonomy can experience more ownership and have a more direct impact on work outcomes. Some studies showed that employees with high job autonomy are less likely to have negative emotions such as job burnout, fatigue, etc. [78,79,80]. Job autonomy inspires them to take responsibility for their work and thus promotes a positive attitude in front of opportunities and challenges [59]. Therefore, in the face of a complex external environment such as the COVID-19 pandemic, female employees can actively deal with difficulties, adjust their mindset, and finish the job task effectively. In fact, the relationship between job autonomy and employee proactive behavior has been investigated by scholars [81,82,83]. For example, Chang (2021) [84] proposed that job autonomy was an important factor for employees to realize job crafting, which could further enhance their career commitment. During the COVID-19 pandemic, female employees perceive higher job autonomy, which helps them balance work and family affairs and stimulates their awareness of changing working patterns. They could develop a sense of control over their work, which helps them enhance the meaning of their work, developing a perception of responsibility. Therefore, job autonomy provides female employees with the motivation and opportunities to accomplish higher goals and encourages them to change the boundaries of tasks and relationships. Thus, it enables female employees to show more proactive behaviors, namely, job crafting. In sum, we propose the following hypotheses:

Hypothesis 3a (H3a).

Job autonomy mediates the relationship between work-related social media usage and job crafting for female employees.

Hypothesis 3b (H3b).

Job autonomy mediates the relationship between social-related social media usage and job crafting for female employees.

2.3. Mediating Effect of Workplace Friendship

A workplace friendship is an informal interpersonal relationship in an organization, which is not only different from friendship in the general context, but also transcends the general colleague relationship in the workplace [85]. It is a kind of high-quality interpersonal relationship [86,87]. In particular, in the context of Chinese organizations, influenced by Confucian culture and the conventional view of “Harmony”, employees are willing to establish workplace friendships and benefit from such positive interpersonal interaction [88]. Workplace friendships have a significant influence on individual emotion and behavior [89]. Here, this study suggests that different purposes of social media use have double-edged benefits on female employees’ workplace friendships.

On the one hand, the “instant feedback” of work-related social media usage can make employees use communication tools more frequently during non-working time. Female employees have to take on both work and family roles at home. This interferes with female employees’ normal work, disturbs work–life balance, and creates huge job demands. Such excessive job demands will lead to job insecurity, role ambiguity, and negative emotions, which further reduce the workplace friendships [90,91]. Not only that, but such job demands result in a greater loss of family resources for female employees. According to Hobfoll and Lilly (1993) [92], individuals have the tendency to acquire, maintain, and cultivate their precious resources. At this time, female employees will protect family resources from loss by reducing work-related interaction and communication. The lack of interaction and communication within the organization is a key factor affecting the formation of workplace friendships. Therefore, this study believes that work-related social media usage has a negative effect on female employees’ workplace friendships.

On the other hand, as an information exchange platform, social-related social media usage can display identity information such as social relationships, interests, and hobbies. Identity transparency enables employees to understand one another and creates conditions for improving interpersonal trust and workplace friendships [60]. Specifically, the social-related social media usage can help female employees maintain and build interpersonal relationships, effectively realize the exchange of social and personal information, and provide social and emotional support [93]. Female employees can quickly find common ground (such as the same hobbies, favorite stars, and delicious food) with colleagues. Previous studies revealed that similar interests and visions could further promote harmonious relationships and psychological safety, which are important factors in the formation of workplace friendships [94]. Therefore, workplace friendships, the informal relationships between female employees, are increased through frequent and visible social activities [59]. Moreover, the job resources brought by the social-related social media usage can help female employees establish professional and interpersonal relationships, make individuals feel safe and confident, strengthen social exchange relationships with each other [95], and thus increase workplace friendships. Hence, we propose the following hypotheses:

Hypothesis 4a (H4a).

Work-related social media usage is negatively associated with workplace friendships for female employees.

Hypothesis 4b (H4b).

Social-related social media usage is positively associated with workplace friendships for female employees.

This study argues that workplace friendships can further improve female employees’ job crafting, mainly for the following two reasons. First of all, the workplace friendships of female employees helps them increase a variety of job resources and create a mutually supportive climate [85], which provides security for job crafting behaviors. Therefore, workplace friendships can help female employees cope with the uncertain changes in the environment, tasks, and relationships that come with from job crafting and help them make better decisions. Namely, female employees with high levels of workplace friendships are more likely to change working patterns through job crafting to successfully complete their job tasks. Secondly, high-quality workplace friendships will enhance female employees’ trust in the organization [96]. This will not only meet the psychological needs of employees, but also generate more organizational identification and affective commitment [97,98]. Research showed that high affective commitment can promote employees’ self-motivation and feedback-seeking behavior [99]. Therefore, female employees with high levels of identity and commitment are more likely to view their work in a positive light and engage in the process of job redesign, namely, job crafting. Thus, we suggest the following hypotheses:

Hypothesis 5a (H5a).

Workplace friendships mediate the relationship between work-related social media usage and job crafting for female employees.

Hypothesis 5b (H5b).

Workplace friendships mediate the relationship between social-related social media usage and job crafting for female employees.

The theoretical model is shown in Figure 1.

3. Methods

3.1. Sample and Collection

In the past two years, China’s first tier cities such as Beijing have been more seriously affected by the COVID-19 pandemic than other second and third tier cities. Therefore, the female samples of this study mainly come from 11 state-owned organizations and institutions in Beijing, Shanghai, and Shenzhen. In this study, random sampling was adopted to ensure the randomness and universality of the female samples. During the data collection process (March 2022 to June 2022), we contacted the head of the human resource management department of the organizations in advance. Moreover, we explained our research purpose to the female participants and declared to them that all the questionnaires were anonymous and would not be used for personal career development. Affected by the COVID-19 pandemic, 200 questionnaires were distributed on site, and 500 questionnaires were distributed by web-links such as email, APPs, and other forms. We removed the questionnaires with a missing rate of more than 10%, and finally obtained 563 questionnaires with a recovery rate of 80.4%.

In the sample, most female employees were mainly engaged in positions of design, finance, accounting, translation, and consulting, and all of them have clear boundaries of working hours (e.g., from 8:30 a.m. to 4:30 p.m.). They have been working from home many times during the COVID-19 pandemic. They were often required to use social media platforms such as the OA system, WeChat, and Tencent Meeting for work communication. In addition, the majority of respondents indicated that they also need to deal with work-related tasks during non-working hours. All these responses showed that the COVID-19 pandemic has significantly affected the working pattern of the female participants.

As shown in

Table 1. Demographic characteristics of female respondents (N = 563).

Demographics	Category	Frequency	Percentage
Age	18–25 years old	45	8.0%
	26–35 years old	281	49.9%
	36–45 years old	190	33.7%
	Over 46 years old	47	8.3%
Education	Junior college and below	109	19.4%
	Bachelor	252	44.8%
	Master	169	30.0%
	Post Graduate	33	5.9%
Working years	10 years and below	223	39.6%
	11–15 years	203	36.1%
	16–20 years	94	16.7%
	Over 20 years	43	7.6%
Marital status	Unmarried	223	39.6%
	Married	289	51.3%
	Others	51	9.1%

, 8.0% of the female employees were 18–25 years old, 49.9% were 26–35 years old, 33.7% were 36–45 years old, and 8.3% were 46 years old and above. In terms of female employees’ education level, 19.4% had junior college degrees and below, 44.8% had bachelor’s degrees, 30.0% had master’s degrees, and 5.9% had post graduate degrees. Around 39.6% of female employees had less than 10 years of working experience, 36.1% had 11–15 years, 16.7% had 16–20 years, and 7.6% had more than 20 years. In terms of marital status, about 39.6% of female employees were unmarried, 51.3% were married, and 9.1% were “others”.

3.2. Measures

All scales (social media usage, job crafting, job autonomy, and workplace friendship) are obtained from international journals, and the bibliographic references and items are shown in Appendix A. As this study investigated the work status of Chinese female employees in the context of COVID-19, a double-blind translation method was used to design the English scale into a Chinese scale. We combined Chinese context and culture as much as possible in the translation of the Chinese scale. Then, we invited team members to translate the scale from Chinese to English again to check the differences with the original scale. After two rounds of double-blind translation, the questionnaires were preliminarily formed. Finally, in order to ensure the consistency of the questionnaires, three management professors from Beijing were invited to revise all the scales. All variables were scored on a Likert 5-point scale from 1 to 5, with 1 = strongly disagree to 5 = strongly agree. A brief description of all scales is given below.

Social media usage. We assessed social media usage using Gonzalez’s scale (2012) with two dimensions [42], namely, work-related social media usage (8 items) and social-related social media usage (5 items). This scale was widely used and showed good reliability in China. Sample items of work-related social media usage such as “I use social media to share information about organizational policies and procedures”, and social-related social media usage items such as “I use social media to take a break from work”. The Cronbach’s α of these two dimensions were 0.943 and 0.892, respectively, indicating that the scale has good reliability.

Job crafting. Petrou et al.’s scale (2012) was used to assess job crafting [30]. This scale includes 13 items such as “I try to learn new things at work”. The Cronbach’s α for this scale was 0.974.

Job autonomy. This scale was measured with Kirmeyer and Shirom’s (1986) 7-item scale [63]. For example, “I have freedom to decide how to do my work”. The Cronbach’s α for this scale was 0.974.

Workplace friendship. This scale was developed by Nielsen’s 6-item scale (2000) [75]. A sample item is “I have the opportunity to develop close friendships at my workplace”. The Cronbach’s α for this scale was 0.941.

Control variables. Previous studies proposed that the age, marital status, education level, and working years are important demographic variables affecting the emotion, cognition, and behavior of female employees. Therefore, this study controlled for these variables in the hypotheses testing process.

4. Results Analysis

4.1. Descriptive Analysis

The descriptive analysis (means, standard deviations, and correlation coefficients) was conducted by SPSS 23.0 software. As shown in

Table 2. Means, standard deviations, and correlation coefficients for each variable.

Variable	Mean	SD	1	2	3	4
1. Work-related social media usage	2.45	1.09
2. Social-related social media usage	3.54	1.07	−0.72 **
3. Job autonomy	3.66	1.06	−0.25 **	0.28 **
4. Workplace friendship	4.01	0.84	−0.25 **	0.29 **	0.33 **
5. Job crafting	3.93	0.85	−0.32 **	0.38 **	0.32 **	0.38 **

Note: ** p < 0.01.

, the majority of female employees used social media for social-related activities (mean = 3.54), and their workplace friendships (mean = 4.01) and job crafting (mean = 3.93) were at high levels. Moreover, work-related social media usage was negatively associated with job crafting, job autonomy, and workplace friendship (r = −0.32, r = −0.25, r = −0.25, p < 0.01). Moreover, social-related social media usage was positively related with job crafting, job autonomy, and workplace friendship (r = 0.38, r = 0.28, r = 0.29, p < 0.01). The results of correlation analysis have provided preliminary evidence for the hypotheses of this study.

4.2. Normality Test

This study used the Q-Q plot method to test the normality of the Chinese female samples. The Q-Q plot reflects the degree of agreement between the actual distribution and the theoretical distribution of variables, which can be used to test whether the data are normally distributed. As shown in Figure 2, the abscissa is the observed value of the variables, and the ordinate is the expected normally distributed value. Work-related social media usage, social-related social media usage, job autonomy, workplace friendship, and job crafting of Chinese female employees are normally distributed.

4.3. Confirmatory Factor Analysis

Self-reported data from female employees were used in the process of hypotheses testing. Therefore, in order to avoid the common method bias (CMB), this study drew on the method of Podsakoff et al. (2003) [100] to verify the measurement model through confirmatory factor analysis (CFA). As shown in

Table 3. The results of confirmatory factor analysis.

Model	Factors	${\mathit{\chi }}^2$	${\mathit{\chi }}^2/\mathbf{d}\mathbf{f}$	CFI	GFI	IFI	CFI
One-factor model	WSMU + SSMU + JA + WF + JC	18,218	12.74	0.43	0.26	0.43	0.15
Two-factor model 1	WSMU + SSMU + JA + WF, JC	13,133	9.19	0.60	0.35	0.60	0.12
Two-factor model 2	WSMU + SSMU, JA + WF + JC	12,783	8.95	0.61	0.38	0.62	0.12
Three-factor model 1	WSMU + SSMU, JA + WF, JC	8496	5.95	0.76	0.55	0.76	0.09
Three-factor model 2	WSMU + JA, SSMU + WF, JC	10,917	7.65	0.68	0.48	0.68	0.11
Four-factor model	WSMU + SSMU, JA, WF, JC	5138	3.61	0.88	0.85	0.87	0.07
Five-factor model	WSMU, SSMU, JA, WF, JC	4189	2.95	0.90	0.91	0.90	0.06

Note: WSMU = work-related social media usage, SSMU = social-related social media usage, JA = job autonomy, WF = workplace friendship, and JC = job crafting.

, our proposed model (five-factor model: work-related social media usage, social-related social media usage, job autonomy, workplace friendship, and job crafting) had better fit indices (${\chi }^2=4$189, ${\chi }^2/df$ = 2.95, CFI = 0.90, GFI = 0.91, IFI = 0.90, RMSEA = 0.06) than other alternative models. For example, the four-factor model (combine work-related social media usage and social-related social media usage) was conducted to test the distinctiveness of the independent variable, and the fit indices were poorer (${\chi }^2=5$138, ${\chi }^2/df$ = 3.61, CFI = 0.88, GFI = 0.85, IFI = 0.90, RMSEA = 0.07). The CFA results confirmed that there was no CMB between these five scales.

4.4. Multicollinearity Test

According to the correlation analysis in Table 1, the correlation coefficients among all variables were less than 0.7, which provided preliminary evidence for the multicollinearity test. Then, the study further tested the multicollinearity among variables by calculating the variance inflation factor (VIF). The results showed that the VIF values of work-related social media usage, social-related social media usage, job autonomy, workplace friendship, and job crafting were less than 2 (VIF > 10 indicates multicollinearity). Therefore, there was no multicollinearity between the variables.

4.5. Reliability and Validity

To ensure the reliability and validity, the Cronbach’s α, composite reliability (CR), and average variance extracted (AVE) of each variable were estimated. As shown in

Table 4. Reliability and validity for each variable.

Variable	α	CR	AVE
1. Work-related social media usage	0.943	0.86	0.59
2. Social-related social media usage	0.892	0.81	0.51
3. Job autonomy	0.974	0.91	0.62
4. Workplace friendship	0.941	0.82	0.55
5. Job crafting	0.950	0.89	0.57

, all CR values were above 0.8, and all AVE values were above 0.5, indicating that the scales had good reliability and validity [101].

4.6. Tests of Hypotheses

Hierarchical regression model was used to test the proposed hypotheses, and the results are shown in

Table 5. The results of hierarchical regression analysis.

Variable	Model 1	Model 2	Model 3	Model 4
	Job Crafting	Job Autonomy	Workplace Friendship	Job Crafting
Age	0.019	0.021	0.077	0.095
Education level	−0.016	0.040	0.079	−0.039
Working years	0.052	0.072	0.094 *	0.090 *
Marital status	0.042	0.026	0.039	0.013
Work-related SMU	−0.253 **	−0.236 *	−0.280 **	−0.196 *
Social-related SMU	0.412 ***	0.307 **	0.309 **	0.245 **
Job autonomy				0.092 *
Workplace friendship				0.126 **
R2	0.306	0.216	0.286	0.326
F	40.78 ***	25.51 ***	37.09 ***	33.44 ***

Note: * p < 0.05, ** p < 0.01, *** p < 0.001. Bootstrapping = 5000 times. SMU = social media usage.

. In Model 1, after controlling for age, education level, working years, and marital status of female employees, work-related social media usage was negatively associated with job crafting (β = −0.253, p < 0.01), while social-related social media usage was positively related to job crafting (β = 0.412, p < 0.001). Thus, H1a and H1b were supported. In Model 2 and Model 3, two similar regression models were conducted to investigate the relationship between social media usage and the mediators. The results in Model 2 showed that work-related social media usage was negatively associated with job autonomy (β = −0.236, p < 0.05), while social-related social media usage was positively related to job autonomy (β = 0.307, p < 0.01); thus, H2a and H2b were supported. In Model 3, the coefficient between work-related social media usage and workplace friendship was significant (β = −0.280, p < 0.01), as was the social-related media usage (β = 0.309, p < 0.01). Thus, H4a and H4b were also supported. In Model 4, the demographic variables, independent variables, and mediators were entered into regression. It was found that job autonomy and workplace friendship were significantly related to job crafting (β = 0.092, p < 0.05; β = 0.126, p < 0.01). After controlling for job autonomy and workplace friendship, the effect of work-related social media usage on job crafting was significantly decreased from −0.253 (p < 0.01) to −0.193 (p < 0.05), and the influence of social-related social media usage on job crafting was significantly decreased from 0.412 (p < 0.001) to 0.245 (p < 0.01). Based on the three steps of the mediation effect test proposed by Baron and Kenny (1986) [102], these findings provided preliminary evidence for the mediating effects of job autonomy and workplace friendship.

Furthermore, this study adopted the bootstrap analysis proposed by Hayes (2003) [103] to test the double-mediating effects of job autonomy and workplace friendship. A bias-corrected 95% confidence interval (CI) was estimated by the bootstrap samples (=5000 times). The results are shown in

Table 6. Bootstrapping mediation testing results.

Pathway	β	SE	95% Confidence Interval
			Lower	Upper
Path 1: Work-related SMU $\to$ Job autonomy $\to$ Job crafting	−0.0195 *	0.011	−0.049	−0.005
Path 2: Social-related SMU $\to$ Job autonomy $\to$ Job crafting	0.0195 *	0.012	0.003	0.051
Path 3: Work-related SMU $\to$ Workplace friendship $\to$ Job crafting	−0.0246 **	0.013	−0.060	−0.006
Path 4: Social-related SMU $\to$ Workplace friendship $\to$ Job crafting	0.0272 **	0.014	0.007	0.063

Note: * p < 0.05, ** p < 0.01, SMU = social media usage.

. In Path 1, job autonomy had a significant indirect effect between work-related social media usage and job crafting (β = −0.0195, p < 0.05, 95% CI = [−0.049, −0.005], exclude 0). In Path 2, job autonomy had a significant indirect effect between social-related social media usage and job crafting (β = 0.0195, p < 0.05, 95% CI = [0.003, 0.051], exclude 0). Similarly, the path coefficients of workplace friendship were both statistically significant (in Path 3, β = −0.0246, p < 0.01, 95% CI= [−0.006, −0.06], exclude 0; and in Path 4, β = 0.0272, p < 0.01, 95% CI = [0.007, 0.063], exclude 0). These results indicated that job autonomy and workplace friendship mediate the relationship between social media usage and job crafting. Therefore, H3a, H3b, H5a, and H5b were all supported.

In order to more directly reveal the impact of social media usage on job crafting and the intermediate mechanism of job autonomy and workplace friendship, the results are summarized in Figure 3.

5. Discussion

Most Chinese organizations during the COVID-19 pandemic have changed their working patterns from “offline” to “online”. This change has significantly affected the cognition, emotion, and behavior of female employees in China, which has brought great challenges to the sustainability of organizational human capital. This study investigated the relationship between social media usage and job crafting of female employees during the COVID-19 pandemic. The mediating roles of job autonomy and workplace friendship on this relationship were also explored. Our findings reveal that work-related and social-related social media usage, the two different dimensions of social media usage, have opposite effects on job crafting of female employees.

Firstly, our study builds on previous research by scholars [2,3,10,11,12,13,14,15,16,17] looking at the effect of social media usage on female employees’ job crafting during the COVID-19 pandemic. Different from previous studies, our results indicate that the social media usage by female employees for different purposes has a double-edged effect on job crafting. It reflects the nature of female employees’ tendency to reduce job demands and access to job resources. Specifically, when female employees use social media platforms for work purposes, the frequency of work communication in non-working hours will be increased. Based on JD-R theory, this information overload makes female employees perceive more job demands. Thus, negative effects such as burnout and stress will be easily generated [20,21,22,23,24], which will significantly reduce their job crafting. In contrast, previous studies have shown that social-related social media usage could break the boundaries of work and strengthen the positive interpersonal relationships between colleagues [60,61]. These interactions help employees obtain more job resources [59], which promotes job crafting. Our finding is consistent with these studies; that is, social-related social media usage of female employees is positively related to job crafting.

Secondly, this study sheds light on the different mediating effects of job autonomy between social media usage and job crafting. During the COVID-19 pandemic, the widespread use of social media increased the frequency of communication between female employees in work and non-work hours. Our results show that work-related social media usage is negatively related to job autonomy, which further decreases the female employees’ job crafting. One possible explanation is that female employees have taken on more family responsibilities than male employees during the COVID-19 pandemic. Frequent use of social media for work increases their work demands, breaks the work–family balance, and promotes their negative emotions such as burnout and exhaustion [20,23,57]. All of these reduce their job autonomy. However, social-related social media usage is positively related to job autonomy. This lines up with the previous studies, in which employees who perceived a higher level of social-related social media usage showed more positive emotions and attitudes, such as psychological security, well-being and mental health status, and so on [59,61,74]. Social communication can help female employees establish, maintain, and develop interpersonal relationships, increase employees’ positive emotions and social support, and create a good working atmosphere. As a job resource, it increases the job autonomy and job crafting of female employees.

Thirdly, we found that workplace friendship mediates work-related/social-related social media usage and job crafting. Work-related social media usage of female employees can increase their work stress and emotional exhaustion, which further depletes their family resources. Especially during the COVID-19 pandemic, it can be difficult for female employees to balance work and family. They tend to cut down on work-related social media usage and avoid unnecessary work communication, which can be detrimental to the formation of workplace friendships. However, social-related social media usage creates conditions for the development of workplace friendships. Even in the home-based working pattern during the COVID-19 pandemic, social-related social media usage can help female employees effectively communicate social and personal information, build and maintain interpersonal relationships [93], and thus improve workplace friendships. This finding is in line with previous studies [59,60,93]. Moreover, we found that the workplace friendships of female employees are an important antecedent variable of organizational identification and affective commitment, which further enhance female employees’ job crafting. This is consistent with the previous research [97,98,99]. Therefore, workplace friendships of female employees play different mediating roles in the relationship between work-related/social-related social media usage and job crafting.

5.1. Theoretical Implications

There are three theoretical implications. First, previous studies on social media usage have focused on the variables related to work or family, and have mainly investigated how social media usage affects individual emotions, cognitions, and behaviors [2,3,10,11,12,13,14,15,16]. In response to the call of Hruska and Maresova (2020) [104], this study takes gender and environmental background as important factors to investigate how the social media usage of female employees influences individual behavior under the special COVID-19 pandemic prevention policy in China. The COVID-19 pandemic increases job uncertainty and poses greater challenges to the career development of female employees. Therefore, social media usage is both a job demand and a job resource, and the impact on female employees is obviously different from that on male employees. This study makes up for the lack of research on gender and background perspectives in social media usage.

Second, previous studies have explored the effects of social media usage on job performance, job satisfaction, job engagement, and work–family conflict, and obtained relatively consistent conclusions. At the same time, as a transformative behavior, scholars have estimated its influencing factors from the individual level, team level, and organizational level. However, existing studies proposed that there were few studies exploring the impact of social media usage on job crafting. Therefore, our study examines the impact of the social media usage of female employees on job crafting, which enriches the empirical research on the outcomes of social media usage, as well as the antecedents of job crafting. Moreover, based on the JD-R theory, this study explores the double-edged impact of social media use on job remodeling from the perspectives of job demands and job resources. This further enriches the relevant theoretical achievements.

Third, the mediating effects of job autonomy and workplace friendship are estimated, further enriching the research on the double-mediation mechanism of social media usage and job crafting for female employees. Based on previous research, this study integrates the perspectives of work and socialization to explore the relationship between social media usage and job crafting. We find that for female employees, work-related social media usage has a negative effect, while social-related social media usage has a positive effect on their job autonomy and workplace friendships. In addition, the results reveal that the mediating effect of workplace friendships are significantly greater than the job autonomy. Hence, our findings not only lay a theoretical foundation for future research on the mechanism between female employees’ social media usage and job crafting, it also expands the application of job autonomy and friendship.

5.2. Practical Implications

First of all, since social media usage has the dual attributes of “work” and “social”, organizational managers should pay more attention to how to effectively use social media platforms to balance such demands and resources during the COVID-19 pandemic. Female employees are encouraged to communicate with each other on social media platforms. For example, the “Enterprise WeChat” provides the “corporate contacts” function, which allows them to quickly find other employees in the organization and start a chat without adding friends, which greatly promotes information exchange among individuals. At the same time, managers should make flexible and suitable policies to allow female employees to properly use social media for leisure and social interaction during working hours, and to guide them to better balance the relationship between work and family. When the interpersonal communication needs of female employees are met, they can be more engaged in their jobs, which further helps them to improve job crafting behaviors.

Moreover, the study also finds the “dark side” of social media usage in organizations; that is, female employees’ social media usage, especially work-related use, will reduce their job autonomy and workplace friendships. Therefore, managers should formulate a series of strategies to manage female employees’ work-related social media usage and avoid work interference and overload caused by the excessive use of social media.

Our findings reveal that female employees’ workplace friendships are positively related to job crafting. Therefore, managers should create a pleasant and harmonious interpersonal atmosphere and provide opportunities for female employees to get to know each other. For example, managers could combine the COVID-19 pandemic prevention policy to adjust the working environment, so that female employees can work in a free and relaxed environment. Moreover, relevant managerial strategies should be formulated to promote communication and social interaction, so as to reduce the possible conflicts and frictions between female employees; thus, further enhancing their job crafting.

5.3. Limitations

Although the double-edged relationship between female employees’ social media usage and job crafting is confirmed by the JD-R theory, this study also has the following limitations. First, this study focuses on female employees’ social media usage and job crafting in China. In fact, China’s COVID-19 prevention policies are significantly different from those of other countries. Therefore, the findings obtained in this study should be further extended to other countries. Second, industry differences are not considered when examining the relationship between social media usage and job crafting. In China, female employees are predominantly employed in the service sector, in which COVID-19 has had the most effect. Different industries may lead to different results. Therefore, in the future, it is necessary to further investigate the impact of the industries engaged in by female employees. Third, the cross-sectional data were used in the data analysis process, which may lead to unclear causal relationships between variables. Longitudinal surveys could be considered in future studies. Last but not least, job autonomy and workplace friendship were selected as mediating variables. Moderators were not considered in this study. Future research can explore the moderating effect between social media usage and job crafting, such as female employees’ personalities, organizational climate, leadership styles, and other potential variables.

6. Conclusions

This study investigated the impact of social media usage on job crafting of female employees based on the COVID-19 background. We found that work-related social media usage was negatively related to job crafting, while social-related social media usage was positively associated with job crafting. It also revealed that job autonomy and workplace friendships mediate social media usage and female employees’ job crafting. This study provides empirical evidence and new insights into the different purposes of social media usage by female employees to promote job crafting.