Relationship Between Work-Family Conflict, Job Embeddedness, Workplace Flexibility, and Turnover Intentions

The present study seeks to propose and test a research model that investigates job embeddedness as a mediator and workplace flexibility as a moderator of the effect of family-work conflict on turnover intentions. This study uses a survey method and a structured questionnaire to collect data from 187 nurses working in various hospitals in Islamabad, Pakistan. The results showed that on-the-job embeddedness partially mediated the effect of work-family conflict on nurses' turnover intention. Furthermore, workplace flexibility moderated the relationship between work-family conflict and turnover intention. Management of the hospitals should take decisive steps to establish and maintain a supportive and flexible work environment because such an environment would help nurses to balance their work (family) and family (work) roles and lead to increased job embeddedness. Otherwise, it would be very difficult to retain high performing nurses in the workplace. The current study contributes to the existing knowledge base by testing job embeddedness as a mediator and workplace flexibility as a moderator of the impact of work-family conflict on turnover intentions of nurses.


Introduction
Nowadays, employee retention is becoming a major challenge for organizations. Turnover intention is defined as the probability that an employee will leave his or her job within a certain time period (Chao, Jou, Liao, & Kuo, 2015). High turnover intention in nursing has become a critical problem in the development of the healthcare system. Turnover has both direct and indirect costs. The direct costs include reduced performance, December 2017 | Vol. 21 | No. 2 recruitment costs, training of new hires, and paying workers overtime to cover vacated posts (Lambert & Hogan, 2009). Turnover intention is always affected by various aspects, for example job strain (Heponiemi, Presseau, & Elovainio, 2016), job burnout (Moreno-Jiménez, Gálvez-Herrer, Rodríguez-Carvajal, & Sanz Vergel, 2012), job satisfaction (Tziner, Rabenu, E., Radomski, & Belkin, 2015), organizational commitment (Lambert, Hogan, & Barton, 2001), etc.
Although employee turnover is driven by different reasons, having stressful working conditions is considered one of the most common factors in pushing employees out of an organization. Given this reality, research should delve deeper into the more intricate details of the relationship between job stress and turnover intention. An enriched perspective can be gained by unveiling the underlying cause that discourages stressed employees to stay with their employer. With this being said, numerous studies have identified work-family conflict as a significant cause of undesired influences on both the individuals and the organizations. Amongst many reasons that contribute to an employee's turnover intentions, one of the most important is work-family conflict. Work-family conflict is defined as a form of inter-role conflict that appears when it is difficult to balance the pressure of work and family (Greenhaus & Beutell, 1985). Some studies in other countries indicate that work-family conflict could positively affect turnover intention (Blomme, Van Rheede, & Tromp, 2010;Özbağ & Ceyhun, 2014). Some studies reported that there were neither direct nor indirect relationships between work-family conflict and turnover intentions (Boamah & Laschinger, 2016). Individuals find it difficult to balance their roles between family and work and thus suffer from stress, leading to dissatisfaction, burnout, and leaving intention (Shockley & Singla, 2011). Due to continuous pressure to perform better in order to compete and stay successful, the organizations heavily depend on their employees and hence transfer this pressure onto employees to work for longer hours with more commitment and involvement. In the context of hospitals, this pressure is even more aggravating. The success of a hospital depends on the quality of service provided to its patients, and a major part of that quality relates to the employees in the hospital.
In hospitals, nursing staff and doctors face emergency situations leading to unpredictable work scheduling requirements, nontraditional working hours, including evenings, nights and rotating shifts, long working hours, uncertain work situations leading to additional work, minimal control over work hours and continuous pressure to keep the patients and their attendants satisfied and happy, despite laborious and exhausting work hours (Flinkman, Leino-Kilpi, & Salanterä, 2010).
In such circumstances, when the nature and tasks of the job are overwhelming, emotionally as well as physically, it becomes extremely difficult to balance the demands of work life with the demands of family life. Under consistent pressure to balance these two sides, individuals become dissatisfied with the organization and may want to quit their job in order to find another organization with better balance (Stordeur & D'Hoore, 2007).
Research has shown that in complex, uncertain and high quality service oriented jobs, managers should provide flexibility to employees to keep them motivated to be more productive. Work life and family life can be balanced effectively if the individual has the authority to flexibly decide their working hours to avoid paradoxical situations between family and work (Podsakoff, LePine, & LePine, 2007). The alternate working arrangements such as flexible working hours and work sharing have been found to improve work family balance. The consequences of work-family conflict are well understood in literature (Eby. Casper, Lockwood, Bordeaux, & Brinley, 2005). Work-family conflict has implications for employee work performance (Haar, 2013), satisfaction and commitment (Shockley & Singla, 2011).
Much less is understood about the impact of workfamily conflict on employee turnover. Work-family conflict prompts a series of role conflicts that create personal stress (Eby et al., 2005;Farquharson et al., 2012). In some circumstances, this stress will prompt the employee to disengage from their job and the organization and begin to seek alternatives outside the organization (Greenhaus, Collins, Singh, & Parasuraman, 1997). In these cases, stress is likely to lead to increased leaving intention. In the same vein, work-family conflict dampens employee job satisfaction and affective commitment (Rode, Rehg, Near, & Underhill, 2007). This paper suggests a complementary account of how work-family conflict may lead to employee turnover, drawing on the Job Embeddedness Theory. Specifically, this paper proposes that an employee's job embeddedness mediates the effect of work-family conflict on turnover intentions. Job embeddedness refers to the formal or informal connections and perceived compatibility between a person and an institution or other people, and the perceived material and psychological costs of leaving a job (Zhang, Fried & Griffeth, 2012). Furthermore, workplace flexibility moderates the relationship between job embeddedness and turnover intentions (Figure 1). Workforce flexibility refers to the autonomy provided to employees to decide working hours, shifts, work sharing, and alternate working arrangements (Eby et al., 2005).
The arguments and evidence of this paper bring several contributions to literature. First, researchers such as Chang, McDonald and Burton (2010) have called for a broadening of the research methods used to analyze work-family conflict. This paper supplements existing work-family conflict literature by demonstrating the non-affective consequences of this conflict on employee attachment to the organization and thus, on employee turnover intention. Second, this paper adds to the Job Embeddedness Theory literature by demonstrating how work-family conflict influences the formation and evolution of employee embeddedness. Despite the rapid development of the JET literature, our understanding of the antecedents of embeddedness remains limited (Zhang et al., 2012). Finally, by understanding how work-family conflict impacts an employee's embeddedness and employee retention intentions, organizational management may become better able to assess the organizational consequences of poorly managed work-life balance.
Theoretical framework and hypotheses. Workfamily conflict and turnover intention. Work-family conflict in the nursing profession is inescapable, because the employees in this occupation experience high levels of physical, cognitive, and emotional demands (Buonocore & Russo, 2013;Shacklock & Brunetto, 2012). Both employees and institutions have been affected by the negative outcomes of work-family conflict such as increasing stress levels, lower performance, and decreasing job and life satisfaction (Adkins & Premeaux, 2012). Greenhaus and Beutell (1985) define work-family conflict as inter-role conflict that occurs when one role (work/personal life) imposes responsibilities and requirements that are not compatible with the other role. There has been a lot of research on employee turnover over the last 20 years (Griffeth, Hom, & Gaertner, 2000). Work-family conflicts influence employee behavior and outcomes at the workplace. Previous literature has found that workfamily conflict is associated with lower job satisfaction (Ayree, Luk, & Fields, 1999), burnout (Rupert, Stevanovic, & Hunley, 2009), higher rates of absenteeism (Burke & Greenglass, 1999), overall performance (Witt & Carlson, 2006), turnover intention (Bhave, Kramer, & Glomb, 2010;Cullen & Hammer, 2007) and employee turnover (Henly et al., 2006).
The interference between work life and personal life has been classified as one of the main stressors in the workplace (Gao, Shi, & Wang, 2013). Work stressors, as numerous studies have found, cause negative at-work behavior such as absenteeism and employee turnover. Conflict between family and work settings occur due to a trade off between an individual's role as an employee and as a member of a family. This tradeoff or conflict usually results in tension or stress in two distinct but inter-related dimensions. The first dimension is the tension or stress caused by conflict that arises as a result of roles, duties, responsibilities, tasks, obligations, and requirements of workplace being incompatible with roles and tasks at one's home life, known as Work-Family Conflict (Netemeyer, Boles, & McMurrian, 1996). For example, an employee is unable to fulfill his/her family roles such as parenting duties, a dine out with spouse and/or kids, weekend time with family, etc. due to long and laborious hours of work, tough and extremely important deadlines, over demanding tasks, meetings and travelling. The second dimension is the conflict experienced by employees when they sacrifice their work roles and job duties, responsibilities, tasks, and obligations due to family commitments.
As human beings with a limited amount of energy, employees have a finite number of hours to allocate to work as well as home life. It is difficult to balance between the two because as an employee gives more time to either, the other life suffers a cut back. This zero sum situation results in a paradox in one's ability to reconcile obligations to attend to both lives equally. These conflicts prove detrimental to an employee's potential to perform tasks with peace of mind, thus resulting in job dissatisfaction and lower organizational commitment, leading to increased turnover intentions. The inability to manage both of these important facets of life leads to a state of continuous stress. It is highly likely that an individual would become disengaged in work if he/she is unable to address family responsibilities and would carry that burden to work where he/she might be unable to perform as expected. This would consequently lead to lack of motivation to stay in their current organization, and he/she might decide to quit the organization in order to balance work and family lives. These aforementioned job characteristics seem to be applicable to jobs in hospitals. Hence, we hypothesize the following: Hypothesis 1: Work-family conflict is positively associated with leaving intention.

Job embeddedness as a mediator. The Job
Embeddedness Theory (JET) from Mitchell, Holtom, Lee, Sablynski, and Erez, (2001) has emerged as an effective paradigm for the analysis of employee retention (Jiang, Liu, McKay, Lee, & Mitchell, 2012). Employee embeddedness has been found to be a reliable predictor of turnover intention and to reliably account for actual turnover when used in conjunction with the Unfolding Theory of Turnover (Zhang et al., 2012). JET provides a useful framework for the analysis of the impact of work-family conflict on employee retention intentions for three inter-related reasons. First, unlike the dominant tradition of turnover theorizing -based on the dissatisfaction-job alternative framework -the JET account recognizes the role of non-affective and structural forces in creating the terrain in which retention and turnover decisions are made. Second, it explicitly recognizes an employee's life out of work. The initial formulation of the JET held the employee to exist within a Lewinsian life-space (Mitchell et al., 2001), subject to the various constraints and opportunities of the different aspects of their life. Third, it enables a context-specific analysis of how employee attachment is affected. An employee's overall job embeddedness is the net result of the employee's accumulated experiences within and outside the organization, coupled with the operational policies and practices within the organization, along with the array of other forces from their life out of work. These adhesive connections go beyond the psychological connections implicit in theories of employee engagement or organizational commitment (Meyer & Allen, 1991) and acknowledge a broader set of nonaffective attachments, such as the person-job fit, the ties represented by friends at work, the opportunity costs of leaving an employer and losing the quality of life enabled by the job.
Job embeddedness is the sum of on-and off-the-job embeddedness. Various studies have repeatedly found that an employee's off-the-job embeddedness influences employee retention (Jiang et al., 2012;Zhang et al., 2012). This finding, however, seems to be influenced by ethnicity and geography (Tanova & Holtom, 2008). Mallol, Holtom and Lee (2007) and Ramesh and Gelfand (2010) have demonstrated that offthe-job embeddedness is a more important factor in determining leaving intention for Indian and Hispanic employees compared to American employees. On-thejob embeddedness is understood as having three components (Mitchell et al., 2001). The first component relates to the degree of fit of the employee to the job and organization and to their life outside work. This fit is a broader version of person-job and person-organization fit and relates to the extent to which the employee possesses congruent skills, attitudes, approaches, and values with the organization, the job and other employees and to their life out of work (Mitchell et al., 2001).
These studies assume that increased attachment to family and community prevents an employee from leaving. Research on the effect of work-family conflict on employee embeddedness has taken several directions. Portoghese, Galletta and Bastelli (2011) found that leader-member exchange mediated workfamily conflict and off-the-job embeddedness in predicting employee satisfaction. Ng and Feldman's (2012) study examined how changes in employee engagement might lead to changed perceptions of workfamily conflict and found, contrary to expectations, that increased employee embeddedness led to increased reporting of work-family and family-work conflict. Karatepe's (2013) study of Romanian hotel managers found that Work-Family and Family-Work Conflict were both negatively related to employee job embeddedness and were mediated by employee exhaustion.
On-and off-the job embeddedness are likely to be affected in different ways and to different extents. Work-family conflict may directly impact employee onthe-job fit at the level of the job and organization. Work-family conflict may lead to reduced person-fit in several ways. First, regardless of the origins of the conflict at home (Family-Work Conflict) or in the workplace (Work-Family Conflict), the existence of this conflict may reveal − or highlight − to the employee their difficulty in meeting the requirements of their job. This awareness may manifest itself as strain, as the employee physically and emotionally experiences the conflict and seeks to reconcile the differing demands. The conflict may manifest itself behaviourally, as the employee realizes that that they are unable to resolve this conflict, leading to disengagement and a decline in organizational citizenship behaviours.
The experience of role conflict may lead to an awareness of poor person-organizational fit -with the realization that work arrangements, goals and personnel of the organization are in conflict with those that are preferred by the employee. Source Attribution Theory suggests that an employee who experiences work-family conflict, may not only dislike the conflict itself, but may attribute this resentment to its source -the organization (Grandey, Cordeiro & Crouter, 2005) -resulting in lower levels of person-organization fit. Strain-based work-family conflict may lead to reduced linkage embeddedness. Employees with time-based family and work conflict may find themselves prevented from developing connections with other employees, because of an inability to participate in unscheduled organizational activities outside ordinary work hours. Conflict over work and family responsibilities may see the employee choosing not to engage with office activities, resulting in reduced involvement with organizational activities (Ten Brummelhuis, Haar & Van Der Lippec, 2010).
Work-family conflict can potentially impact employee off-the-job embeddedness. As the demands of work and home conflict, these demands may potentially aggravate an employee's actual or perceived attachment to their social and family lives. Similarly, the conflict experienced at home because of work demands and the conflict created by home demands and manifested at work, will also impact employee perceptions of their embeddedness outside work. This can occur in several ways. The demands of work, leading to work-family conflict, may lead to time-based strain in the domestic domain and a re-evaluation of their home lives, or at least resentment to the impositions represented by their home lives, or their work lives. Work-family conflict may manifest itself in personal strain, in the form of disengagement from family activities, resentment or estrangement, as well as behavioural changes as the employee seeks to find ways of accommodating the competing demands. Hence, we hypothesize the following: Hypothesis 2 (a): On-the-job embeddedness mediates the relationship between work-family conflict and leaving intention.

Hypothesis 2 (b):
Off-the-job embeddedness mediates the relationship between work-family conflict and leaving intention.
Workplace flexibility as a moderator. Work-family conflict has a positive effect on the occupational stress and level of well-being in general (Cooper et al., 2001). Work-family conflict may affect stress-strain relationships in several ways. To begin with, some authors argue that work-family conflict can act as a moderating variable between stressors and work-family conflict. As work-family conflict puts an employee in a state of stress and inability to balance work and home life, this experience of continuous stress creates a feeling of occupational stress, so we can argue that workforce flexibility may act as a moderator between work-family conflict and turnover intention.
Workplace flexibility may play a buffering role in explaining the relationship between job embeddedness and turnover intentions. Indeed, employers seem to be motivated to offer their workforces greater temporal/spatial flexibility and work-family support in part because they believe doing so will attract and retain valuable employees (Allen et al., 2013). Workplace flexibility is considered an important organizational practice that helps employees effectively manage their work demands and family responsibilities (Allen et al., 2013). Kattenbach, Demerouti, and Nachreiner (2010) showed that flexibility at workplace allowed 167 German company employees to have more autonomy over their work schedules, thus reducing negative workfamily/family-work spillover. In a similar vein, De Sivatte and Guadamillas (2013) revealed that workplace flexibility helped to minimize negative workfamily/family-work spillover among 480 employees. By giving employees increased autonomy over when and how to carry out work, workplace flexibility provided employees with the means to manage their resources, alleviating the need to quit their jobs to protect these resources.
Workplace flexibility is an important variable which influences the relationship between the work stressors described above and work-family conflict (Swanberg et al., 2008). Flexible work arrangements help employees to attain balance between work and family roles. This leads to positive attitudes as employees feel greater autonomy in deciding about how, when, and where to perform job duties, thus enhancing their job performance. These positive attitudes counterbalance any negative behaviors and thus employees no longer find any reason to quit their organizations. In a situation where an employee has to perform parenting duties of his/her kids, job satisfaction levels will be greater if he/she is provided with flexibility in deciding working hours and work sharing arrangements with peers who are unmarried or have little to no parenting duties. This is then translated into positive feelings and intent to stay in the organization instead of thinking about quitting the organization. Hence, we argue that an increase in workplace flexibility results in lower work-family conflict, in turn resulting in a reduction in employee turnover intentions. On the basis of the above arguments, we hypothesize that: December 2017 | Vol. 21 | No. 2

Hypothesis 3(a): Workforce flexibility moderates the relationship between on-the-job embeddedness and turnover intention.
Hypothesis 3(b): Workforce flexibility moderates the relationship between off-the-job embeddedness and turnover intention.

Methods
Participants. This research studied the relationships between workforce flexibility, job embeddedness, workfamily conflict and turnover intention. We selected two private and two public sector hospitals in Islamabad. Our respondents were the nursing staff in these hospitals. Following Anseel, Lievens, Schollaert, and Choragwicka, (2010)'s finding that prior notice of a survey can increase survey response rates, an email introducing the project and the imminent survey was sent to potential respondents several days prior to the survey being distributed. The advance notice survey was drafted and signed by the research team, had a university logo and was distributed by a member of the organization's HR unit. No incentives were offered for survey participation. An invitation to participate with a link to the online survey was emailed to potential respondents several days later. Potential respondents were assured of the anonymity and confidentiality of responses and that the research project had been approved by a university ethics committee. Two followup reminders were distributed after each survey to remind employees to participate.
The online survey was completed by 187 nurses, representing a response rate of 22.3%. Following Blair and Zinkhan (2006)'s method for identifying nonresponse bias, several demographic characteristics (age, gender) of the sample were compared against the characteristics of the workforce. The average age of the sample was 36.8, equal to the workforce average. Although 59% of the sample were female, Analysis of Variance testing found that male and female respondents did not differ on several variables (such as work-family conflict or leaving intention) used in this study. Accordingly, we have no evidence of non-response bias.
Measures. All items were measured using 7-point Likert scales, ranging from 1 (Strongly disagree) to 7 (Strongly agree). Cronbach alpha scores for each measure are reported in Table 1, along with descriptive statistics and correlations.
Work-family conflict. In this study, work-family conflict is measured as the un-weighted average of all ten items of Netemeyer et al.,'s (1996) Family-Work and Work-Family Conflict sub-scales. A sample item is "The amount of time your job takes up makes it difficult to fulfill your family or personal responsibilities". The Cronbach's alpha of the scale was 0.83. Turnover intention. Turnover intention was measured using Abrams, Ando, and Hinkle's (1998) three item scale. The unweighted mean of the three items represents the turnover intention score. A sample item is "Do you intend to leave the organization in the next 12 months?". The Cronbach's alpha of the scale was 0.88.
Workplace flexibility. We measured workplace flexibility with four items derived from Ten Brink's validated questionnaire (2004), which is based on the Workplace Flexibility scale (Meyer, 1997). A sample item is "My organization allows me to change set working hours (including changing shift pattern)".The Cronbach's alpha of the scale was 0.92. Off-the-job embeddedness. This measure was made up of the unweighted mean of the z-scores of the remaining twelve items in Felps et al., (2009) job embeddedness scale. Following the method used by JET researchers zscores were taken because several of the items were dichotomous (Mitchell et al., 2001). A sample item is "The area where I live offers the leisure activities that I like". The Cronbach's alpha of the scale was 0.87.

On-the-Job
As the data for this study came from a single source (i.e., employees), there was a potential threat of common method bias. To address this issue in more depth, we conducted Harman's one-factor test on all the items (Podsakoff et al., 2003). We extracted five distinct factors accounting for more than sixty seven percent of the total variance, with the first factor explaining about twenty two percent. No single factor therefore emerged, nor did one factor account for most of the variance. These two conditions reduced the likelihood of common method variance being a serious problem in this study.

Results
Off-the-job embeddedness. This measure was made up of the unweighted mean of the z-scores of the remaining twelve items in Felps et al., (2009) job embeddedness scale. Following the method used by JET researchers zscores were taken because several of the items were dichotomous (Mitchell et al., 2001). A sample item is "The area where I live offers the leisure activities that I like". The Cronbach's alpha of the scale was 0.87. Note: ∆χ 2 presents differences between model and the following model. Fit indices criteria refers to Hair et al., (2006). ** p <0.001

. Structural Equation Modeling
As the data for this study came from a single source (i.e., employees), there was a potential threat of common method bias. To address this issue in more depth, we conducted Harman's one-factor test on all the items (Podsakoff et al., 2003). We extracted five distinct factors accounting for more than sixty seven percent of the total variance, with the first factor explaining about twenty two percent. No single factor therefore emerged, nor did one factor account for most of the variance. These two conditions reduced the likelihood of common method variance being a serious problem in this study.
To test Hypothesis 3a, we entered the main effect of onthe-job embeddedness as well as the interaction effects (on-the-job embeddedness * workforce flexibility) on leaving intention. If the interaction paths are significant, moderator hypothesis is supported. A significant beta coefficient for each interaction term (on-the-job embeddedness * workforce flexibility) or values of the incremental F-statistic indicate that the moderator variable acts as a moderator. We first entered control variables and then the main effects of on-the-job embeddedness and workforce flexibility were entered along with the control variables. In the last step, the interaction effects of on-the-job embeddedness with workforce flexibility were entered along with the control variables.
When the control variables were entered, job tenure and job position were related positively and significantly to leaving intention (t = 3.91, p < 0.01 and t = 2.80, p < 0.05), respectively.
Step 2 provided a significant increase in the variance explained in step 1 (ΔR² = 0.18; ΔF = 12.23, p < 0 .01) for leaving intention. Workforce flexibility was positively and significantly related to leaving intention (t = 2.40, p < 0.01). Hypothesis 3a would be supported if the interaction terms accounted for a significant incremental variance in explaining leaving intention either individually, manifested by beta values, or collectively, revealed by the values of the incremental F-statistic.
Step 3 provided a significant increase in the variance explained over step 2 (ΔR² = 0.23; ΔF = 15.95, p < 0.01). Hence, Hypothesis 3a was supported. Similarly, Hypothesis 3b would be supported if the interaction terms (off-the-job embeddedness and workforce flexibility) accounted for a significant incremental variance in explaining leaving intention either individually, manifested by beta values, or collectively, revealed by the values of the incremental F-statistic.

Discussion
The purpose of this paper was to examine the impact of work-family conflict on employee leaving intention through the perspective of the Job Embeddedness Theory and workforce flexibility. In literature dominated by analysis of stress and employee affective response, the JET framework enables the recognition of non-affective influences on the turnover process (Karatepe, 2013;Lang, Kern, & Zapf, 2016;Peltokorpi, Allen, & Froese, 2015). The model also showed workfamily conflict positively influenced turnover intention, which had also been mentioned in other studies (Blomme et al., 2010;Chiang & Chang, 2012;Kim & Kao, 2014). Sperlich and Geyer (2015) revealed that women who suffered from higher work-family conflict showed higher family-related stress. Work-family conflict can be regarded as a stressor. Work characteristics, such as night shifts, minimal control over work hours and unpredictable scheduling requirements are important stressors which are positively related to work-family conflict (Blomme et al., 2010;Cooklin et al., 2015). These work characteristics are applicable to nurses. Hence, there is a positive relationship between work-family conflict and turnover intention with job embeddedness as a mediator. Some studies (Han, Han, An, & Lim, 2015;Thakur et al., 2017;Treuren, 2017) mentioned that work-family conflict had indirect effects on turnover intention, but had no direct impact on turnover intention. As the results showed, the indirect effect is much greater than the direct effect, so the small samples of other studies did not have enough power to detect the direct effect.
This paper examined how work-family conflict may impact employee turnover intention. It recognizes that conflict may affect how people relate to their work. This effect may have the effect of weakening their attachment to their job and organization, not only in terms of how they feel about their work, but in terms of their actual involvement with the organization. Workfamily conflict, an earlier section of the paper proposed, may erode an employee's attachment to the organization. It is this embeddedness that affects an employee's turnover intention. This paper sketches several ways in which work-family conflict acts to influence an employee's embeddedness.
Work-family conflict was understood as having consequences on an employee's satisfaction with their work -their satisfaction with their job and the organization, and on an employee's affective commitment -their willingness and enthusiasm to support the organization (Lu et al., 2017;Nohe, Meier, Sonntag, & Michel, 2015). Dissatisfaction and disengagement activates the process of employee withdrawal, manifesting as turnover intention and culminating in the decision to leave. The availability of alternative opportunities is also necessary. Work-family conflict taken together, employee embeddedness partially mediates the effect of work-family conflict on employee turnover intention. Interestingly, of the two components of job embeddedness examined, only onthe-job embeddedness had a statistically significant relationship with the independent and dependent variables. For these employees, about three-quarters of the effect of work-family conflict directly impacted employee turnover intention, with the remaining effect mediated through employee embeddedness.
Practical implications. This study has important implications for hospital management. Nurses can be retained in hospitals if management establishes a working environment where family and work lives are balanced. Understanding how work-family conflict influences turnover intention is increasingly important to employers facing growing labour, skill and talent shortage and growing pressure on organizations by increased competition in the product and equity markets. If nurses are provided with flexibility in deciding their working hours, shifting duties among each other, and a relaxing environment where they can share their burdens of work with others, chances of leaving the organization minimize. This paper suggests that work-family conflict directly affects an employee's attachment to the organization. As noted earlier, in addition to the consequences on employee turnover, a decline in employee embeddedness is likely to lead to a decline in employee discretionary effort. This is likely to assist employers in working out the cost consequences of better managing work-family conflict and to reduce the gap between management and December 2017 | Vol. 21 | No. 2 employee perceptions of the importance of work-family arrangements.
Limitations and further research. Although this paper points to some new areas of investigation, this research has several limitations. Like much research in the social sciences and in work-family conflict studies, it relies on cross-sectional data to make inferences about the dynamics of employee behaviour. In this case, an employee's work-family conflict at a point in time is compared with their job embeddedness and turnover intention at that same point in time, in order to make inferences about how work-family conflict impacts employee decision-making over time. The second limitation relates to the self-reported nature of the data. Subsequent research could avoid this problem by collecting the independent and dependent variables separately, at different points in time, or obtaining data from another source -such as a partner, manager or coworker.

Conclusion
This study examined the mediating effect of job embeddedness and moderating effect of workplace flexibility on the relationship between work-family conflict and turnover intentions of nurses. This study has made important contributions to existing knowledge. First, work-family conflict literature has emphasized the affective consequences of work-family conflict on employee turnover intention. Karatepe and Karadas (2014) tested the mediating effect of psychological capital and proposed that future research should take into account additional variables, both at individual and organizational level, to further explain the indirect effect of work-family conflict on turnover intentions. Therefore, this study examined the mediating effect of job embeddedness to explain the effect of work-family conflict on turnover intentions. Second, although the JET has prompted many studies, the majority of this literature has focused on the outcomes of employee job embeddedness (DiRenzo et al., 2017;Peltokorpi et al., 2017). Although employee embeddedness has been demonstrated to be a reliable predictor of employee outcomes such as employee performance and turnover, much less is understood about how embeddedness forms and evolves. This paper extends existing literature by examining how a particular antecedent − work-family conflictinfluences the formation and evolution of employee job embeddedness and demonstrating how work-family conflict is an antecedent of aspects of employee job embeddedness. Third, despite the predictive powers of workplace flexibility, little is understood about its role as a moderator (Rudolph & Baltes, 2017). In this study, we examined whether workplace flexibility moderates the effect of work-family conflict on employee turnover intention.