The Impact of Employee Engagement on Employee Retention: The Role of Psychological Capital, Control at Work, General Well-Being, and Job Satisfaction

Purpose – Employee engagement and retention is a promising area in management as well as psychology. The purpose of this paper is to examine the effects of employee engagement (EE) on their retention (ER). We proposed a theoretical framework modifying (M. Gupta, 2018) model explaining how employee engagement affects their retention. The original model linked engagement with control at work (CAW) and general wellbeing (GWB) through psychological capital (PsyCap). We extended this model by linking CAW and GWB to employee retention. We further added job satisfaction (JS) as a mediator between EE and ER. <br><br>Design/methodology/approach – To establish its empirical validity, we conducted a survey from 200 employees working in different companies in Pakistan by using a closed-ended Likert scale type questionnaire. Data were analyzed using confirmatory factor analysis and structural equation modeling.<br><br>Findings – Findings suggested a positive significant relationship between EE and ER. PsyCap, CAW, and JS play a significant mediator role for EE and ER, only GWB does not mediate the relationship between EE and ER.<br><br>Practical Implication – This will help in retaining employees and mounting psychological capital through training and development since when employees will be happy and motivated, they can perform well and will have job satisfaction. As a result, the level of the intention of employees to leave would below. <br><br>Originality/Value - This study is preliminary the first to investigate the association between EE and ER through many other constructs (PsyCap, CAW, GWB, and JS) in the Pakistani context.<br>


Background of the Study
The idea of employee engagement has emerged in gaining recognition among professional health psychologists due to its positive impact on employee's wellbeing and control at work (Shuck, 2011). An astounding 90 percent of the workforce thinks, that a policy on employee engagement, would help to

Gap Analysis
There has been some research on employee's wellbeing and life satisfaction earlier, which is found in various fields like community health, education, and criminal justice. However, they usually do not observe the nature and background of work, depicting just correlation results (Erdagon, 2012). In spite of the lately growing consideration, practical studies on wellbeing are still lacking. Former studies have merely focused on non-working populations including students, patients, children and others. Thus, it is felt that there is a research gap, due to the lack of consideration paid to workers wellbeing in the management field, as the literature on wellbeing has overlooked the work related domain (Erdagon, 2012). It has only happened in recent times that employee wellbeing is alleged to be connected with an individual's personal life and work related results has begun to be researched in the field of organizational behaviors and human resources. The research of (Erdagon, 2012) has recommended an idea of workplace contentment, stating that it is based on satisfaction with one's surroundings, direction, job advancement, work distinctiveness and person's atmosphere fit. Contributing to an understanding of well being at work. Similarly, the research of (M. Gupta, 2018) explains how employee engagement is linked with control at work (CAW) and general wellbeing (GWB) through psychological capital (PsyCap). Specifically in the healthcare industry of India.
Findings indicated that PsyCap mediates completely among EE and CAW and moderately between EE and GWB. However, the study of (M. Gupta, 2018) was limited for the healthcare sector in India, so the results cannot be generalized. Moreover, the role of culture was neglected as well. We Modified (M. Gupta, 2018) model to include retention in the framework. The original model linked engagement with control at work (CAW) and general wellbeing (GWB) through psychological capital (PsyCap). We extended this model by connecting CAW and GWB to employee retention. We further added job satisfaction (JS) as a mediator between EE and ER. This will give a more detailed picture including incorporating the effect of other factors like employees satisfaction relevant to the generation. Currently, there has been limited research conducted in Pakistan on employee engagement and retention. This research will not only see the impact between these two variables however, it will also include the mediating role of PsyCap, GWB, CAW and JS.

Research Objectives
The objective of this study is to empirically test the impact of employee engagement on employee retention and to see how different variables such as psychological capital, control at work, general wellbeing and job satisfaction mediates with each other and among variables in Pakistani context. In this study we are investigating that how employee engagement (EE) impacts on employee retention (ER) and how Psychological capital (PsyCap), control at work (CAW), general well being (GWB) and job satisfaction (JS) mediates between the two variables. Even though, employee engagement, job satisfaction and employee well being have specific and significant affiliation with each other. Though, in the current study we consider them to be distinctive constructs.

Research Question
1. How will PsyCap affect the relationship between EE and ER, between EE and CAW, between EE and GWB and between EE and JS. 2. How will GWB affect the relationship between EE and ER and between PsyCap and ER. 3. How will CAW affect the relationship between EE and ER and between PsyCap and ER. 4. How will JS affect the relationship between EE and ER, between PsyCap and ER.

Significance
The role of employee engagement cannot be neglected in today's workplace. In order to make employee content and stay within their organization for a longer period of time, employers need to understand the importance of engaging employees. Hiring workers with high level of psychological capital, providing resources to increase their well being and making them satisfy for the jobs through different means is as necessary as anything because employees are assets that cannot be copied, however, if not provided with proper opportunities they intend to leave. This study will help managers to recognize how employee level of engagement will impact their retention and how other construct will catalyze this relationship (PsyCap, GWB, CAW and JS). Engagement is not only taking a year end survey; however, it is something that goes on for the whole year and each day at the job with an employee. This study will help managers and organizations to understand that how to make engagement the core of their corporate strategy to achieve its objectives and to meet the demands of diversified workforce in Pakistan in order to retain them. This has been backed by (MacLeod, 2009) that the outcomes of highly engaged employees include higher performance, lower level of intention to quit and sound business acumen and sense.
The current research will contribute in building the theoretical knowledge as it will incorporate new factors affecting engagement level of an employee as well as to be able to retain employees in organizations in a demanding environment. The generational gap within organizations makes it difficult for leaders to come up with strategies that will manage staff from all age brackets. Even many organizations in Pakistan are still unable to accept the fact that employee engagement and retention are and should be the part of overall HR KPI's and matrix as organizations performance is associated with both of these variables. Furthermore, this study will enable the HR and OD professionals to incorporate strategies that can build employees level of engagement. Further, certain competencies can be included to identify the individuals before hiring and to know if they are psychosocially positive or not. OD can work on cultural aspect by making culture more conducive for all new employees as well as old employees .

LITERATURE REVIEW AND HYPOTHESIS DEVELOPMENT
This has been argued by many scholars that engaged workers need to be psychologically strong to live a high quality life (Fredrickson, 2009) in terms of well being and CAW to attain personal and organizational goals (Richman, 2006). In the below section, we will review the study construct and subsequently drive the hypothesis.

Employee Engagement and Employee Retention
According to (Law, 1998) , engagement is a multidimensional construct. It is the expression of an individual's complete self (physical, emotional and cognitive) in their job roles (Kahn, 1990). It consists of three major facets that are: (1) Vigor: which is the work related energy and resilience that an individual possesses. Engaged employees are flexible when hard times arise, they can easily cope up with pressure and know how to use their time and energy to complete each task. Employees high in vigor are good team player and support to other team members when required. With their stimulating and brisk attitude, they do every task with immense zeal (i.e. physical and behavioral engagement).
(2) Dedication: engaged employees are highly involved in their task, and have considerate quest of work. They feel that their work is extremely imperative for the organization and its success. These workers feel pride in the work they do and have strong job identity and job significance. These are also aware about how their job will affect other jobs in the organizations and they know their responsibilities and duties fully. They know how to get a task completed; they put their full energy and potential in any new task assigned to them.
(3) Absorption: it is the degree to which the employees are entirely alert and immersed in their task, which they find captivating and appealing. These employees are fully concentrated and happy while performing work; it is that state whereby the time of an employee passes very rapidly and one finds it hard to detached oneself from assigned work. With the presence of these three constructs, it is believed that the employee is engaged. According to (Harter, 2002), engaged employees have a tendency to demonstrate positive organizational outcomes which includes customer satisfaction, efficiency, increased profits, and lower turnover intentions (Harter, 2002), higher organization commitment (Schaufeli W. a., 2004) and work-family satisfaction (Bakker, 2008). (Maslach, 2001), analyzes why employee feel burn out by taking a contemporary approach to which he said that the employee being engaged at their work is the antidote of burnout. Studies on the construct job engagement to date have tried to identify the key drivers of engagement. The Study of (Schaufeli W. a., 2004) discovers that there are job resources that act as job motivators, which brings engagement and through which positive attitude, high productivity and good mental health is depict in engaged employees. These empirical results suggest that engaged employees are psychologically strong and vice versa. They are resilient in hard times and they have control over their work with career satisfaction, and this may lead to low turnover intention. According to (Mitchell, 2001), when an employee thinks to quit any organization due to poorly designed work polices, role conflicts, least motivation, the concentration of same work, lack of growth opportunities, lack of training and development, issues with management or any other voluntary reason, is called turnover intention. Reports suggest that engaged and committed employees can generate many benefits for the organization such as lower absenteeism and reduce turnover (de Lange, 2008). Due to extreme competition from the market, companies need to come up with different strategies to retain employees; otherwise organizations will be prone to many issues. Retention strategies referred by different scholars comprise financial reward, job enlargement, training and development opportunities, work life balance, and supportive work environment (Ghosh, 2013). Other scholars present retention strategies as employee engagement (Palmer, 2012), career development (Agarwala, 2003) and learning attitudes (Kyndt, 2009). As per former literature, some of the reasons why an employee aims to leave an organization are the lack of vigor, dedication and absorption in one's work-life. Hence, the below hypotheses can be stated:

2.2.PsyCap
Organizations in current times are trying to increase psychological capital beyond financial capital (what you have), human capital (what you know), and social capital (who you know) (Luthans F. a., 2004). It is broadly defined as a creature's positive state of development related to "who you are" and "what you can become" (Luthans F. Y., 2007). It consists of four dimensions: (a) Selfefficacy: take on a challenging task and implementing it through enough self confidence. It is basically the control of oneself on its behaviors, level of motivation and social environment. Belief of an individual in his or her capability to show behaviors that can produce specific performance attainment. Believing in one self to do certain tough and challenging task and completing those tasks efficiently. (b) Optimism: it is about having an optimistic view point on being victorious now and in the near future, being positive about changes taking place at the workplace and in life and taking all the changes optimistically and believing that success will be achieved in near future. (c) Hope: representing determination and elasticity in achieving goals, employees aiming high about the future and expecting certain things to happen for them like achieving their assigned objectives. (d) Resilience: being persistent in the face of difficulty and adversity, facing the difficult time with courage and determination and being cheerful during tough time considering that it is also the part of learning and shall pass too. These four dimensions buildup the psychological capital of individual. Individuals with high level of psychological capital are very important for organizations as the work pressure and work demands are increasing each day. Today, many companies are opting towards lean management style, restructuring the entire layer of management, and bringing an overall organizational change. Employees with high level of PysCap can be able to deal with such drastic changes in the workplace. (Luthans F. Y., 2007) study further suggests that PsyCap is the higher order and comprehensive structure for understanding and capitalizing on human assets in the organization. Many related factors that influence the level of efficacy, hope, resilience and optimism. The author further claims in his study that employees with a high level of PsyCap are more motivated and positive at work fundamentally. Work stress is managed in a much better way by a positive employee than those employees who frequently feel low and down (Gupta M. a., 2017). Furthermore, due to certain accomplishments at work, an engaged worker may experience higher level of satisfaction and feel happy, which can ultimately leads to the state of GWB (Tinline, 2010). These achievements makes employee feel more certain at work and refill their ability of persistence and resilience. The employees are more optimistic now about the future. Hence, it can be said that because of involvement and engagement in work, employee experience success and positive feeling (Bakker, 2008). This will improve their special resources such as efficacy, hope, optimism and resilience (PsyCap). Further, these special resources help individual employees to assist in managing stress and pressure at work (i.e.CAW) and keep them contented and motivated in work (i.e. GWB) (Luthans K. L., 2016). These workers are then satisfied with their work (JS) and the intention to leave the organization also decreases. Hence, the following hypotheses are proposed:

General Well Being
An effect-orientated evaluation of wellbeing is happiness that entails predominance of positive affect over negative affect (Diener E. , 1984). In his study, (Christopher, 1999) mentioned that a person with more positive feelings than negative, is apparent like doing well (happy), or experiencing well (i.e. individuals appraisal) and possessing higher emotional intelligence. Health, steady employment, wealth (income), encouraging relations and happy recent life events (child birth, marriage, buying house) are associated with higher GWB (Weimann, 2015). GWB means that people will have high levels of positive emotions and are satisfied with life in general. A person with a high level of GWB would be inclined by his professional experience and will be independent of his work situation to a certain extent. (Zelenski, 2008) argued in his study of happy productive workers that happy people are indeed more productive. According to (Diener E. , 1984) and (Erdagon, 2012) SWB involves two components: affective balance (comparisons between the level of positive and negative effects in someone's life) and views about life satisfaction.GWB is an emerging topic in the management area as today's organizations are more versatile and work hours have been increased with job complexity. GWB bring positive outcomes in employees work and personal life. Individual with a high level of GWB are supposed to be retained in the organizations for longer period of time as they are generally happy with the surrounding and whatever is going on in their life. Consequently, they take changes at workplace as positive changes. Also, they are more productive and may possess high level of PsyCap. In the current study we suppose that sense of wellbeing is not rigid and can be influenced by individual's intentions and related factor. We are examining GWB as an outcome of EE and PsyCap and in turn seeing its effect on ER. Hence, we have proposed the following hypotheses:

Control at Work
The extent to which an individual can exert control over their decisions at work is CAW. It is basically an individual's capability to control what happens in his or her work, mainly to influence matters that are important to his or her personal goals. In addition to this, it can also be called as autonomy in the work place, whereby an employee can have the freedom to determine what he or she can do at work and how can he or she do that work. In today's era, employees prefer jobs that provide them autonomy to make certain decisions along with responsibilities. People prefer not to be dictated but rather be encouraged in what they do by themselves, which also give rise to many positive feelings. In his study, According to (Weigl, 2010) engaged employees can easily implement control over their work by job enlargement and by creating additional demanding responsibilities for them. Higher CAW motivates employees to achieve predefined goals. Job related results are affected by the extent to which an individual thinks positively. Researchers think that there are jobs which are critical and require greater focus and control like in health care industry (Gupta V. A., 2016). Literature suggests that CAW is a potential consequence of EE and PysCap. EE fosters CAW (Torp, 2013). There has been a positive relationship found between these two variables. Engaged employees are cognitive vigilant and high on energy which enables them to control their work. Hence, we propose the following hypotheses:

Job Satisfaction
Job Satisfaction (JS) is a measure of career success, that means positive work related and psychological outcomes that are achieved by an individual as a result of work related experiences (Judge, 1995). It is basically the response of an employee toward his/her work experience. (Gruneberg, 1979) said that it is an emotional response towards their job, as to how they feel about their job and work conditions (Bushra, 2012). It is one of the important factors leading to success in the organization. In one study by (Ng, 2005), it was classified that the predictors of career success are organizational sponsorship (support from supervisor, learning and growth opportunities and organization size), human assets, steady individual differences and socio demographic variables. It is also obvious that satisfied employees show a positive attitude towards the organization, while dissatisfied shows opposite. It is observed in the study by (Lee. H. Y., 2009) that levels of dissatisfaction in jobs such as low motivation, absenteeism, grievance and high turnover are all effected by job satisfaction. Work satisfaction plays its role in bringing positive behavior among employees and hence increasing employee retention. In his study, (Lounsbury, 2003) in his study that assertiveness, carefulness, emotional toughness and extraversion were absolutely linked with career satisfaction. Hence, we suggest the following hypotheses:

H12. JS will positively mediate the relationship between employee engagement and employee retention H13. JS will positively mediate the relationship between PsyCap and employee retention
Research theoretical model of the study is illustrated in figure (1) which clarifies the relationship between study variables.

Sample and Procedure
The sample consists of employees working in different companies within Pakistan. The questionnaire survey was used to gather data from respondents. The questionnaires were distributed to 250 employees, while 230 were returned, out of which 200 usable responses were considered as the final sample for the study, giving a response rate of about 80 percent. The sample in this research includes 166 employees from different service companies, while 34 were from different manufacturing companies. The demographic variables include gender, marital status, age, education level and working experience. Most respondents were male (62 percent), among which (60 percent) were in their 30's, and (23 percent) in their 40's. The education level of respondents includes, 57 percent graduated from a university. The length of employment at the current employer is between one to five years (45.5 percent), between six to ten years (18 percent) and between eleven to fifteen years (11.5 percent). The age of the respondents ranges from 20 to 60 years and work experience varied from 1 to 16 years and above. In summary, most respondents were educated males working for less than five years in the service industry. Summary of the demographic profile is presented in table 1.

Values
Response  Y., 2007). Sample items include "I always look on the brighter side of things at work" and "I feel confident contributing to discussions about the company strategy". Both CAW and GWB were measured using (Easton, 2013), CAW, 6 items, (Cronbach's α = 0.845) : GWB, 3 items, (Cronbach's α = 0.843), One of the Sample item for CAW is "I am able to voice opinions and influence changes in my area of work" and a sample item for GWB is"In most ways my life is close to ideal". JS scale developed by (Stringer, 2006) was used for JS, a sample item is "I have the chance to work alone at my job" and Employee Retention was measured using developed scale of (Seashore, 1982) (Cronbach's α = 0.886) and a sample item includes "I will not change this organization easily"

Descriptive Statistic& Analysis
The mean of the variables mentioned in Table II below ranges from 2.995 to 3.68 and SD from 0.99 to 1.262. Similarly, the mean of seven items of the variable EE range from 3.17 to 3.655, whereas SD ranges from 0.993 to 1.144. The outer loading range from 0.539 to 0.820. Outer loading above 0.7 is significant.
In variable EE, outer loading of two statements are below 0.7 and the remaining five are above 0.7, which are all significant. The t statics range from 7.866 to 33.136.The t statics is the value of differentiation represented in units of standard error, the higher the t statistic value, the greater the evidence against null the hypothesis. All the P values are 0.000 for the items of EE which is quite significant. A low p-value (normally ≤ 0.05) indicates strong proof against the null hypothesis.
Next, the mean of ten items of the variable PsyCap ranges from 2.955 to 3.68 whereas SD ranges from 1.041 to 1.115. The outer loading range from 0.526 to 0.800, the outer loading of two statements is below 0.7 and the remaining eight are above 0.7 which are all significant. The t statics range from 7.639 to 20.081. All the P values are 0.000 for the items of PsyCap which are quite significant. Similarly, the mean of three items of the variable CAW ranges from 3.23 to 3.44 whereas SD ranges from 0.999 to 1.083. The outer loading range from 0.836 to 0.920, which are all significant. The t statics ranges from 33.958 to 70.697. All the P values are 0.000 for the items of CAW which are quite significant.
Furthermore, the mean of ten items of the variable GWB ranges from 3.245 to 3.51 whereas SD ranges from 0.99 to 1.238. The outer loading range from 0.0006 to 0.892, the outer loading of one statement is below 0.7 and the remaining five are above 0.7, which are all significant. The t statics range from 0.052 to 61.907. All the P values are 0.000 for the items of GWB except for item six which is 0. 958. In addition to this, the mean of eight items of the variable JS range from 3.08 to 3.54 whereas SD ranges from 0.994 to 1.262. The outer loading range from 0.620 to 0.844, the outer loading of two statements are below 0.7 and remaining six are above 0.7 which are all significant. The t statics range from 8.793 to 24.435. All the P values are 0.000 for the items of JS which are quite significant. Lastly, the mean of three items of the variable ER ranges from 3.08 to 3.19 whereas the SD ranges from 1.154 to 1.181. The outer loading ranges from 0.882 to 0.917. The t statics ranges from 45.612 to 66.614. All the P values are 0.000 for the items of ER which are quite significant. (Refer table II).

Measurement validation
In the initial stage, the relationship of all variables in study with each other was identified through their coefficient correlation. All the major variables are positively correlating with each other as values are near to 1. (Refer table III).

DATA ANALYSIS AND RESULTS
In this part, the fallout from Partial Least Square Analysis (PLS), reliability and correlation are reported. To examine hypothesis, (SEQ) structural equation modeling is used, which is considered to be one of the strong methods to examine measurement and structural models. The second step, includes the measurement model, which is the projected framework, and the variables validity (which includes the reliability, convergent and discriminant validity) of all variables were measured with the help of confirmatory factor analysis (CFA) using PLS Smart 3.0. The study is completed in two parts; in the first step CFA was calculated for individual variables to check its factor structure. The outcome from this step inveterate that the EE and PsyCap are multidimensional constructs, that have four and three primary factors as mentioned earlier in the literature (Schaufeli W. T., 2006) (Luthans F. Y., 2007). Respecification of the model was not required to improve its fit and the measurement model had a good fit to the data. CFA showed that CAW, GWB, JS and ER are unidimensional constructs.
In the next step of the CFA analysis, all the latent variables were co-varied with each other in the overall measurement model (i.e. proposed theoretical model).
To examine construct validity, composite reliability was calculated and all the values were above 0.8, equal to or greater than .80, which is considered good for confirmatory research (Daskalakis, 2008). CR is the most vigorous determinant of reliability as contrast to Cornbach's α (Petersonandkim, 2013). The results of the scale are reliable in the current study. Furthermore, the average variance extracted (AVE) was also premeditated to ascertain convergent reliability.

Discriminant validity
Discriminate validity is defined as any single construct when differs from other constructs in the model which are meant to be related in a particular study (Carmine, 1979). Discriminate validity results are acceptable when the variables have an AVE loading more than 0.5 which indicates that a minimum 50% of the variance was taken by the construct (Chin, 1998). Discriminate validity is recognized if the elements, which are in diagonal, are considerably higher than those values, which are in off-diagonal in the parallel rows and columns. Discriminant Validity analysis are being conducted to examine whether non-related ideas or measurements are in fact unrelated or not. A useful estimation of discriminant legitimacy demonstrates that a trial of an idea isn't exceptionally associated with different tests intended to quantify hypothetically various ideas. As per the result of the current study, the top numbers (which are the square root of AVE) in each factor column are almost higher than the values (correlations) below it, so there exists a disriminat validity. However, the correlation value of JS and PsyCap where EE is at the top and correlations of PsyCap with GWB on top are higher than AVE values. (Refer Table V

Model fit measures
The fitness of the model in SEM-PLS is explained by various measures such as standardized root-meansquare residual (SRMR), and the exact model fits like d_ULS and d_G, Normed Fit Index (NFI), and χ2 (Chi-square). The model fit measures consisting of the measured value of both saturated model as well as the estimated model is reported in the table present below. The saturated model assesses the correlation between all constructs. The estimated model, on the other hand, takes model structure into account and is based on the total effect scheme. The measurement model has a good fit with the data (SRMR is 0.064, NFI is 0.793). The lower the SRMR the better the fit. By rule, a model has a good fit when SRMR is less than 0.8 (Hu, 1998

Assessment of hypothesized Model
All study hypotheses were supportive, presenting statistically positive path coefficients, except GWB on ER, (T > 1.96 and p < 0.05). In size effect, EE has higher path coefficients for ER at 0.887 and GWB has a higher path coefficient for ER at 0.850. All the research hypothesized direct relationship was supportive except with a slight significant low effect of EE on ER at P values = 0.887 and GWB on ER at p value = 0.850. Likewise, with the former literature on work engagement, it was identified that EE has a positive effect on GWB (P = 0.048, p < 0.01) and on PsyCap (p = 0.000, p> 0.001). However, GWB has a negative path coefficient of 0.20 on ER in the current study and a p-value of 0.850, which is great than 0.50. (Refer  table VII).
Here, we accept H1 as EE has a moderately positive effect on ER at P value = 0.887. Engaged employees plan to stay in an organization more often than unengaged employees do as per (Konrad, 2006). Because of the increased cognitive, emotional and behavior engagement, employee have a high level of organizational satisfaction and lower turnover intentions. When employees are more attachment towards any organization they do not tend to quit easily.

Test of Mediation
To examine the mediation effect of PsyCap, CAW, GWB and JS, two models were compared, one of which includes mediating effect and another one depicts direct total effect among variables.
H2. PsyCap will positively mediate the relationship between employee engagement and CAW By including a direct path from EE to CAW, the first substitute model was examined for statistical analysis. Though not hypothesized, it is implicit that EE can certainly influence employees level of CAW as the correlation coefficient was strong at (r=0.666). The higher the engagement level at work the more control employees will have there. The path coefficient from EE to CAW in our analysis is 0.16, which is a highly positive correlation among these two variables. In the first alternative model, with PsyCap as a mediating variable the path coefficient is 0.504, t = 6.937, p = 0.000. The results of the mediation analysis discovered that PsyCap fully mediates the relationship between EE and CAW. EE has both direct and indirect effects on CAW. Here, we accept H2 as PsyCap positively mediates the relationship between EE and CAW. It is also supported by (Nguyen T. , 2012). Their research finding showed that psychologically competent individuals have personal assets and they do not depend on external resources. This advantage enables them to have more control over their work in hard times, which improves their engagement level.
H3. PsyCap and GWB will positively mediate the relationship between employee engagement and employee retention There exists a positive direct relationship between PsyCap and GWB. The level of hope, resilience, optimism and efficacy of workers would definitely and directly affect their overall life satisfaction. The direct path from PsyCap to GWB has a path coefficient = 0.658, t = 7.814 and p = 0.000 which is significant. However, GWB does not mediate the relationship between EE and ER and PsyCap and ER and shows a negative correlation as mentioned above. Hence, we fail to accept H3.
H4. PsyCap and JS will positively mediate the relationship between employee engagement and employee retention Adding a direct path from EE to ER, the path coefficient = 0.511, t = 8.587, p = 0.000 which is a significant positive relationship between these two variables. The direct path results from PsyCap to JS having the path coefficient = 0.496, t = 5.597 and p = 0.000. In the alternative model, with PsyCap and JS as mediating variables for EE and ER, the path coefficient = 0.258, t = 4.326, p = 0.000, the results of the mediation study discovered that PsyCap and JS positively mediates the relationship between EE and ER. Hence, we accept H4.
H5. PsyCap will positively mediate the relationship between employee engagement and GWB Adding a direct path from EE to GWB, the path coefficient = 0. 708, t = 17.396 and p = 0.000. In the alternative model, with PsyCap as the mediating variable, the path coefficient = 0.536, t = 7.485 and P = 0.000, the results of the mediation study identified that PsyCap positively mediates the relationship between EE and GWB. Thus the direct as well as in direct effect is significant and we accept H5.
H6. PsyCap will positively mediate the relationship between employee engagement and JS Adding a direct path from EE to JS, the path coefficient = 0. 742, t = 18.784 and p = 0.000. In the alternative model, with PsyCap as mediating variable the path coefficient = 0.404, t = 5.340 and P = 0.000, the results of the mediation study discovered that PsyCap positively mediates the relationship between EE and JS. Thus, the direct as well as in direct effect is significant. In sum, PsyCap played a essential role not only for EE and JS but for GWB as well. Hence, we accept H6.
H7. GWB will positively mediate the relationship between employee engagement and employee retention Adding a direct path from EE to ER, the path coefficient = 0.511, t = 8.587, p = 0.000, in the alternative model, with GWB as mediating variable, the path coefficient is -0.003, t = 0.175 and P = 0.861, the results of the mediation study identified that GWB negatively mediates the relationship between EE and ER. Thus the direct effect is significant and indirect is not. As mentioned above there exists a negative relation between GWB and ER as per statistical analysis. Hence, we fail to accept H7.It is not necessary that employees with high GWB stay in a similar organization for a longer period of time as millennials now a day's tend to leave every organization in two years for better prospects.
H8. GWB will positively mediate the relationship between PsyCap and employee retention By including a direct path from PsyCap to ER, the path coefficient = 0. 340, t = 3.800 p = 0.000. In the alternative model, with GWB as the mediating variable the path coefficient is -0.013, t = 0.183 and P = 0.855, the results of the mediation study revealed that GWB negatively mediates the relationship between PsyCap and ER. Thus the direct effect is significant and indirect is not. Hence, we fail to accept H8.
H9. CAW will positively mediate the relationship between employee engagement and employee retention Adding a direct path from EE to ER, the path coefficient = 0.511, t = 8.587, p = 0.000. In the second alternative model, with CAW as mediating variable the path coefficient is = 0.010, t = 0.591, p = 0.555, the results of the mediation study identified that CAW fully mediates the relationship between EE and ER.
Hence, it can be said that EE has both direct and indirect effects on ER. Here, we accept H9. (Kanten, 2012) suggested that EE has a positive major relationship with positive behavior at work as these employees are vigilant and energetic at work in a way that they manage their work and which in turn this leads to retention.
H10. CAW will positively mediate the relationship between PsyCap and employee retention By including a direct path from PsyCap to ER, the path coefficient = 0. 340, t = 3.800 p = 0.000. In the third alternative model, with CAW as mediating variable the path coefficient is 0.037, t = 0.591, p = 0.557, the results of the mediation analysis discovered that CAW fully mediates the relationship between PsyCap and ER. Here, we accept the H10. As per previous literature PsyCap positively influences CAW and in turn may improve employee retention.
H11. PsyCap and CAW will positively mediate the relationship between employee engagement and employee retention Adding a direct path from EE to ER, the path coefficient = 0.511, t = 8.587, p = 0.000 which is a significant positive relationship between these two variables. The direct path results from PsyCap to CAW has the path coefficient = 0.619, t = 7.706 and p = 0.000. In the alternative model, with Psycap and CAW as mediating variable the path coefficient = 0.030, t = 0.584, p = 0.560, the results of the mediation analysis discovered that PsyCap and CAW positively mediates the relationship between EE and ER. Hence, we accept H11.
H12. JS will positively mediate the relationship between employee engagement and employee retention Adding a direct path from EE to ER, the path coefficient = 0.511, t = 8.587, p = 0.000. In the alternative model, with JS as mediating variable the path coefficient = 0.215, t = 2.613 and p = 0.009, the results of the mediation analysis discovered that JS positively mediates the relationship between EE and ER. Hence it can be said that EE has both direct and indirect effects on ER through JS and we accept H12.The total effects including the direct and indirect (mediating) effects are summarized in table VIII. H13. JS will positively mediate the relationship between PsyCap and employee retention By including a direct path from PsyCap to ER, the path coefficient = 0. 340, t = 3.800 p = 0.000. In the alternative model, with JS as mediating variable the path coefficient is = 0.316, t = 4.502 and P = 0.000, the results of the mediation analysis discovered that JS positively mediates the relationship between PsyCap and ER. Thus, the direct as well as in direct effect is significant. Hence, we accept H13.

Research Findings
The results obtained fully support H1, H2, H4 H5, H6, H9, H10, H11, H12 and H13. In contrast, regarding H3, H7 and H8, the mediating relationships were negative and we are unable to accept these hypotheses. Considering mutually, the results of the current research provide empirical facts for the points from (Bakker, 2008) that personal resources in the form of PsyCap can optimistically relate to the level of work engagement. The research validates the study in a way that it supported the idea of engaged employees being easily able to be retained, Furthermore employees with a high level of PsyCap does posses CAW and JS that again affects the turnover intentions of employees. The result further supports the study of (Kanten, 2012) EE has a constructive major relationship with positive behavior at work as workers with high level of engagement are quite alert and energetic at work. They control their work, which in turn leads to retention of these employees because they know how to control a stressful situation and pressure at work and how to be positive about the new changes at work life. It is not necessary that employee with a high level of GWB intend to stay in a single company for a longer period of time, GWB does not guarantee employee loyalty to one company which is also backed by the result of the study that GWB does not positively mediates the relationship between EE and ER and between PsyCap and ER. The study further validates the findings of (Xanthopoulou, 2009) about positive relation between EE and PsyCap. Engaged employee's posses' positive feelings about their work and are emotionally attached towards their work which increases their self efficacy, hope, optimism and resilience. Also, it supports the study by (Avey J. W., 2008), which empirically depicts the direct relationship between PsyCap and EE. Workers that have a high level of self efficacy and optimism with demanding goals are expected to be extra internally motivated and engaged in their jobs.
However, a small number of studies examined the relationship between PsyCap, CAW, GWB and JS. The outcomes of the current study further extended the findings of (Nguyen T. , 2012) that PsyCap has a positive relationship with workers behaviors by giving insights into CAW and GWB. Several studies found the positive relationship between PsyCap and GWB, however, these both together did not mediate positively between EE and ER as per our result. These results are in accordance with the available literature discussed above. Engaged individuals are entirely concerned and passionate about their jobs and organizations and they are willing to be able to add more towards organization's success. Furthermore, the results also suggest that the employees with elevated intensity of optimism and self efficacy along with demanding objectives are fundamentally motivated and emotionally, cognitively and physically engaged in their jobs. Moreover, they, demonstrate characters that are not only related to fulfillment, which results productively, but also career satisfaction. The essential contribution of this research lies in establishing the mediating role of PsyCap, CAW, GWB and JS between EE and ER which was never examined earlier. In that case, all variables are positively mediating except GWB between EE and ER and Between PsyCap and ER.

Theoretical Contribution
This research is linked to the rising constructs in HR, OB, and Psychology. This study examined five positive organizational behaviors simultaneously. One of the contributions of this study lies in identifying the fact that EE and PsyCap are antecedents of JS. It means that the employees with a high level of efficacy, resilience, hope and optimism tend to have high level of EE, which results into work satisfaction and happiness and high retention. This study will guide in further extension of the research on the same topic in the future. This research can be helpful for managers and organizations to understand how to come up and devise strategies that can help to increase employee engagement and why change is required to devise policies on work life balance, new task assignment, training and development.

CONCLUSION
In this research, we tried to understand the relationship between employee engagement and employee retention through the mediating role of psychological capital, control at work, general well being and job satisfaction. The result of the study showed a positive relationship among variables except for GWB's roles between EE and ER and PsyCap and ER. The role of employee engagement cannot be neglected by any organizations in today's era, where employee's demands keep changing and they need to be engaged in order to be retained happy. Several other factors impact the level of engagement and retention in the organizations.
This research further improved the findings of (M. Gupta, 2018) , as it has been revealed that PsyCap fully mediate between EE and GWB as per results. Whereas, interpretations of (M. Gupta,2018) for finding the relationship among WE, GWB and CAW were based on the correlation values. This study goes a step ahead and provides the extent to which WE is associated to GWB and CAW. PsyCap has been found impacting notably on both GWB and CAW.

Limitation and future direction
The present study is not exclusive of limitations. This study relied on self reported answers by employees who volunteered to contribute, so common method bias is anticipated. Though, all variables can be measured by individual perceptions. Furthermore, focusing the study on Pakistan, the sample of the study is likely to be limited with certain groups with similar demographic characteristics.