Understanding work engagement through multi-level investigations: From antecedents to outcomes

Work engagement is a valuable organizational resource, and it has many positive outcomes. In today’s dynamic and competitive business environment, organizations can only be successful when they have an engaged workforce. The current study examines how supervisors support and fairness is important for employee’s work engagement and how the role of perceived organizational support is significant in an employee's workplace. In line with Organizational Support Theory (OST) and Conservation of Resource Theory (CRT), the current study is an empirical attempt to explain antecedents that could increase work engagement and resultantly fetch productivity and profit in the context of South Asia (the case of Pakistan) by considering the outcomes. This cross-sectional study draws data through structured questionnaires from 310 employees of eight mega retail stores by using simple random sampling. The outcomes of Structural Equation Modeling (SEM) analysis revealed a positive effect of supervisors’ support and organizational fairness on work engagement through the mediation of Perceived Organizational Support (POS) and explained the impact of work engagement on task performance and career satisfaction of employees. The current study tested the model for work engagement; future research might test the model using other employee factors (employee sustainability or motivation) in order to test continuous employee behaviors in their workplaces.


Introduction
*Employee engagement refers to the intensity and altitude of commitment, eagerness, and inclination, which an employee demonstrates towards his work, and these levels are the signals of his/her devotion towards the work. In today's highly dynamic and competitive business patterns, organizations can only be successful when their workforce is engaged in their everyday jobs. In the Gallup Survey 'State of the Global Workplace (GC, 2017)', it has been exposed that only 15 % of full-time employees are engaged at work whereas in Asia this percentage is down to 6% (State of the Global Workplace, (GC, 2017)). This survey also revealed that there were foremost monetary penalties for businesses all over the world tune to $7 trillion in the shape of lost productivity. Geographically splitting the loss, it spears that Western Europe faces 10%, Eastern Europe confronts with 15%, Post-Soviet states encounter 18%, the Middle East and North Africa cope with 58%, Sub-Saharan Africa handles 14%, East Asia deals with 57%, South East Asia faces 22%, South Asia stands on 28%, Australia and New Zealand receives a challenge of 14%, Latin America realizes 32%, the United States and Canada meet 31% and overall best companies in the world are confronted with 70% of the monetary loss.
Next year, in the report of Gallup Survey 'The Engaged Workplace (GC, 2018)', results were quite similar to the previous year's report, and the fact has been established that eighty-seven percent of worldwide employees are not engaged at their work (The Engaged Workplace, (GC, 2018)). Employee engagement market review by Bersin (2015) highlighted that various companies are spending massively on employees engagement, about $720 million per year spent in the US and the forecast is that it will soon grow to about $1.5 billion, and individuals who are truly engaged in the workplace have consistently averaged less than 33% (Kowske, 2012;Mann and Harter, 2016). By keeping the facts in mind, this study intends to reveal a secret that offers an answer to; what is the key to upsurge work engagement, and what benefits an organization can earn by engaging employees?
Supervisors are considered as the representatives of organizations; therefore, supervisors' support and role of fairness pacify employees' emotions and inspire them to be engaged in work additionally (Rhoades and Eisenberger, 2002). The organizations which care about the well-being and fulfill the socioeconomical needs of employees start gaining the confidence and trust of employees, and academically it can be termed as Perceived Organizational Support (POS) (Baran et al., 2012). Employees with positive POS feel less pressure, more motivation, and increased engagement at work (Shaw et al., 2013). High work engagement has exposed its association with positive upshots such as career satisfaction and high task performance (Schaufeli et al., 2002;Salanova et al., 2003;Schaufeli and Bakker, 2004).
From the South Asian perspective, Ram and Prabhakar (2011) took a holistic view and presented a matrix of work engagement predictors but ignored the factors that strengthen the predictors. Other important studies regarding antecedents of work engagement were conducted by Biswas et al. (2013) and Nikolova et al. (2019), but they did not consider outcomes of the work engagement. Research work by Dajani (2015) on the Egyptian banking sector revealed job performance and organizational commitment as the resultant of work engagement but remained unsuccessful to inculcate the factors that are important from the employee perspective, whereas the study of Bano et al. (2017) in the context of Pakistan talked about important factors from an employee perspective, but they missed work engagement.
By keeping in view the research gap, the current study is an empirical endeavor to explain three questions pertaining to work engagement in the South Asian setting; a. Is there any positive influence of supervisors' support and fairness on employees' work engagement? b. Is there any mediating effect of POS between the relationship of supervisors' support, fairness, and employees' work engagement? c. Is there any impact of work engagement on task performance and career satisfaction?
These questions offer various perspectives relating to work engagement; on one side, what influences work engagement: Supervisors' support, fairness, and on the other side the outcomes of work engagement: Task performance, career satisfaction, which are highly valuable for employees as well as the organizations in South Asian setting. There is a dearth of explicit contextual clarification of the topic at hand, and the present study explains the individual and organizational level outcomes that are directly influenced by employee work engagement.

Literature review and hypotheses development
In subscript, the review of the literature explains the variables and rationalizes the study hypotheses, respectively. Eisenberger et al. (1986) explained the tenets of organizational support theory (OST) that employees assess the work-benefit ratio through the fulfillment of social and emotional needs, and during this assessment, process employees develop a generic opinion about how much the organization is supportive and fair in appreciating their work input and looks after to their welfare. Moreover, OST elaborates that employees who observe a high level of organizational resource support feel the responsibility to repay the organization through appropriate behavior and a positive attitude (Eisenberger et al., 1990). Moreover, conservation of resource theory also suggests that individuals with extra or saved work resources develop a habit to invest in the firm again (Hobfoll, 2001).
Work engagement is a positive psychological state comprising of drive, determination, and dedication for work tasks (Schaufeli et al., 2002). It is a motivational impression that signifies the vigorous distribution of individual properties toward work duties and roles (Rich et al., 2010). Work engagement is a valuable organizational resource as it shows many positive outcomes (Bakker and Albrecht, 2018;Schaufeli et al., 2002). Apparently, it seems like job satisfaction and job performance, but scholars such as Newman and Harrison (2008) clarified that work engagement is a separate construct, and an academician like Salanova et al. (2003) argued that it is an outcome of employee performance. London (1993) defined employee engagement as employees' positive perception that supervisors and organizational support permit them to control their own work, performance goals, and career plan. Kahn (1990) proposed the resultant of employee engagement that employees experience an emotional connection with their work, implement their duties/roles interactively, enthusiastically, and emotionally, in a general and immediate manner. The current study is using Kahn's approach as an operational definition.

Relationship between supervisor's support and work engagement
The supervisor's support is subordinates' psychological recognition that his/her supervisor is concerned, caring, and offers all possible support when needed at the workplace (Burke et al., 1992). Stinglhamber and Vandenberghe (2003) explained supervisory support as the belief of employees about the quantity in which the quality of helping relationship is available by the supervisor during or for the work task. The supervisor's behavior helps to enrich performance, adds to organizational efficiency as well as employees' well-being (Chou and Stauffer, 2016). In organizational work structure, supervisors are recognized to play an important role in developing subordinates expectations (Graen and Scandura, 1987). Supervisory support with regard to organizational culture and managerial dimension is relatively the most important area and has a significant potential to develop the employees' engagement (Rhoades and Eisenberger, 2002;Baran et al., 2012). Supportive supervisors are those organizational resources that provide a role of giving importance, guidance, feedback, and assistance to employees because employees typically rely on their supervisors for work engagement outcomes and prospective professional growth (Loi et al., 2014). Wayne et al. (1999) mentioned that directions and career support by a supervisor tend to engage employees more in work. Supervisory support is observed as representative of the work atmosphere that offers emotional and noticeable resources that affect the emotional level of employee engagement (Bakker and Demerouti, 2008;Hobfoll, 2001). The previous research finds that supervisors' thoughtfulness towards employees concocts the needs towards the contribution of employee engagement (Swanberg et al., 2011). Therefore the connection between employees and supervisors is important in stimulating and encouraging the employees. Considering the above deliberations, the current study presumes: H1: Supervisor's support positively influences work engagement.

Relationship between fairness and work engagement
Probably the first explicit suggestion that fairness should be considered was made by Scott (1941) when he explained a principle of accounting and categorically stated, "Accounting rules, procedures, and techniques should be fair, unbiased, and impartial." The concept of fairness was further developed by Greenberg (1987) in the context of organizational justice. Research on fairness has boomed in many previous years (Kunnan, 2004). Individuals are obviously attentive to the fairness of situations and proceedings in their daily lives, across a change of contexts (Tabibnia et al., 2008). Fairness is often of dominant concern to organizations because the perceptions of fair and unfair can influence job attitudes and behaviors (Barsky et al., 2011), motivation in work (Latham and Pinder, 2005), and job satisfaction (Al-Zu'bi, 2010), therefore it is evident that an individual's observations of any organizational decision as fair or unfair can affect the person's attitudes and behaviors at the work environment. Studies also propose that the significance of employees' emotions and affect in the assessment of fairness at the workplace, and thus result in an attitudinal reaction (Barsky et al., 2011).
More lately, advocates of fairness theory proposed that "deontic responses" natural, involuntary, and adaptive responses to unfair treatment (Folger and Cropanzano, 2001) and those responses catalyze and can lead to explosive and potentially destructive behavior in organizations (Barclay et al., 2005;Krehbiel and Cropanzano, 2000;Weiss et al., 1999). Employee work engagement is also related to fairness because fairness has the core element of perceived organization support (Kurtessis et al., 2017). The literature has extensively studied the concept of fairness related to the employees' work engagement in different fields of area (Baran et al., 2012;Bies, 1986). Based on the said arguments current study expects that significant positive relation exists between fairness and work engagement. Thus, the hypothesis is: H2: Fairness positively influences work engagement.

Relationship of work engagement with task performance and career satisfaction
In previous studies, researchers figured out some outcomes of work engagement, i.e., career satisfaction, organization commitment, intention to quit (Andrew and Sofian, 2012; Bothma and Roodt, 2012), and organizational citizenship behavior (Afacan, 2015). Organizations' ultimate goal is to improve employees' task performance and satisfaction to retain them as top performers (Ulrich, 1997). In the present study, task performances and career satisfaction are proposed as work engagement outcomes. Task performance is any assessment that examines the employees' exhibition of their learning, knowledge, understanding, and work execution that yield a tangible performance as proof of learning (Marzano et al., 1993). Locke (1969) defined it as the satisfying emotional state resulting from the valuation of the individual job as attaining or simplifying individual job beliefs. In an exploratory study of Engelbrecht (2006), it has been observed that highly engaged employees are more client-oriented in their workplace and found peaceful and persistent in dealing with customers. Schaufeli and Bakker (2004) suggested that engaged workers earn higher evaluation scores from their supervisors on actual-role and extra-role performance, and such workers even reach the stage of enthusiasm and pleasure to work for an organization. Shaver and Lacey (2003) categorized employee work-related satisfaction into two broad types: First, career satisfaction that denotes to worker's choice of career and the manner in which experience of work justifies that choice. Second is job satisfaction that indicates the reaction of workers to their proximate work environment and employer. It is well anticipated with career satisfaction because of its connection with career choice and expectations of workers' jobs. It might have a more lasting relationship with work engagement, parallel to specific psycho-social features of the work environment more conventionally associated with the job demand (Demerouti et al., 2001;Alarcon et al., 2011). Thus, research on the significance of work engagement has exposed its association with positive upshots such as career satisfaction and high task performance (Salanova et al., 2003;Schaufeli and Bakker, 2004;Schaufeli et al., 2002). On the basis of the above arguments, the following hypotheses can be proposed: H3a: Work engagement positively influences the task performance of employees. H3b: Work engagement positively influences the career satisfaction of employees.

Mediating role of perceived organizational support
Perceived organizational support can be defined as employees' belief that the organization shows concern about their welfare and meets their socially embedded emotional needs (Baran et al., 2012). OST claims that to fulfill social and emotional needs, employees assess the work-benefit ratio, and during this process, employees develop a generic opinion about how much the organization appreciates their work input and looks after their welfare. Such perceived organizational support (POS) results in higher employees' thoughtful sense of responsibility toward the organization, increases their sentimental commitment with the organization, and encourage employees' level of expectancy; that better performance would result in a better reward. Thus behavioral outcomes of POS would include an upsurge in in-role and extra-role performance and fall in retreat behaviors such as absence or permanent departure from work.
OST proposes that employees denote human-like features to their working place, which results in the development of POS (Eisenberger et al., 1986). Supervisors are considered as agents who act on behalf of the organizations and exhibit organizations' scheme of thought instead of their own (Levinson, 1965). It means employees base their judgment about organizational rules, standards, and philosophy, on the behavior exerted by the supervisors over the individuals. Researchers have found that subordinates with high POS feel less pressure at the workplace and are more motivated to work (Shaw et al., 2013). Rhoades and Eisenberger (2002) established the fact that there are three key types of POS: Supervisor support, fairness, and favorableness of organizational rewards. Therefore, POS leads to improved workers' engagement when a supportive supervisor helps, and organization fairness is available (Baran et al., 2012).
The study of Eldor and Harpaz (2016) originated an inspiring and motivating learning environment where organizational fairness and supportive supervisor are positively related to employee work engagement, and employees are able to practice in many resources to build a higher level of engagement with the help of favorable organizational support. Therefore, the present study argues that the supervisor's support and fairness facilitate employees' perceived organization support, and this kind of organizational resource transforms them into greater work engagement. Thus, the following hypotheses can be proposed from the above-stated arguments: H4a: Perceived organizational support mediates the relationship between supervisory support and work engagement. H4b: Perceived organizational support mediates the relationship between fairness and work engagement.

Sample and procedure
The current study is based on a quantitative research design and used a survey of a questionnaire as a tool to gather data from employees of mega retails stores in Pakistan. Out of 30 megastores/ hypermarkets (Balochistan=4, Islamabad=8, Punjab=12, Sindh=6) enlisted with Securities and Exchange Commission of Pakistan (SECP), 12 retail stores of province Punjab were selected for this study by using a purposive sampling technique. The sample size was comprised of 310 respondents, decided with the technique of the total number of items in instrument multiply with the number of 10 (Tabachnick et al., 2007). A total of 600 questionnaire survey forms were distributed in 12 megastores, 350 filled forms were received back, out of which 310 were usable for data analysis. The response rate of the study was 58.33%. To test the proposed model, six constructs have been used out of which two are independent variables, namely supervisor support and fairness, and work engagement, task performance, and career satisfaction are used as dependent variables in this study, whereas perceived organizational support has been used as a mediator between supervisor's support and work engagement and fairness and work engagement. The unit of analysis of this crosssectional study were the employees of retail stores working at any organizational level. In Table 1, the demographic statistics of the sample are summarized.

Measures
This study used established scales for all the constructs. To measure the supervisor's support, the five items scale of London (1993) is used. The sample item is "my supervisor provides enough time for me to attend training." The four items scale of Colquitt et al. (2001) is used to measure fairness. The sample item is "Your Company treated you with dignity." Perceived organization support (POS) is measured by four-item scales established by Hofmann and Morgeson (1999). The sample item is "Your company value your contribution." Work engagement is measured by the nine items scale established by Schaufeli et al. (2006). An example item is "At my work, I feel bursting with energy." Task performance is measured by three items scale developed by Griffin et al. (2007).
A sample item is "Subordinate carried out the core parts of your job well." Career satisfaction is measured by five items scale from Greenhaus et al. (1990). An example item is "I am satisfied with the success that I have achieved in my career." All measurement scales were based on existing literature and responded on a "five-point Likert scale ranging from 1= (strongly disagree) to 5=(strongly agree)."

Data analysis technique
Data analysis was conducted in SMART PLS 3 software. PLS structural equation modeling is better to measure the relationship among variables (Hair et al., 2012). It followed two steps modeling approach. In the first step measurement model was analyzed, and then the structural model was evaluated. In the reflective measurement model, internal consistency, convergent validity, and discriminant validity were analyzed first, and in the second step of analysis in the structural model, multicollinearity assessment, size, and significance of the path coefficients and coefficients of determination (R 2 ) were measured. The researchers used IBM SPSS version 20 for finding descriptive statistics and SMART PLS 3 for Path coefficient analysis to study the relationship and the effects of the independent and mediated variables on the dependent variables.

Evaluation of measurement model
For the first step, the PLS algorithm was run to measure the value of Cronbach's alpha, Composite Reliability, and Average Variance Extracted (AVE). The internal consistency is significant because Cronbach's alpha was greater than 0.7 (Shemwell et al., 2015;McCrae et al., 2011), and the composite reliability value is also greater than 0.7 of all variables the AVE value is greater than 0.5 (Table 2).

Internal consistency
Internal consistency is generally measured by using Cronbach's alpha values, but in the case of PLS-SEM, composite reliability is measured by considering different outer-loadings of the indicator variable. The value of composite reliability is between 0 and 1 and can be interpreted the same as Cronbach's alpha values. The following Table 3 represents the values of composite reliability of all variables.  The composite reliability of all variables was greater than 0.7 and less than 0.95, so all variables are reliable.

Convergent validity
To establish convergent validity, outer-loadings of indicators (Table 4) and average variance extracted (AVE) ( Table 5) are considered. The value of outer-loadings should be greater than 0.7. However, values with weaker loading can be retained if these indicators are important to establish content validity; therefore, values with loadings less than 0.4 must be deleted (Hair et al., 2017). Secondly, the value of the average variance extracted should be 0.5, or higher is acceptable to establish convergent validity.
All the indicators of factors have outer-loadings greater than 0.4; therefore, no indicator is deleted in the analysis. Table 5 indicates that all the value of the average variance extracted for all variables is greater than 0.5. Thus, it is proved that both of the conditions to establish convergent validity have been met.

Discriminant validity
There is a new approach to access discriminant validity using the heterotrait-monotrait ratio of correlations (HTMT). HTMT value below 0.90 is the criteria for establishing discriminant validity ( Table  6). As shown in Table 6, heterotrait-monotrait ratio (HTMT) for all variables is less than 0.90. Hence discriminant validity is achieved.

Evaluation of structural model
The second step in PLS-SEM is to evaluate the structural model. In this step, evaluation of the structural model was performed by multicollinearity assessment, size, and significance of path coefficients and the coefficient of determination (R 2 ) the structural model.

Multicollinearity assessment
In PLS-SEM, a tolerance value of 0.20 or less and a VIF value of 5 and higher shows collinearity problems. Table 7 represents the VIF values of all variables. Table 7 presents values for VIF, and all are less than 5. Therefore, there are no collinearity issues. Table 8 shows the values of R 2 . The value of R 2 lies between 0 and 1, and higher values show a higher level of predictive accuracy.   The coefficient of determination (R 2 ) of all variables shows a high level of predictive accuracy.

Significance and relevance of structural model relationships
In Fig. 2, the results of statistical testing for the structural model with path coefficients have been shown. The supervisor support positively relates to work engagement (β=0.435, t=6.073, p<0.001), so H1 is accepted, and it also positively relates to POS (β=0.503, t=5.397, p<0.001).
The results express that fairness has a negative relation with work engagement (β=-0.025, t=2.08, p<0.03), so results rejected the H2, but it positively relates to POS (β=0.116,t=2.503,p<0.01). Moreover, the results showed that work engagement is positively related to career satisfaction (β=0.114, t=3.953, p<0.001) and task performance (β=0.464, t=6.480, p<0.001), which shows that the H3a and H3b are accepted. In addition, mediation analysis is done by following the procedure PLS bootstrapping, and results showed the indirect effects between variables and direct effect for path coefficients, so mediation results found that the indirect effect explained more variance for work engagement because the ratio of direct to an indirect effect of supervisory support and fairness on work engagement was 0.635 and it is greater than the ratio of direct relation with work engagement it shows mediation and supports H4a and H4b. Table 9 shows a summary of the structural model. ***Significant at the 0.001 level; **Significant at the 0.01 level; *Significant at the 0.05 level

Discussion
This research investigates how supervisors' support and fairness are important for employee's work engagement and how the role of POS is significant in an employee's workplace. Specifically, the study supports the Organization Support Theory, which highlighted that employees who observe a high level of organizational resource support will feel the responsibility to repay the organization through appropriate behavior and positive attitude (Eisenberger et al., 1990) and Conservation of Resource Theory which suggested that individuals with extra or saved work resources develop a habit to invest in the firm again (Hobfoll, 2001).
Organization support theory offers clarifications for possible causes of such factors of employees, in theory, proposed model. A conceptual framework links the key constructs to each other and provides a base for emerging empirically testable hypotheses. The current paper extends prior research (Rhoades and Eisenberger, 2002;Baran et al., 2012;Salanova et al., 2003;Schaufeli and Bakker, 2004;Schaufeli et al., 2002;Biswas et al., 2013;Abbas and Wu, 2018;Nikolova et al., 2019) in several ways. More specifically, current research tried to reveal a knowledge understanding of work engagement in three ways. First, the supervisors will recognize the circumstances that they should make categorization in employees' careers and should support in the daily routine of performing tasks to create a satisfied employee and increased productivity. While most supportive supervisors' related studies look at employee's task performance (Swanberg et al., 2011), the current study has extended the supervisor related support debate within the organization and established the fact that supervisors enhance the employees work engagement with the support of POS and ultimately it improves the employee's career satisfaction and their task performance. This study remained successful in revealing the second truth that supervisor's support, fairness, and perceived organizational support are associated with organization support theory (Eisenberger et al., 1990). Fair treatment with the employee in the organization and supervisory support plays a significant role in organizational as well as employee development. Both factors show a significant impact over work engagement if an employee perceives that he is getting organizational support. The positive relationship specifies that higher levels of POS and supervisors support among employees higher the employees work engagement. Thirdly, this study amplifies the importance of perceived organizational support as a mediator between supervisor's support and work engagement and fairness and work engagement. This study strengthens the arguments of Baran et al. (2012) that perceived organizational support leads to improved workers engagement when a supportive supervisor helps and organization fairness is available at all possible levels.

Theoretical and practical implication
Present research creates three important theoretical contributions. First, this study contributes to employee work engagement and its outcomes literature with organization support theory by representing fairness, supervisory support, and POS mediator as one key factor of employee work engagement and its outcomes in terms of better task performance and career. Second, from an organization support theory perspective, the present research purposed model offers a more broad understanding of how fairness, supervisory support, and POS relates to employees work engagement and its outcomes in terms of task performance and career job satisfaction, whereas the extant literature (Yang et al., 2018;Ng and Sorensen, 2008;Chen et al., 2016) usually work of Kahn (1990) theory on work engagement to clarify the effects work engagement. The Organization support theory viewpoint offers numerous fascinating novel research commands to extend these theories of the effects of career-related satisfaction with work engagement. Thirdly, current research also contributes that fairness is not any positive connection with employees work engagement in the present study context the past research's arguments that fairness is positively affected on work engagement are not proved (Baran et al., 2012;Wang et al., 2015), it would be proved may be in other contexts.
This study has practical implications for supervisors and organizations that they should closely observe their workers' problems regarding unfair treatment or any discriminating behavior and then provide useful suggestions to help and support so that employees feel more motivated and engage well in their work. The employees having minimal work engagement can be devastating for organizational productivity. Their organization and supervisors should offer valuable ways for these workers in the form of training/mentoring sessions. As a result, employees will get more engagement in their tasks and duties.

Limitations and future directions
The current study carries some limitations; firstly, the coefficient of fairness in path analysis of the structural model was negative, which is a clear indication that fairness and work engagement has a negative relationship with each other, which is a deviation from literature support. In the future, this relationship should be carefully studied and investigated through a larger data set for a clear understanding of deviation. Secondly, the present research only tested the model for work engagement. Consequently, future research might test the model using other employee factors (e.g., employee sustainability, motivation, etc.) in order to test continuous employee behaviors in their workplaces. Thirdly, this model should be replicated with a broader scope in different industries and contextual settings.

Conclusion
The results of this study suggest that there is a meaningful positive relationship between supervisory support and work engagement, and perceived organizational support is a mediator between their relationships. The relationship between fairness and work engagement is proved as negative in this study, but after intervening in perceived organizational support, this relationship turns positive. This study also showed that task performance and career satisfaction are two significant outcomes of work engagement. So, in this way, the current study completed its journey of revealing the secret of understanding work engagement through multi-level investigations by discussing antecedents and outcomes. Organization support theory and conservation of resource theory are proved to be the underpinning theories of this study.
The theoretical framework of this study gives an in-depth understanding of engagement, along with relevant measures and the results of the engagement. Therefore, organizations using this framework will know which factors have a noticeable impact on employee engagement.

Conflict of interest
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.