Work-Family Stressors and Work-Family Satisfaction: Effect of Sense of Coherence at a Metropolitan Municipality

This article contributes to emergent research by examining the linkage between work–family stressors and their effect on work-family satisfaction (WFS) at a metropolitan municipality in South Africa. Extant literature on stress has either not adequately examined the linkage between domain specific stressors and domain specific satisfaction or have suggested models with direct connections. The present study suggests a mediating model and assesses the mediation. Specifically, it claims that sense of coherence (SOC) plays a mediating part in the in the work and family stressors - WFS relationship. This mixed methods research applied a variance-based structural equation modelling (Partial Least Squares) to a sample of 307 professional level employees at a metropolitan municipality in South Africa. The finding supports the importance of SOC and its influence on WFS. Additionally, mediation hypotheses theorise how SOC plays a critical mediating influence in the work-family stressor-WFS relationship. Data analysis suggest that (a) work stressors and WFS interrelated in a manner that SOC fully mediated the effect of work stressors on WFS (b) SOC partially mediated the relationship between family stressors and WFS. The findings have both theoretical and practical implications.


Introduction
Individuals encounter stress in their everyday life at work and home. Because of this, contemporary workplaces have designed and adopted benefitting practices that assist employees with the management of their work and family stressors towards achieving satisfaction and improved performance (Michel, 2015). Ko and Hur (2014) suggest that there is a need to understand the effects of these benefits on employee outcomes at work and home. Most scholars of these practices examine their outcome on performance and behaviour (Kossek, Baltes & Matthews, 2011).However, in addition to workplace benefits, employees also use their personal resources (Hobfoll, 2001)to address stress in their environment in order to achieve satisfaction with work and family situations. In their study, Caprara, Barbaranelli, Steca and Malone (2006) found that a personal resource (self-efficacy) significantly contributed to satisfaction with work among teachers. Satisfaction with work is simply the perception of contentment that individuals derive from their daily work activities (Judge, Thoresen, Bono, & Patton, 2001). Ordinarily, it is not expected that an individual burdened with work and family stressors should derive satisfaction from such situations. Interestingly, scholars have reported that irrespective of high work stressors, individuals find satisfaction with work situations (Klassen & Chiu, 2010). This implies that intervening factors could be responsible for the derivation of such satisfaction in the face of stressors. The present study examined SOC as one of those intervening factors.
To appreciate the influence of work-family stressors on the achievement of work-family satisfaction among employees at a metropolitan municipality, it is vital to consider the mediating role of SOC. Scholars have recommended SOC as an individual's personal resource for coping with stress (Feldt, Leskinen, Kinnunen & Ruoppila, 2003;Nielsen & Hansson, 2007). Literature reveals that individuals with strong SOC cope more adequately with stress and achieve more satisfaction with work and family situations than those with weak SOC (Moksnes & Haugan, 2015;Griffiths, Ryan & Foster, 2011;Antonovsky, 1987). However, to the knowledge of these scholars, no available published studies were discovered that investigated the role of SOC in the linkage between work-family stressors and work-family satisfaction among employees of a metropolitan municipality. Therefore, the purpose of this empirical analysis was to explore the ways in which SOC directly or indirectly influenced the achievement of work-family satisfaction among professional level employees at a metropolitan municipality in South Africa.
Job stress are those work-related factors in the work environment that adversely affect individual employees (Beehr, 1976;Kahn, Wolfe, Quinn, Snoek & Rosenthal, 1964). This article appraised job stress as those situations that place an individual employee in a position where s/he is ill disposed to perform his/her function maximally because of strain resulting from the job and/or work environment. Job stress could associate negatively with SOC and lead to dissatisfaction at work. Persistent job stress could negatively affect strong SOC on the long run. Empirical studies report that job stress affects satisfaction with work situations adversely (Khamisa, Oldenburg, Peltzer, & Ilic, 2015;Hayes, Douglas, & Bonner, 2015).Job autonomy refers to the level of self-determination, liberty, discretion intrinsic to a job offered an individual in arranging work, and deciding the processes to be engaged in performing the job (Saragih, 2015;Tufail et al., 2016;Parker, Axtell, & Turner, 2001). Accordingly, job autonomy results in the essential mental state of "experienced responsibility for outcomes of the work", that conversely results in increased efficiency at work and increased work enthusiasm. Conversely, absence of autonomy is the lack of discretion to make important decisions concerning one's job (Kim & Stoner, 2008;Liu, Spector & Jex, 2005). When an employee feels that his/her input into what he does is not contributory to the organisational achievements, he/she may be dissatisfied with work situation. Absence of autonomy could have a significant negative impact on the level of an individual's SOC. COR claims that employees would preserve their personal resources from stressors that drain those resources rather than gain another resource for example satisfaction. In this context, the authors presume that absence of autonomy will negatively influence SOC and lead to reduction or dissatisfaction with the work situation. Liu et al. (2005) report that felt autonomy associated positively with satisfaction (at personal and organisational levels). Other scholars found that absence of job autonomy associated negatively with satisfaction with work (Wu, Luksyte & Parker, 2015).
Function vagueness/role conflict result from the diverse and clashing sources of governmental influences in the Public Service (Pandey & Wright, 2006). Features of such organisation, for example, centralised decisionmaking hierarchies (Nicholson & Goh, 1983), and formalities (Ramaswami, Agarwal & Bhargava, 1993), are reported to heighten function vagueness. Hamner and Tosi (1974) report that function vagueness is an outcome of the absence of necessary information concerning what an individual's task is. Scholars have also attributed dissatisfaction with work situations to function vagueness and role conflict (Hoffman, Hoffman, Kelley, Kelley, Rotalsky & Rotalsky, 2016). This article assumes that work stressors are capable of adversely affecting an employee's SOC and their achievement of WFS. Additionally, individuals with strong SOC would use it to address work stressors but in the long run, this could deplete their SOC and impair the achievement of WFS. Accordingly, H1: An inverse but significant association exists between work stressors and SOC Family Stressors: Several issues resulting from the interaction of family members and the family environment confront and task personal resources of family members. Family stress is characterised by multifaceted needs for modification or adaptation in behaviour. Generally, families take actions against particular stressors and resultant challenges from these stressors (McCubbin & Sussman, 2014). Aseltine Jr, Gore, and Gordon (2000) confirm that family stressors could have negative influences on a person's internal consistency. Hence, there is a need to identify and examine these stressors and their effects, as well as coping factors such as SOC and their influence in buffering or eliminating the adverse stressors. Contemporary workplaces have numerous dual-earner families that are playing multiple roles and struggling to find satisfaction with work and family situations (Lin, Chen & Li, 2016). According to Hepburn and Barling (1996) in Jones, Burke and Westman(2013), responsibility for providing care for elders could upset work schedules, lead to fractional absenteeism and cause disruptions at work.
Depending on the type of family, most stressors emanate from the challenges of parenting and elder care (Keating & Eales, 2017); relational tension (Clout & Brown, 2016) and absence of spousal support (Barger & Cribbet, 2016). These difficulties tax individual abilities to cope with work and family duties. The resultant stressor from giving more attention to work is absence of spousal support, which could be challenging to an individual's coping abilities. For example, Saleem (2015) reported that due to the high level of competition in the workplace, most employees give more time and attention to their work than they give to their families. This study examined parenting, elder care, relational tension and absence of spousal support as family stressors to ascertain the influence on an individual's SOC and the achievement of WFS. This article suggests that family stressors could negatively influence an individual's WFS, but SOC could mediate the effect of family stressors on WFS. In accordance with these ideas, it is postulated that; H2: A significant but inverse relationship exists between Family stressors and WFS.
The mediating role of Sense of Coherence (SOC): Although there are numerous factors affecting individuals at work and family domains respectively, the influence of stress and need to manage it has been attracting attention in contemporary times. This article examined work and family stressors among professional level employees at a metropolitan municipality in South Africa. Scholars have investigated the mediating role of psychological traits such as locus of control and self-efficacy (Prati, Pietrantoni, & Cicognani, 2011), positive and negative affect (Green, DeCourville & Sadava, 2012), and mindfulness (Black, Sussman, Johnson & Milam, 2012), relevant to stress and other outcome variables. This article sought to demonstrate that SOC occupies a mediating role between work and family stressors and WFS. According to Yamazaki (2008), SOC is generally conceptualised as the capacity to manage individual stress. Individuals with strong SOC demonstrate positive disposition in the presence of stress and accept stressful conditions as challenges that need to be addressed. SOC is more perceptible when stress increases in an environment (Takeuchi & Yamazaki, 2010) and as a stress-cushioning factor; it kicks into action with the introduction of a stressor (Yamazaki, 2008). Scholars report that SOC had an inverse relationship with stress and that continued exposure to stressors causes SOC to deteriorate (Moksnes, Rannestad, Byrne & Espnes, 2011;Wolff & Ratner, 1999;Ryland & Greenfeld, 1991). Extant literature further reports that SOC acts a buffer from adverse work stressors (Feldt, 1997;Urakawa, Yokoyama & Itoh, 2012). Scholars report that SOC mediated the effect of religiousness on satisfaction with life (Cowlishaw, Niele, Teshuva, Browning & Kendig, 2013). Kinman (2008), in a study among academics in the UK, found that statistically significant associations were found among SOC, job stressors and health outcomes. Antonovsky (1987) claimed that positive work experiences could improve an employee's SOC (Antonovsky, 1987). This implies that organisations can make the employees work experiences better for them in order to harness the benefits of SOC. However, Mayer and Boness (2011) report that little or no suggestions about particular interventions to stimulate SOC have been made. Although much work has been done to ascertain the role of SOC as a mediator between stress and other outcome variables, few research has been done to evaluate mediation of work-family stress and WFS. This study examined SOC to ascertain its mediating role in the interface between work and family stressors and workfamily satisfaction (WFS). In accordance with the scholarly ideas, H3a: SOC mediates the connection between work stressors and WFS at a metropolitan municipality in South Africa H3b: SOC mediates the connection between family stressors and WFS at a metropolitan municipality in South Africa

Sense of Coherence (SOC) and Work-family satisfaction (WFS):
Work-family satisfaction is one of the endogenous (outcome) variables examined in this article. This article proposed that the relationship between work and family stressors and WFS is adversarial. Where the levels of work-family stressors are high, individuals will not achieve WFS. Additionally, where personal resources and organisational benefits available to the individual are not adequate (Lazarus, 1991), WFS will be low. The COR model (Hobfoll, 1989;Grandey & Cropanzano, 1999) and transactional theory (Lazarus, 1991) underpin the theoretical concepts of assumption. COR model theorises that people endeavour to gain and preserve resources such as things for example, (food), personal resources (SOC, self-esteem, self-efficacy), and energy (e.g. time) (Goh, Ilies, & Wilson, 2015). A loss of or threat to these resources is significant to the individual (Hobfoll, 1989). Where individuals with high SOC for example, are exposed to continued work and family stressors, this resource stands the risk of being drained over time (Moksnes et al., 2011;Wolff & Ratner, 1999). In this context, this study posits that the work and family stressors examined could deplete an individual's personal resources (SOC) through energy drain and time consumption. Increase in work and family stressors transmutes into the need for extra resources to address them. However, resources are scarce and limited, hence much of them are not available to satisfy work and family demands (Ilies et al., 2007;Edwards & Rothbard, 2000). For example, the need to work for longer hours could result in the problem of spouses spending fewer hours at home. This could leave lesser time to fulfil home demands (Thompson, Beauvais, & Lyness, 1999) or play the role of a supportive spouse. Conflict created because of this situation between the work and family spheres could lead to the draining of the individual's personal resources (e.g. SOC) and lead to dissatisfaction with work and family situations. Inversely, SOC could buffer these stressors and lead to the achievement of WFS Based on these claims, H4: A significant positive relationship exists between SOC and WFS Conceptual Framework guiding the study: As presented in Figure 1, work and family stressors are conceptualised as the independent variables influencing SOC and WFS at a metropolitan municipality in South Africa.  The study is a sequential transformative mixed method research undertaken among professional level employees in a metropolitan municipality in South Africa. Data was collected in two ways; (a) concurrently with a questionnaire that contained both open-ended and close-ended questions respectively and (b) sequentially by interviewing 11 members of top management at the metropolitan municipality. Priority was given equally to both the quantitative and qualitative data collected from the survey and interview (Creswell & Garrett, 2008;Hanson, Creswell, Clark, Petska, & Creswell, 2005;Morgan, 2007). Outcomes of quantitative data needed to be refuted, authenticated or supported by the qualitative data outcomes; hence, the equity in prioritising both data sources (i.e. quantitative and qualitative). Summarily, this paper presents the information gathered and appraised based on Hanson et al. (2005) sequential transformative mixed methods.

Sample recruitment and questionnaire administration:
Participants were informed about the study through the municipal in-house mail and online information system. Thereafter, the researchers personally administered self-response questionnaires to the participants. All the questionnaires were written in the English language. Only employees at the professional level cadre who were born before January 1994 were permitted to participate in the study. The study targeted a population of7,000 employees and a sample of 370 professional level employees were drawn conveniently from the HR, Governance, Finance and Procurement, Safety and Social Services, Infrastructure, and Human Settlement clusters of the examined Metropolitan municipality (Krejcie & Morgan, 1970). Data was collected over a six-month period; response rate was 84%. The ethical committee of the University of KwaZulu-Natal (UKZN) reviewed and approved the questionnaires. All the scales used in this study to analyse SOC, work-family stressors and WFS were those whose validity and reliability and reliability had been proven in other studies. However, their reliability and reliability were also validated in the present study. Structural equation modelling using the SmartPLS was engaged in the analysis of data. The measures used are explained in broader terms below.
Measures: Standardised questionnaires were used in collecting all the items. The scholars relied on existing and validated measurement scales from existing literature where possible. For example, in measuringSOC: Antonovsky's SOC-13 version of Orientation to Life Questionnaire Antonovsky (1987) to assess SOC. The scale comprises of four-items measuring meaningfulness, five items evaluating comprehensibility and four other items assessing manageability (Feldt et al., 2003) and one evaluating controlwas added. Responses were based on a 7-point semantic differential scale with twofold anchoring sayings. Responses to items measuring comprehensibility ranked from 1) very often, to 7) very seldom or never; manageability ranked from 1) never happened to 7) always happened; and meaningfulness ranked from 1) very seldom or never, to 7) very often (Feldt et al., 2003). Past studies have generally reported the validity and reliability of the OLQ scale with primary evidence validating the significance of the core concept (Antonovsky, 1993). One hundred and twenty-seven (127) studies reported that the SOC-13 Cronbach's alpha range from 0.70 to 0.92 (Eriksson & Lindström, 2007). This study reported an acceptable Cronbach alpha reported of 0.78 Work Stressors: A 29-itemWork Stressor Scale measuring job stress, absence of autonomy and function vagueness was used in assessing work-stressors. Job stress was evaluated on two items, while Spector and Fox's (2003) Factual Autonomy Scale was used in assessing absence of autonomy in order to ascertain the predictability of perception by the individual about limitations imposed on their ability to perform tasks without supervision and the result of other people's decisions on their job. However, a 17-item scale adopted fromRizzo, House and Lirtzman's (1970) 30-item scale was used in evaluating function vagueness/role conflict. This assessed how employees perceived their jobs, supervision, interpersonal relationship at work, time needed to complete a task and guidelines/ policies' clarity. The Cronbach alpha value reported in this study for this scale is 0.87 Family Stressors: Parental workload/relationship tension, absence of spousal support, and elder care were evaluated with a 13-item Family Stressor Scale. The items were adopted from Barling, MacEwen, Kelloway and Higginbottom (1994) and Brestan, Eyberg, Algina, Johnson, and Boggs (2003). Response was on a fivepoint frequency based answering scale. The Cronbach alpha reported in this study is 0.89WFS: was assessed on a two-item scale measuring satisfaction with work and family situations. The items the saliency of the satisfaction that respondents derived from their work and family situations respectively. "The major satisfaction in my life comes from my job "is one of the items from this scale. The Cronbach alpha value in this sample for the family stressor scale is an acceptable 0.70 Data Analysis: Data was initially captured and analysed with the IBM SPSS version 22 software; there were no missing data. Thereafter, the tested data were saved as Comma Delimited (*CSV) file for ease of importation into SmartPLS 3.3 software. Partial Least Squares (PLS) permits the simultaneous evaluation of constructs validity, reliability, and appraisal of the associations among variables (Barroso, Cepeda & Rolda'n, 2010). Calculation of the path model and estimation of parameter was done with SmartPLS 3.0 based on the 'path weighting scheme' (Henseler, Ringle, & Sarstedt, 2012). The latest guidelines for PLS-SEM were followed in the assessment and measurement models prior to the evaluation of the structural model (Hair, Hult, Ringle, & Sarstedt, 2013a). However, qualitative data was organised for ease of content analysis through the NVIVO software.

Findings
Quantitative data: Below is a presentation of the findings of the quantitative data collected for this article.

Table 1: Quantitative Analysis
Measurement models: The indicators were found to be at acceptable reliability levels (Klarner, Sarstedt, Hoeck, & Ringle, 2013). To doubly ascertain that the internal consistency of the major indicators were reliable, the Cronbach alpha and composite reliability coefficients were tested (Hair, Sarstedt, Hopskins, & Kuppelwieser, 2014). Table 1 shows that the composite reliability of the measurement models exceeded 0.812, while Cronbach alpha exceeded 0.700. These are clear indications that the variable measures' internal consistency are reliable. Similarly, all average variance extracted (AVE) values exceed the critical base value of 0.50, supporting the convergent validity of the measures. Furthermore, the Fornell and Larcker (1981) principle that requires the square root of the AVE to be of greater value than any of those inter-factor correlations was applied. The test provided evidence that the variables exhibited discriminant validity (Campbell & Fiske, 1959). Table 2 and Figure 2 below, present the results of the PLS-SEM analysis. They illustrate findings of the structural model estimation and assessment of the associations between work and family stressors, SOC and WFS (Hypotheses H1 to H4).To test for mediation based on the SEM, where zero is absent in the interval for a mediation hypothesis, it implies that the indirect effect is significantly different from zero with 95% confidence (Castro & Roldan, 2013).From Table 2, the p-values indicated that a significant relationship exists among the variables at (p < 0 and p < 0.05). However, the linkage between work-family stressors and WFS are not significant. The core criterion used in assessing structural model (Henseler et al, 2012), i.e. the coefficient of determination R-square is 0.098 (see Fig. 2) for the principal target construct (WFS) in this study. This Rsquare, which is greater than zero, confirms the predictive validity of the model (Hair, Sarstedt, Ringle & Mena, 2012b). The recorded R-square for mediating construct (SOC) is 0.117. Summarily, the predictive validity of WFS is 9.8% and SOC is 11.7% (Hair, et al., 2012a). The indirect path WS (work stressors) to SOC (sense of coherence) indicated a negative and statistically significant beta loading of (r = -0.295, p < 0.001), while the direct path (WS -WFS) is not significant at (r = -0.021, p > 0.05). Therefore, the finding that an inverse but statistically significant association exists between WS and SOC supports H1. Furthermore, the direct path FS to WFS (r = 0.142, p, 0.05) and indirect path FS through SOC (r = -0.120, p < 0.05) have inverse but statistically significant associations. This result supports H2 by showing that family stressors significantly influence WFS and SOC negatively. The direct path (FS -WFS) showed a higher significant but inverse impact (r = -0.142, p < 0.05) than the indirect path (FS -SOC) at (r = -0.120, p < 0.05). In summary, the inverse relationships indicate that WS and FS confronting employees at the metropolitan municipality were adversely influential on employee SOC. Hence, SOC buffered 11.7% of work and family stressors among employees at the metropolitan municipality, supporting extant literature.

Table 2: Path coefficient Mean, STDEV, T-Values, P-Values
The path coefficient from SOC to WFS (work-family satisfaction) resulted in (r = 0.245, p < 0.001), indicating that SOC has a positive and highly significant effect on WFS among employees in the metropolitan municipality investigated and supports H4. The findings show that SOC fully mediated the connection between WS and WFS at (p = 0.000) because, the direct link which was not significant previously (r = -0.021, p > 0 .05), became significant through the indirect route (r =-0.295, p < 0.01). This implies that work stressors exerted a significant but indirect influence on WFS through SOC. This finding as explained in Figure 2 and Table 1 support hypothesis H3a of this study. Therefore, SOC was expedient in buffering work stressors among respondents and instigating WFS. However, the path analysis in Figure 2 showed that SOC partially mediated the linkage between family stressors and WFS. The association between FS and WFS remains significant (r = -0.120, p < 0.05), however with a difference. The difference of 0. 022, is lesser than when SOC was not present (r = 0.142, p < 0.05). Although the variance accounted for (VAF) in this study is 0. 022, Helm, Eggert, and Garnefeld (2010) suggest that partial mediations between 0% and 100% is normal, but the higher the value, the stronger the partial mediation. In this study, the partial mediation has moderate VAF of 22%. Hypothesis H3b is partially supported in this study. This means that the effect of family stressors on WFS is reduced in size but not completely buffered when SOC is controlled at the metropolitan municipality in South Africa.  Respondents agreed that they were facing numerous and varied work stressors but suggested that they could manage some of these stressors. The responses of some of the participants buttress this claim as follows, "Naturally, local government is a bureaucratic organisation and ambiguity/role conflict and absence of autonomy although expected, cause stress to employees. We know and struggle with these things, optimistically we manage" (Interviewee 7). "Nothing much stresses me at home. I have a great child and supportive partner" (Respondent 058).

Original
Employees at the municipality investigated generally accept that though they were faced with stressors at work, they have a disposition that suggests that they were sentimentally optimistic about their ability to cope with support from their families. This clarifies why SOC fully mediated work stressors and WFS and partially mediated family stressors and WFS at the metropolitan municipality investigated. Several respondents accepted some of the work stressors as normal but. Others believed that through their faith, stress was manageable. Take for example, this response "I am capable of coping with work and life stressors. I have a strong spiritual foundation which helps me cope with the trials and tribulations of life." (Respondent 071). Extant literature supports this finding that SOC positively associates with religious faith (Cowlishaw et al., 2013). Work stressors: Fig. 3presents a number of work stressors suggested by the respondents as those taxing them at work.

Figure 3: Notable work stressors among respondents
Although these stressors are numerous, the quantitative data revealed that SOC fully mediated the linkage between work stressors and WFS. This implies that if the management of the municipality invests in the adoption of interventions that build up SOC as suggested by scholars (Mayer & Boness, 2011), the effect of work stressors on employee WFS will be adequately mitigated while boosting employee satisfaction with work and family situations.
Family stressors: Respondents in this study listed fewer family stressors than work stressors challenging the achievement of WFS in the metropolitan municipality investigated. However, a particular respondent's comments summarised the overall feeling of most of the respondents.
Interviewee 3: "It is mainly family responsibilities which affect work, either extended family or your own family. In most cases, almost all workers even my colleagues would have people that they need to take care of and that becomes a problem. You will either be irresponsible which you don't want to be because this is Africa (our culture demands that you belong to a family and the community) and no matter what you do you still want to belong and take care of your family and extended family responsibility. In addition, that becomes a conflict, which in most cases becomes irreconcilable, because the individual employee has to live a certain level of standard of living. You are responsible to the family and needed there too while your work demands your presence also. This crumbles so many of them and of course affects their coping mechanism. Numerous people are facing divorce, and others spend time at work to escape problems at home. Many of the women are single mothers and seem to have challenges with family and work involvement. They marry today and in three months' to one year's time, the marriage fails and they are divorced! In addition, in most cases the divorce is very bitter and messy. Some leave with little children that require attention and care. Most people are lucky to leave such marriages without 'baggage' but they (divorce) are always messy, very messy. Then of course, that causes a lot of stress to the employee and affects work. We expect something to be done to alleviate these challenges" This could explain why SOC could not fully mediate family stressor -WFS linkage. Most respondents were of the opinion that family stressors were what they had to live with. Although these stressors have outcomes that may be adversarial, respondents would rather address work stressors to ensure that their earnings are secured rather than attend to family stressors. Hence, to these employees, family stressors do not influence their satisfaction with work situation and mean little or nothing in their appraisal of satisfaction. Therefore, their indifference to family stressors may account to the reason why SOC partially mediated the linkage between family stressors and WFS. Fig. 4 presents the family stressors found among professional level employees at the metropolitan municipal workplace examined. Theoretical implications: A statistically non-significant association was found among work stressors and WFS; this is supported by extant literature (Hoffman et al, 2016;Khamisa et al, 2015;Hayes et al, 2015;Wu et al, 2015). From this finding, it implies that strengthening individual employee SOC is more needful than merely establishing the achievement of WFS, since SOC fully mediated work-stressor and satisfaction. In this study, work stressor is the construct that is more negatively associated with WFS in line with extant literature (Ryland & Greenfeld, 1991;Wolff & Ratner, 1999;Kinman, 2008). However, transactional theory (Lazarus, 1991) and COR theory (Hobfoll, 1989(Hobfoll, , 2001(Hobfoll, , 2011a(Hobfoll, , 2011b give supportive explanations to the outcome of the work-family stressor -SOC-WFS connexion. According to the COR theory (Hobfoll, 2011a), organisational and personal resources are required to address work and family stressors deemed to be a threat to the individual. SOC as a personal resource assembles resources from within the individual and/or organisational environment to address stressors. Additionally, SOC works in conserving some of the resources for the future (Antonovsky, 1993). Therefore, the highly significant (p = 0.000) mediation of SOC among work stressors and WFS indicates that where employees have strong SOC, the effect of work stressors on WFS will be fully buffered. The inverse relationship among the variables (work and family stressors and SOC; and work and family stressors and WFS) validates the expectations of this article and is substantiated by extant literature (Moksnes et al., 2011), that stressors strongly and negatively impact an individual's SOC. In this study, SOC is also authenticated as a buffering factor (Antonovsky, 1993;Feldt, 1997;Urakawa et al., 2012) in the positively significant path from SOC to WFS. Astonishingly, the family stressors -WFS link was not fully mediated by SOC. SOC partially mediated since the effect of family stressors on WFS was barely reduced in size, family stressors were not completely eliminated when SOC is controlled. Another aspect is that, although work and family stressors jointly predicted SOC work stressors contributed more to the model. The qualitative data clarified this standpoint by highlighting the numerous work stressors confronting the employees at a metropolitan municipality in South Africa. The implication of this on SOC is that, it (SOC) will consistently deplete because of the continued exposure of employees to numerous work stressors (Moksnes et al., 2011;Wolff & Ratner, 1999). This study achieved its objective because all the hypotheses tested were supported in the study.

Implication for practice:
The growing importance of stress in the global work environment highlights the question of how organizations should support employees in managing their work and family stressors towards achieving WFS (Kossek et al., 2011). On these lines, Falk and Miller (1992) suggests organisations should appraise the coping means available to an individual in line with stressors in making decisions that should reduce the negative results of these stressors. A good understanding of SOC (Griffiths et al., 2011) and work place benefits (Ko & Hur, 2014) is needed to answer this question. In this empirical analysis, the findings show that individual employee SOC is important in managing work and family stressors and achieving WFS. Therefore, the management of the metropolitan municipality examined should take cognisance of the way in which the constructs influence each other in order to design and adopt workplace practices affecting each construct. This requires the adoption of a holistic method of appraising the design, adoption and implementation of workplace strategies to ensure that they integrate SOC enriching factors such as those listed by Mayer and Boness (2011). Human resource practitioners in the examined metropolitan municipality and researchers should look into ways of identifying such SOC enriching factors and tailoring them into practices and studies that assist employees in managing stress.
Qualitative data found that numerous work and family stressors ( Fig. 3 and Fig 4) confront employees at the municipality examined. Extant literature suggests that employees SOC deplete with continued exposure to myriads of stressors (Moksnes et al., 2011;Wolff &Ratner, 1999). In order to circumvent the total weakening of employee SOC and promote the achievement of WFS, management could revisit concerns around those work-stressors highlighted by this study such as bureaucracy, complex holdups and practices that hinder employee satisfaction with work. More use could be made of fundamental technological solutions that simplify work processes in order to assist employees. Adopting major family aligned strategies (Abe, Fields & Abe, 2016) such as flexible work arrangement, telecommuting, domestic relationship counselling, and employee wellness programmes may be invaluable in addressing family stressors. Management has to introduce SOC enhancing elements (Mayer & Boness, 2011;Kinman, 2008)into workplace practices that assist employees in addressing their work and family stressors.

Conclusion
This study adds value to literature by objectively investigating the mediating role of SOC in the interface between domain specific stressors and work-family satisfaction. It found that SOC is meaningful in fully mediating work stressors and partially mediating family stressors. SOC can be improved by positive experiences at work (Antonovsky, 1987), although particular SOC promoting interventions are scarcely suggested in extant literature (Mayer & Boness, 2011). This area of finding SOC promoting interventions could be a profitable area for further research to explore. The findings have both practical and theoretical implications.

Limitations and recommendations for further studies:
This study appraised the mediating influence of SOC between work and family stressors among professional level employees at a metropolitan municipality in South Africa. The study recognised various limitations in its findings and conclusions. The primary limitation is associated with the concept of causality. Even though evidence supporting the model's causality is postulated, it has not been actually verified. This study considered a soft modelling method that leaned more towards prediction rather than causality. According to Falk and Miller (1992), the capacity to control events is guaranteed by causation while predictability permits a limited level of control. The suggested model was based on structural equations, which assumes that the association between latent variables is linear (Hair, Anderson, & Tatham, 1998), is another limitation. Additionally, the cross-sectional design of this study is a limitation to its generalisation. However, future studies should assess the variables over a period, bearing in mind the subtleties of the individual parameters of assessing work and family satisfaction. Furthermore, the use of SOC as the only mediating variable may not have done justice to the linkage between work and family stressors. Other intervening variables such as work-life balance strategies, self-efficacy and support should be explored for mediation in the linkage, with SOC as a possible moderator.