Abstract
This paper examined how academic staffs’ extrinsic and intrinsic job satisfaction influenced their turnover intention. The role of the work–life balance was investigated in this relationship by distinguishing between work–family and family–work conflict. Data from 450 academics in Malaysian universities were collected and analysed using Partial Least Squares Structural Equation Modeling. The results revealed that both extrinsic and intrinsic job satisfaction significantly affected academics’ work–family and family–work conflict, as well as their turnover intention. Work–family conflict partially mediated the effect of extrinsic and intrinsic job satisfaction on academics’ turnover intention. Conversely, family–work conflict had no significant mediating effect in this relationship. This study has contributed to the existing body of knowledge on academics’ work–life balance by validating the differing (mediating) impacts of work–family and family–work conflict along with the relationship between job satisfaction and turnover intention. The results have important implications for universities’ human resource management by providing valuable insights on potential work–life balance policies for academics to reduce their turnover. However, the study's limited scope, comprising only five private universities in Malaysia, may constrain the generalisation of the findings.
Keywords
Introduction
Similarly to other developing nations, higher education takes centre stage in Malaysia as the crux of growth and innovation (Azman et al., 2014; Rajandram, 2020). It is one of the most critical sectors propelling Malaysia's talent development to spearhead its socio-economic growth, in line with the 11th Malaysia Plan and the Malaysia Higher Education Blueprint 2015–2025. Thus, both public and private higher education institutions (HEIs) need to attract, develop and retain a talented and competent workforce of academics to promote the nation's knowledge growth (Ganapathy, 2016; Tee et al., 2021). Unfortunately, the turnover of academics remains a significant challenge for private and public HEIs (Ainer et al., 2018; Tee & Chan, 2016). A work–life balance (WLB) is essential to enhancing academics’ productivity (McWilliam, 2007). Studies have revealed that a good WLB correlates with job satisfaction (JS), thus, minimising turnover intention (TI) (Forsyth & Polzer-Debruyne, 2007; Noor, 2011). Hussein et al. (2016) depicted that employees who could balance the needs of family and work are more; loyal, productive and motivated. Various research studies on work–family and family–work conflict discovered; increased pressure (Kinman & Jones, 2008), working long hours and jeopardised personal lives (Akkas et al., 2015), stress and juggling work and family duties (O’Laughlin & Bischoff, 2005); consequently, leading to TI (Kinman & Jones, 2008). Due to this significance, considerable efforts have been exerted to analyse the influence of work–family and family–work conflict in academic's JS and TI.
Turnover signifies quitting a job at one company and moving to another company in the marketplace (Al-Suraihi et al., 2021). TI is an employee's behaviour to cease or discontinue membership in an organisation (Cotton & Tuttle, 1986). At the same time, JS has been defined as a person's emotional state and pleasure resulting from appraising jobs and job experience (Locke, 1976). Spector (1997) defined JS as an attitude and the extent to which people like (i.e. satisfied) or dislike (i.e. dissatisfied) their jobs. This definition took into consideration dimensions, such as pay, promotion, supervisor relationship, co-workers’ relationship, fringe benefits, communication, contingent rewards, operational procedures and the nature of work. Indeed, JS is a vital factor affecting one's feelings, cognition, and thoughts. In JS literature, no universal definition has been offered covering the concept of JS. However, some scholars have suggested that it would be more appropriate and comprehensible to conceptualise JS under two main titles, namely extrinsic and intrinsic job satisfaction (Bektaş, 2017). Although the distinction between extrinsic and intrinsic satisfaction has been widely discussed from a theoretical standpoint, limited empirical studies have assessed JS in terms of its two components. An important question remains unanswered about their differing effects on employee outcomes (Fisher & Locke, 1992; Kuvaas et al., 2017). Nevertheless, prior studies have proven the negative relationship between JS and TI in the educational context (Kuwaiti et al., 2020; Pieters et al., 2020; Toker, 2011).
WLB is defined as the ‘accomplishment of role-related expectations that are negotiated and shared between an individual and their role-related partners in work and family domains’ (Grzywacz et al., 2006, p. 416). Prior studies have established that WLB is an essential determinant of TI in various contexts (Asghar et al., 2018; Fayyazi & Aslani, 2015; Oosthuizen et al., 2016), including the education sector (Azhar et al., 2016; Soomro et al., 2018; Watanabe & Falci, 2016) in general. Karatepe and Demir (2014) explained that when a person endures work and family role demands that are incompatible, participation in both roles leads to conflict. Thus, to better understand WLB, some researchers (Carlson & Kacmar, 2000; Karatepe & Demir, 2014; Ng & Feldman, 2012) conceptualised the WLB into two distinct constructs. These constructs were, work-family conflict (WFC), a form of inter-role conflict where role pressures from work and family domains are mutually incompatible, and family-work conflict (FWC), which occurs when family events interfere with work–life bidirectionally. Conflict arises from disagreement in fulfilling work and family duties that lead to work-related stress (Jamadin et al., 2015). To better understand this issue from the perspectives of academics, scholars noted that academic workloads, besides conducting teaching activities, such as demands to publish research, engagement of staff on a contract or non-contract basis (Annink et al., 2016; Jacobs & Winslow, 2004); training and supervision of students, attending meetings and consultancy (O’Laughlin & Bischoff, 2005); irregular work schedules, long hours, high job involvement (Jacobs & Winslow, 2004); negatively impacted on work–family and family–work conflict due to the difficulty in balancing work and family responsibilities (Bijawat, 2013; Walker et al., 2008). Past studies have associated the WLB in general with employees’ attitudes, such as satisfaction and TI (Pouluse & Sudarsan, 2014). To date, there has been a lack of studies that have considered the mediating impacts of the WLB, particularly in distinguishing the differing (mediating) effects of work–family and family–work conflict.
This study investigated the differing effects of extrinsic and intrinsic satisfaction on TI. Second, it probed the mediating role of work–family and family–work conflict in more detail to enhance the understanding of this under-researched topic to fill the gaps in the existing literature. This study drew upon the Social Exchange Theory (SET; Blau, 1964) to investigate the impact of WLB in greater detail by distinguishing how work–family and family–work conflict links between JS and TI. This study intended to contribute to the existing literature on people management, particularly determining the impacts of extrinsic and intrinsic job satisfaction and comparing the mediating effects of work–family and family–work conflict in workplace outcomes like TI.
Literature review
SET
Using the SET (Blau, 1964) as the underpinning theory, this study assumed that social relationships and the perceived support from employers would result in reciprocated behaviours, affecting the lives of individuals (i.e. employees) in an organisation (Renee-Baptiste, 2008). Employers who offered support and satisfaction to their employees expected higher work performance and commitment from them. In contrast, employees sought better JS in exchange for their dedication and loyalty to organisations (Cropanzano & Mitchell, 2005). Employees who received favourable treatment from their organisation would, thus, develop feelings of being cared for, in turn, engendering employees’ positive outcomes, such as commitment and low TI (Al-Suraihi et al., 2021; Riggle et al., 2009). In this regard, the SET was adapted to explain the relationship between JS, the adoption of WLB policies to reduce work–family and family–work conflict, and ultimately to decrease TI.
Predictors of TI
Turnover signifies quitting a job with one company and moving to another company in the marketplace (Al-Suraihi et al., 2021). Contrary to actual turnover, TI reflects an attitude or specific behavioural interest (Alam & Mohammad, 2010). TI indicates the likelihood of an employee leaving an organisation shortly. Since an individual's behaviour predicts their intention to leave an organisation, scholars have commonly used TI as a proxy to predict actual turnover (Cohen et al., 2016; Pitts et al., 2011; Richter et al., 2020).
Prior studies have established that the WLB is an essential determinant of TI. Soomro et al. (2018) noted that a better WLB (i.e. lower WFC/FWC) amongst academics could enhance their dedication and reduce their TI. Muteswa and Ortlepp (2011) found that organisations with a WLB policy created a more conducive work environment for their employees, reducing their TI. In addition, Amstad et al. (2011) and Oosthuizen et al. (2016) validated the impact of work–family and family–work conflict on employees’ work-family outcomes, such as employee turnover. Accordingly, this study investigated how work–family and family–work conflicts were related to TI. Further, this paper explains the reciprocation between academics and universities by affirming the direct relationship between extrinsic and intrinsic satisfaction, work–family and family–work conflict and TI, and the indirect effect of extrinsic and intrinsic satisfaction on TI via work–family and family–work conflict (as mediators).
WFC and FWC
The management of the WLB of employees has been the focal point of organisations as a way to retain human resources (Au & Ahmed, 2014; Bataineh, 2019). Greenhaus and Beutell (1985) pointed out that an individual's participation in work (family) caused them difficulty in participating in family (work). Researchers have used several terms, such as work–home interference (de Wind et al., 2021; Greenhaus et al., 1997) and home–work interference (Greenhaus et al., 1997; Smith et al., 2021) interchangeably to denote work–family conflict and family–work conflict. The WLB encompasses a balance between two separate roles performed by a person, namely work and family roles, and both bring satisfaction for the holder of the roles (Dikkers et al., 2005; Rahman et al., 2018). However, if a person endures work and family role demands that are incompatible, participation in both roles may lead to conflict (Aboobaker et al., 2017; Karatepe & Demir, 2014). To better understand the impacts of the WLB, this study conceptualised the WLB into two distinct constructs: work–family and family–work conflicts. Both work–family and family–work conflicts can be reflected in three different forms; time-based conflict, strain-based conflict and behaviour-based conflict, and these conflicts have been found to predict TI positively (Aboobaker et al., 2017; Carlson & Kacmar, 2000; Gragnano et al., 2021; Hamid & Amin, 2014; Oludayo et al., 2018).
Extrinsic satisfaction (ES), intrinsic satisfaction (IS) and TI
In reviewing definitions of JS, the most comprehensive appeared to be by Spector (1997), who defined JS as an attitude and the extent to which people like (i.e. satisfied with) or dislike (i.e. dissatisfied with) their jobs. Following Herzberg's two-factor theory (1959), the present study classified JS into two different constructs, namely ES and IS. Intrinsic factors, or satisfiers, relate to the person performing the job. In contrast, extrinsic factors or dissatisfiers relate to contextual factors, such as pay, fringe benefits, supervisor/co-worker relationships, contingent rewards, and operating procedures. Following a general evaluation of related previous studies, this study conceptualised JS into two distinct constructs, namely extrinsic and intrinsic job satisfaction (Bektaş, 2017).
Most scholars who have studied JS have confirmed that measures, such as pay, promotional opportunities, work environment, supervisor/co-worker relationships, recognition, policies, and procedures were significant indicators of JS (Spector, 1997). The factors above were found to have implications for employee TI and retention (Al-Suraihi et al., 2021). Extrinsic and intrinsic satisfaction trigger cascading emotional tendencies that cause employees to respond favourably towards their job (Luthans, 2011), which was found negatively related to TI (Kosi et al., 2015; Mughal et al., 2016). Akthar et al. (2016) established that both ES (i.e. pay, contingent rewards, fringe benefits) and IS (i.e. recognition from supervisors) were significantly linked to academics’ higher job enthusiasm and led to lower TI. Similarly, Hussain and Ghulam (2017) proved that ES (i.e. supervisory support, pay and co-worker relationships) and IS (i.e. promotion and nature of work) significantly predicted university lecturers’ TI. Accordingly, the following hypotheses were proposed:
Extrinsic satisfaction (ES), intrinsic satisfaction (IS) and work-family and family-work conflict (WFC/FWC)
The relationship between JS and WLB has also been frequently studied in the education sector. Naz et al. (2017) found that ES, such as support from an individual's immediate superior and co-workers, played a vital role in academics’ maintenance of family harmony and avoiding WFC caused by stress and heavy workloads. Similarly, Tee et al. (2021) noted that factors, such as collegiality (i.e. support from immediate superiors and co-workers), balance (i.e. the ability to balance between professional roles, personal life and family roles), expectations (i.e. clear job duties from universities) played essential roles in contributing to academicians’ success at universities. All these measures contributed to work–family and family–work conflict and led to TI. Achour et al. (2017) noted that superiors who offered advice to academics to resolve family conflicts might enhance academics’ work–family satisfaction and, thus, minimise work–family and family–work conflict.
Some researchers (Gragnano et al., 2021; Mesmer-Magnus & Viswesvaran, 2005) have stated that distinguishing work–family versus family–work conflict is an essential focus of WLB research. They have rightfully argued that the unidimensional nature of the WLB does not provide much insight into the window of employment opportunities, especially TI. Indeed, both forms of work–family and family–work conflict vary in nature and entail different predictors, outcomes, and processes (Gragnano et al., 2021). This study postulated that extrinsic and intrinsic satisfaction could minimise the work-family and family-work conflict and, thus hypothesised that:
Mediating roles of WFC and FWC
WFC links with the mechanism of negative interdependencies between work and home roles. This situation explains why the roles reciprocally affect one another negatively (Edwards & Rothbard, 2000; Ford et al., 2007; Nguyen et al., 2016). Netemeyer et al. (1996) emphasised that FWC denoted the contrary; it occurred when family matters interfered with work. This outcome was further discussed by some researchers who concluded that WFC negatively affected individuals, including their families and organisations (Barling et al., 2004; Nguyen et al., 2016).
Past research has revealed that private universities in Malaysia have more stringent academic policies regarding; assessments, research activities, knowledge dissemination, and student enrolment, which have influenced the academics’ roles (Tee & Chan, 2016). These increases in the demands on academics’ roles have subsequently affected their WLB and led to negative attitudes, such as intention to leave (Doherty & Manfredi, 2006; Mgaiwa, 2021). Moreover, existing WLB literature has traditionally focused on the conflicts between work and family and sought solutions to employees’ negative situations (Bennett et al., 2017; Erdamar & Demirel, 2014). That is an organisation that formulates a robust WLB policy that helps its employees organise their work–family roles, reducing workload and role conflict and lowering TI (Osibanjo et al., 2016; Pouluse & Sudarsan, 2014). Oludayo et al. (2018) further asserted that implementing WLB programmes could reduce work–family and family–work conflict, reducing TI.
Empirical evidence on the relationship between extrinsic and intrinsic satisfaction, work–family and family–work conflict, and TI has been chiefly direct. It has been demonstrated that extrinsic and intrinsic satisfaction significantly decreased TI (Azalea & Lim, 2015; Mughal et al., 2016). Other scholars have noted that extrinsic and intrinsic satisfaction predicted WLB (Khairunneezam et al., 2017; Torp et al., 2018). However, there have only been a limited number of studies concerning the indirect link between extrinsic and intrinsic satisfaction and TI via work–family and family–work conflict (Gözükaraa & Çolakoğlub, 2016). This paper extended this line of research and postulated the mediating effect of WFC and FWC in transmitting JS to lower TI (Byron, 2005; Fabienne et al., 2011; Michel et al., 2011). Concurrent with the SET, support from employers in work/family matters was perceived as ‘extra concern’ by employees, which enhanced their satisfaction at work. In return, employees subsequently reciprocated by demonstrating their commitment and loyalty to their organisation. As such, it is plausible that work–family and family–work conflict have an indirect influence that enables extrinsic and intrinsic satisfaction to lower TI. Along with support for the direct and indirect relationships between extrinsic and intrinsic satisfaction, work–family and family–work conflict, and TI, point towards the following hypotheses:
Figure 1 shows this study's research framework, including extrinsic and intrinsic satisfaction as independent variables, work–family and family–work conflict as mediators, and TI as the dependent variable.

Research framework.
Methodology
Population and sampling
This study applied a two-stage sampling technique. The first step is collecting data from the select private universities. The snowball sampling technique was used to identify private universities with high turnover rates due to TI cases are highly sensitive issues for the universities. Once the private universities were selected, the second stage was to determine the respondents from whom to gather the data by adopting purposive sampling to guarantee the selection of unknown participants (Gary, 2014). According to the Ministry of Higher Education, Malaysia (2018), there were twenty (20) public universities, forty-seven (47) private universities and thirty-four (34) private university colleges in Malaysia. However, only private universities were considered in this study. Simple random sampling was used to select the institutions for data collection. Universities that declined to participate were replaced by other universities that followed them in the list. Only five (5) private universities agreed to participate in this survey. The data collection was conducted from January 2020 to April 2020.
A total of 600 questionnaires were distributed, out of which 450 were returned, reflecting a 75% response rate. However, 50 responses were found to have more than 25% of unanswered items and, thus, were excluded from further analysis, resulting in a final number of 400 (66.67%) usable surveys for analysis. The response rate in this research was considered appropriate and sufficient, in line with the criteria suggested by past researchers (Choong et al., 2013; Fincham, 2008). In addition, the 75% response rate in this study was also higher than other research studies concerning academic TI in Malaysia (Azalea & Lim, 2015).
Most of the participants were female (55.5%), aged between 31 and 40 years old (57.3%), and married (74.5%). The majority of the respondents had attained a Master's degree (57.3%), while 38.8%t possessed a Doctorate degree. Besides, most (58%) of respondents had worked with their university for 1–5 years, whereas 42% had worked at their university for 6–10 years. Regarding job positions, most of the participants were lecturers (52.3%) and senior lecturers (43%). The rest (4.7%) per cent held the positions of Associate Professor and Professor, respectively.
Measurement and scales
The survey used a 5-point Likert scale to measure the respondents’ perception towards ES, IS, WFC, FWC and TI. The scale ranged from ‘1 = strongly disagree’ to ‘5 = strongly agree’. Spector’s (1997) Job Satisfaction Survey (JSS) was adapted to measure JS. Pay, promotion, co-worker relationships, fringe benefits, contingent rewards, and operating procedures were classified as extrinsic indicators, while the nature of work and promotion were considered intrinsic indicators. A total of 10 items were adopted from Netemeyer et al. (1996), of which 5 items were used to measure WFC, and 5 items were used to measure FWC. The TI scale from Boyar et al. (2008) was adopted to measure employees’ likelihood to stay with an organisation.
Data analysis
The Statistical Package for Social Sciences (SPSS) version 21.0 software application was used to compute the distribution of the respondent's profiles. In addition, Partial Least Square Structural Equation Modeling (PLS-SEM) was conducted to validate the measurement, structural and mediation models while verifying the hypotheses. PLS-SEM is helpful in behavioural and social sciences research to assess the multidimensionality, reliability and validity of each construct along with the degree of measurement error (Hair et al., 2018). In addition, PLS-SEM allowed the simultaneous testing of mediation with minimum bias, resulting in a better understanding of the full effects of the independent variables (ES and IS) and mediators (WFC and FWC) on the dependent variable (TI). The t-values and p-values were obtained through the output of bootstrapping with 5000 subsamples and a one-tailed test at a confidence interval of 95% (significance level = 0.05, one-tailed test) for the hypotheses which tested the direct relationships between extrinsic and intrinsic job satisfaction, WFC and FWC variables with TI. This study used a one-tailed test because the estimated value departed from the reference value in only one direction (Pillemer, 1991).
Results
Assessing the reflective measurement model
Internal consistency reliability, convergent validity and discriminant validity were tested to assess the reliability and validity of the reflective constructs (ES, IS, FWC, WFC and TI). Table 1 summarises the measurement model's assessment results, showing that all model evaluation criteria were met and supported the measurement model's reliability and validity. All indicators had a loading of 0.70 or above, and the average variance extracted (AVE) for all constructs (ES = 0.551; IS = 0.573; WFC = 0.624; FWC = 0.604; TI = 0.751) were more than the threshold value of 0.50. Internal consistency was inferred since the composite reliability for all constructs averaged over 0.70 (Hair et al., 2018).
Summary of measurement model assessment.
; FWC: family–work conflict; TI: turnover transaction; WFC: work–family conflict.
The heterotrait-monotrait ratio of correlations (HTMT) values for all pairs of constructs, as shown in Table 2, were lower than the threshold value of 0.85. This outcome added support to the bootstrapping result and showed that neither of the confidence interval values included the value 1.0 (Henseler et al., 2015). This finding indicated that the constructs were distinctive and captured phenomena not represented by other constructs in the model, which confirmed the discriminant validity of the model.
Results of HTMT criterion
ES: extrinsic satisfaction; FWC: family–work conflict; IS: intrinsic transaction; TI: turnover transaction; WFC: work–family conflict.
Assessing the structural model
Table 3 summarises the results of the structural model assessment and justifies the hypotheses for all direct relationships. The findings reported that both ES and IS had significant (t > 1.645, p < 0.05) negative effects on WFC (ES → WFC, β = −0.620; IS → WFC, β = −0.165) as well as FWC (ES → FWC, β = −0.425; IS → FWC, β = −0.122). Also, both extrinsic and intrinsic satisfaction had significant negative effects on academics’ TI (ES → TI, β = −0.111; IS → TI, β = −0.136). Furthermore, the structural model satisfactorily explained the variance of academics’ TI (62.9%) and substantially explained WFC and FWC (WFC = 51.3%; FWC = 22.6%). Overall, the model had an acceptable fit and predictive relevance since all the Q2 values were larger than 0 (Hair et al., 2018). Therefore, all the direct hypotheses, H1, H2, H3, H4, H5 and H6, were supported.
Results of structural model assessment and hypothesis testing.
ES: extrinsic satisfaction; FWC: family–work conflict; IS: intrinsic transaction; TI: turnover transaction; WFC: work–family conflict.
Note: *Significant at p < 0.05, t > 1.645, one-tailed.
Assessing the mediation effects
This study hypothesised that WFC and FWC were mediators between extrinsic and intrinsic satisfaction and TI. The mediation was tested using a 95% confidence interval (Henseler et al., 2015). Tables 4 and 5 report the (bootstrap) direct effect, including work–family and family–work conflict as mediators.
The direct effect with mediator (bootstrapping) for work–family conflict (WFC).
ES: extrinsic satisfaction; IS: intrinsic transaction; TI: turnover transaction.
The direct effect with mediator (bootstrapping) for family–work conflict (FWC).
ES: extrinsic satisfaction; IS: intrinsic transaction; TI: turnover transaction.
Table 6 reports the indirect effect, including work–family and family–work conflict as mediators. WFC was found to function as a mediator in the relationship between ES (β = −0.047; t > 1.645), IS (β = −0.013; t > 1.645) and TI. However, FWC only significantly mediated the relationship between ES (β = 0.05; t > 1.645) and TI but not between IS (β = −0.01; t < 1.645) and TI.
Results of mediation testing.
ES: extrinsic satisfaction; FWC: family–work conflict; IS: intrinsic transaction; TI: turnover transaction; WFC: work–family conflict.
Discussion and implication of study
This study tested the mediating effect of work–family and family–work conflict between JS and TI amongst academics in Malaysian private universities. It postulated that the decision to leave depended on work–family and family–work conflicts and extrinsic and intrinsic satisfaction. The results indicated that IS had a more significant negative effect on academics’ TI than ES. However, ES had more significant impacts on work–family and family–work conflict compared to IS. These findings suggested that fulfilling ES was essential for academics to manage their WLB, but IS was more important in reducing TI. In particular, this research was consistent with the prior research by Khairunneezam et al. (2017); their study pointed out that the WLB contributes to JS or vice versa. Heavy workloads and working overtime are the major contributors to creating work–family and family–work conflict. To restore WLB, family support is a vital source to increase life satisfaction.
Moreover, this study also confirmed that unmanaged WFC would be of more concern than FWC because WFC was found to predict the WLB better and was more significantly related to (increase) academics’ stress and, consequently, their TI (Boles et al., 2001; Noryati et al., 2009; Rode & Near, 2005). The result implied that academics who were extrinsically and intrinsically satisfied with their jobs had more balanced work–family lives (i.e. reduced WFC), leading to lower TI (Rothbard & Dumas, 2013).
On the other hand, FWC was significant only on the link between ES and TI. Still, it was not significant on the connection between IS and TI. The results indicated that support from supervisors and co-workers were essential to improving the working environment and helping to ease the burden of academicians, which resulted from WFC. This finding was in line with previous research findings by Stupnisky et al. (2015) and Achour et al. (2017); they pointed out that collegial support and relationships at work played an essential role in helping academicians reduce stress from FWC and, thus, balanced their work and family life. Academicians who experienced reduced FWC or WFC were less likely to leave roles at HEIs.
This finding also revealed that IS did not help academics resolve FWC if family events interfered with their work–life. It can be presumed that the conflicts encountered by academics would be primarily from job stressors rather than from family interferences that affect work–life. The minor role of IS on FWC meant that Malaysian academics might encounter lesser family interferences on their work–life than work interferences on family lives. This situation may be explained by the flexible nature of academics’ work which provides them with a greater ability to manage their family matters (Gunawan et al., 2018). Thus, it would be reasonable to assume that academics who continually face clashes between their work and family roles (i.e. WFC) would have a stronger intention to find another job. Hence, IS is not the significant predictor for FWC dimensions, such as time-based conflict, strain-based conflict and behavioural-based conflict. As such, this study's results were contrary to previous findings by Bruck et al. (2002), who found that both work–family and family–work conflict related to both types of extrinsic and intrinsic job satisfaction. Practically, this may be good news for private universities seeking to devise ways to minimise how an academic's work intrudes on their family life rather than controlling the extent to which family life interferes with work.
Implication of study
Theoretically, this study has shed light on the underlying framework of the SET in explaining how JS translated into lower TI via managing employees’ WLB. This research tested the SET theory empirically through the domain of JS and WFC and FWC. However, ample evidence exists of the relationships between extrinsic job satisfaction with numerous variables, such as TI (Akthar et al., 2016; Hussain & Ghulam, 2017). However, extrinsic and intrinsic job satisfaction has been examined less with work–family and family–work conflict in general or in the context of education sector.
This study extended the concept of social exchange to the domain of WLB. It delved further by examining how work–family and family–work conflict can separately and consequentially lead extrinsic and IS into a reduced TI. Specifically, the extrinsic job satisfaction domain, such as the social relations of supervisors and co-workers, promoted and strengthened the exchange relationship for academicians, reducing work–family and family–work conflict. It was also noted that extrinsic and intrinsic satisfaction could minimise WFC, which, in turn, led to lower TI. The findings of this research were, thus, in line with the SET, which stipulates that employees who feel being taken care of by a company will engender beneficial consequences (i.e. lower TI) to the employers (Cropanzano & Mitchell, 2005). That is to say that the fulfilment of work–family roles was a significant factor in triggering reciprocal social exchanges amongst academics who were satisfied with their job and exhibited more commitment or loyalty to their organisation.
Practically, this study has provided empirical evidence on the mediation effect of the WLB, specifically WFC and FWC, on academics’ TI. The findings suggested the critical need for organisations to manage employees’ WLB (Adebayo, 2016). A proper WLB programme in conjunction with extrinsic and intrinsic satisfiers is essential to modify employees’ behaviours at work (Asghar et al., 2018). As such, it is recommended that private universities offer more support for academics’ work–family balance, such as implementing the flexi-hours workload, mentoring systems and developing academics’ career profiles to enhance their work processes. This outcome will reduce WFC, leading to a lower intention to leave an organisation.
Limitations and future research
The present study has several limitations that must be highlighted. First, cross-sectional method was used to collect the data at one point in time. Since JS, WLB and TI are time-bounded concepts that reflect individuals’ accumulated experiences and their effects over time (Judge et al., 2010), the results are not guaranteed to be a good representative. Therefore, longitudinal data could be collected at more points of time in future research to help predict WFC and TI over time.
Second, this study's sample was restricted to academics from five private universities in Malaysia, which may not represent the entire population of the Malaysian higher education sector. Thus, future researchers are urged to include respondents from more universities to widen the generalisability scope. Nonetheless, the results of this study did provide insight into the predictor of academics’ TI.
Conclusion
In conclusion, this study uncovered the impacts of extrinsic and intrinsic satisfaction on the WLB and, ultimately, academic staff's TI. These research findings are significant for universities to formulate and implement effective human resource management strategies to manage their academic staff's WLB and retention. Moreover, acquiring and retaining human capital in academia is vital for universities to create a sustainable competitive advantage to cope with the intense competition due to globalisation and the commercialisation of the higher education industry.
Footnotes
Author's Note
Poh Kiong Tee is now affiliated at International University of Malaya-Wales, City Campus, Ground Floor, Block A, Jalan Tun Ismail, 50480 Kuala Lumpur, Malaysia.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship and/or publication of this article.
Supplemental material
Supplemental material for this article is available online.
