The Role of Psychological Engagement in Relationship between Perceived Organizational Support and Withdrawal Behavior and Intentions: An Empirical Study on Small Industries in Egypt

The study of employees' perceptions is gained much importance to determine employee’s behavior at work place.The purpose of this study is to test the direct impact of perceived organizational support and psychologicalengagement on both withdrawal behavior and intentions, and investigates the mediating role of psychologicalengagement in these relationships. Sample data were gathered from 467 subordinates and their supervisors, from20 small factories located in New Damietta industrial zone in Egypt. The study depends on hierarchical multipleregression analysis. The results showed significant negative impact of perceived organizational support andpsychological engagement on both withdrawal behavior and intentions. Furthermore, psychological engagementis partially mediated the relationship between perceived organizational support and both withdrawal behaviorand intentions. The results, limitations of this study, and future research opportunities were discussed.


Introduction
Organizational support researches began by noting that what matters for managers and organizations is employees' commitment with organization, while what matters for employees is organization's commitment with them (Eisenberger, Huntington, Hutchison, & Sowa, 1986;Eisenberger, Fasolo, & Davis-LaMastro, 1990).It is seem as a vicious circle.Allen and Meyer (1990: p. 1) defined employees' commitment, as an affective variable, as "employees' emotional attachment to, identification with, and involvement in, the organization." But from viewpoint of Eisenberger, et al., (1986;1990) commitment is a bound feeling emotionally and intellectually for some actions of one party to another.Therefore, they believe in commitment is phenomenon happen in shape of reciprocity relationship, employees' commitment with organization and organization's commitment with employees.They named employees' perceptions about organization's commitment with them, Perceived Organizational Support (POS).Eisenberger and his colleagues defined POS as employees' beliefs about the extent to which organization values their contributions and cares about their well-being (Eisenberger et al., 1986;Rhoades & Eisenberger, 2002;Eder & Eisenberger, 2008).They believed in organization should start by supporting their employees then wait repay positive outcome from them.
Employees seek a balance in exchange relationship with organization (Wayne, Shore & Liden, 1997) throw develop their perception about organization.Employees' perception is one of the most important determinants of the employees' attitudes and behavior.This is due to; employees use their perception as tool to measure the balance in their relationship with organization (Ali, Rehman, Haq, Jam, Ghafoor, & Azeem, 2010).
On the other hand, the engagement as a term beset a lot of ambiguity and lack of agreement among researchers (Macey & Schneider, 2008).From viewpoint of burnout researchers (e.g.Maslach & Leiter, 1997) engagement, as an opposite pole of burnout, is a state characterized by energy, involvement, and professional efficacy; the direct opposite of the three burnout dimensions: exhaustion, cynicism, and inefficacy.From another viewpoint, Schaufeli and Bakker (2004: p. 295) defined engagement as "a positive, fulfilling, work-related state of mind that is characterized by vigor, dedication, and absorption".They showed that engagement is an insistent and extended affective-cognitive state and is not focused on any particular object, event, individual, or behavior.
Engagement is a psychological concept because it refers to the allocation of personal resources and also to how densely and continuously those resources are applied to work roles (Rich, Lepine, & Crawford, 2010).In this context, Psychological Engagement (PE) was defined by Kahn (1990) as the harnessing of an employee's energies on the levels of physical, cognitive, and emotional, to perform his work roles.The cognitive aspect is employees' beliefs about the organization, its leaders and working conditions; the emotional aspect is employees' attitudes toward each of those three factors, and the physical aspect is the energies exerted to accomplish their roles (Kahn, 1990).This study will adopt engagement as psychological state, for many reasons, remove ambiguity about engagement and distinguish it about many behavioral variables such as innovative, initiative, extra-role, and organizational citizenship behavior (e.g.Macey & Schneider, 2008;Young, 2011); and distinguish it about many attitudinal variables such as satisfaction, affective commitment, and job involvement (e.g.Mauno, Kinnunen, & Ruokolainen, 2007;Macey, & Schneider, 2008;Kular, Gatenby, Rees, Soane, & Truss, 2008); and to test whether engagement will account for the relation between POS and both Withdrawal Behavior (WB) and Withdrawal Intentions (WI).
Social Exchange Theory (SET) argues that obligations are generated because of a state of interdependence relationship between parties (Sack, 2006).This relationship evolves over time into trusting, loyal, and commitment as long as the parties respect reciprocity norm (Cropanzano & Mictchell, 2005).Reciprocity norm means repayment rules such that the actions of one party lead to a response by the other party (Cropanzano & Mictchell, 2005).So that, SET purported that relationship between employees and their organization based on norm of reciprocity, which means sharing of positive outputs; organizations can realize its goals and favorable outcomes through the generous treatment of their employees (Aselage & Eisenberger, 2003).Where, employees trade effort and loyalty to their organization for such tangible benefits as pay and such socioemotional benefits as esteem, approval, and caring (Armeli, Eisenberger, Fasolo, & Lynch, 1998).
Several empirical studies, which investigated the relationship between POS and its consequences, indicated that high levels of POS improved positive employees' attitudes and behavior.Eisenberger, et al., (1990) showed that high levels of POS, in occupational environment, associated with high levels of job performance and job attendance, and in manufacturing environment, also POS realized positively relationship to affective attachment, performance outcome, and provide constructive suggestions to help the organization.Furthermore, in governmental environment, POS realized positively relationship to organizational citizenship behavior (Asgari, Silong, Ahmad, & Samah, 2008).
In addition to the above, Wayne, et al., (1997) indicated to positive relationship between POS and both affective commitment and organizational citizenship behavior in American corporation.Also, Eisenberger, et al., (2001) demonstrated positive relationship between POS and both in-role performance and affective commitment.Furthermore, POS realized positively relationship to job satisfaction and reduce the impact of work condition on job satisfaction (Eisenberger, Cummings, Armeli & Lynch, 1997).Eisenberger, Armeli, Rexwinkel, Lynch, & Rhoades, (2001) indicated that when employees perceive they are valued and supported by their organization, they would reciprocate by being more involved in their work roles.This is due to perception of employees will create obligation feelings (Wayne et al., 1997).These feelings make employees not only feel an obligation to return organization's commitment in shape of employees' commitment, but also make employees engage in behavior which support organizational goals (Wayne et al., 1997;Rhoades, Eisenberger, & Armeli, 2001).In this context, employees who are perceived high levels of organizational support are more likely engaged to their jobs.
From another side, several empirical studies indicated that high levels of POS reduced negative employees' attitudes and behavior.For example, in a manufacturing organization, POS eliminated the relationship between coworker WB and employee tardiness.In retail sales organization, POS reduced the relation between work team withdrawal and employee withdrawal (Eder & Eisenberger, 2008).Another example, POS realized negative relationship with turnover intentions, and organizational commitment partially mediated this relationship (Tumwesigye, 2010).In this context, employees who are perceived high levels of organizational support are more likely reduced their WB and WI.
On the other hand, there is a belief among practitioners and academics that the existence of positive consequences of engagement (Saks, 2006).Kahn, (1992) proposed that high levels of engagement lead to positive outcomes.The meta-analysis conducted by Harter, Schmidt, and Hayes (2002) observed positive outcomes of employee engagement for both employees and organizations.For example, Bakker, Demerouti, and Brummelhuis (2012) demonstrated that engagement realized positively relationship with task performance, contextual performance, and active learning.Henryhand, (2009) showed that positive impact of engagement on job satisfaction, and negative impact on intent to leave.Moreover, Saks, (2006) indicated to positive impact of engagement on job satisfaction, organizational commitment, and organizational citizenship behavior; and negative impact on intentions to quit.Therefore, there is a reason to expect that PE is more likely reduced WB and WI.
SET focused on the importance of understanding employees' motivation to the achievement of organizational goals (Justin & Eisenberger, 2003).This is indicator to try understanding mechanisms of employer-employee relationship.Therefore, Maslach, Schaufelli, and Leiter (2001) suggested that treatment of engagement as a mediating variable for the relationship between the work conditions and work outcomes.Chung and Angeline (2010) showed that engagement partially mediated the relationship between availability of job resources and job performance.
Within the above context, the purpose of this article is to test the following model, as shown in figure 1.

Psychological Engagement Mediator Variable
Perceived Organizational Support Independent Variable Withdrawal Behavior and Intentions Dependent Variable Figure 1.Study model

Methodology
The methodology section will discuss the sample of study, and its specifications.Also, the measures of study variables will explain.Then, study measures will test by principal component factor analysis and Cronbach's alpha to demonstrate their validity and reliability.Finally, descriptive and analysis method will discuss.

Sample
Sample data were gathered from subordinates and their supervisors from 20 small factories located in New-Damietta industrial zone in Egypt, during the period from September to November 2012.These factories represent various sectors which include: furniture, marble, metal, paper, and plastic industries.Of the 800 distributed questionnaires, 544 questionnaires were returned.A number of questionnaires were excluded due to some missing data.The final sample size was 467 supervisor-subordinate matched responses thereby reducing the response rate to 58 percent.The study sample was classified according to the respondent's gender into: 71.5 percent of male and 28.5 percent of female; according to respondent's tenure into: 22.7 percent less than or equal 5 years, 38.5 percent more than 5 years and less than or equal 10 years, and 38.8 percent more than 10 years tenure; and according to respondent's level of education into: 38.8 percent less than college degree, 55.5 percent college degree, and 5.8 percent more than college degree.

Measures
The study used short version scale which developed by Eisenberger et al., (1997) to measure of POS This scale consists of 8-items which attained high loaded factor among 36-items scale developed by Eisenberger et al., (1986).As shown in table 1, the study used a principal components analysis to determine whether the 8-items comprised a unitary dimension or multiple dimensions and to test whether factor loaded for each item exceeds 0.4 or not.The results suggested a single factor and all factor loading exceed 0.4.According to Hinkin (1995), each item added value to the scale.A principal components analysis was an essential step because of translation of original scale to Arabic form.Then, the study tested scale reliability using Cronbach's alpha coefficient.The results demonstrated that alpha coefficient of POS was 0.83.According to Nunnaly and Bernstein (1994), POS scale was reliable.To measure PE, the study used 7-items scale which developed by Young (2011).As shown in table 2, the results suggested a single factor and all factor loading exceed 0.4.Therefore, all items of scale were standardized.The results of Cronbach's alpha demonstrated that alpha coefficient of PE was 0.89.Therefore, PE scale was reliable.
Table 2. Factor analysis and reliability of psychological engagement scale To measure WB, the study used 4-items scale which used by Eisenberger, et al., (2001), as shown in table 3, the results of principal components analysis indicated to a single factor and all factor loadings exceed 0.4.Thus, all loadings are standardized.The results of reliability test demonstrated that alpha coefficient of WB was 0.80.Therefore, the scale of WB was reliable.Furthermore, to measure WI, the study used 3-items scale which used by Abraham, (2005).As showed in table 3, the results of principal components analysis suggested a single factor and all factors loading exceed 0.4.Thus, all loadings are standardized.The results of reliability test demonstrated that alpha coefficient of WI was 0.72.Therefore, the scale of WI was reliable.
Table 3. Factor analysis and reliability of withdrawal behavior and withdrawal intensions scale The study used self-reports to evaluate POS, PE, and WI, and used supervisor reported to evaluate WB of subordinates.All scales evaluated on 5-point scale, (1=strongly disagree, 5=strongly agree).

Analysis Methods
The study used mean, standard deviation, and correlation coefficient to descriptive sample data.The study also depends on hierarchical multiple regression analysis to test study hypotheses.As well as, the study was used Baron and Kenny approach to test the mediating role.According to Baron and Kenny (1986), the three regression models are employed to evaluate the mediating role of PE.The first regression model is used to investigate the effect of POS as independent variable on the PE as a mediator variable.The second regression model is used to investigate the effect of POS on both WB and WI as dependant variables.Then, to test the mediating role of PE, in the third regression model PE was inserted in the second regression model.In the third regression model, if the effect of POS on both WB and WI becomes insignificant, this will indicate to complete mediation, but if effect becomes weaker, this will indicate to partial mediation.

Results
The means, standard deviations and correlation matrix among study variables showed in table 4. The results of correlation analysis indicated to POS realized positive and significant relationship with PE.Also, it is realized negative and significant relationship with both WB and WI.Furthermore, the PE realized negative and significant relationship with both WB and WI.Regression analysis was used to test the direct impact of POS on both WB and WI, and to test the direct impact of PE on both WB and WI.The three steps of regression analysis proposed by Baron and Kenny (1986), was used to test mediating role of PE in the relationship between POS and both WB and WI.
By included PE in the regression model of POS on WB, β of POS changed from -0.49 to -0.28, but still significant.Thus, the relationship between POS and WB became weaker.Therefore, PE partially mediated the relationship between POS and WB.Also, by inserted PE in the regression model of POS on WI, β of POS changed from -0.34 to -0.25, but still significant.Therefore, PE partially mediated the relationship between POS and WI.

Discussion
Consistent with prior research, the results of this study indicated to significant negative impact of POS on both WB (e.g.Wanye et al., 1997;Liu, 2004;Lew, 2009), and WI (e.g.Eisenberger et al., 2001;Eder & Eisenberger, 2008).Results also demonstrated positive impact of POS on PE.Therefore, the study results supported results of literature review throw add new evidence from Arabic country and from small industries sector.Those results can be interpreted in the light of the norm of reciprocity.
The norm of reciprocity means obligating the reciprocation of favorable treatment between the organization and employees (Aselage & Eisenberger, 2003).In this context, organizational support theory focused on monitoring outcomes of organizational support and tries to be interpreted these outcomes in the light of employee-employer relationship (Eisenberger, et al., 1986).Therefore, employees received intangible benefits such as caring and respect; in addition to tangible benefits such as wages and medical care from their organization, most likely to respond to the organization in the form of positive outcomes or reducing the negative outcomes such as WB and WI.
The results were discussed, until now, demonstrate that when employees perceived high levels of organizational support, they will reduce their WB and WI; but it not answer about, how did this happen?The mediating test tries to interpret the relationship between organizational support and its outcomes, throw the employee-state-employer relationships.Consistent with prior research (e.g.Ram & Prabhakar, 2011), the results showed that PE partially mediated the relationship between POS and both WB and WI.These results indicated to the operation of multiple mediating factors.According to Baron and Kenny (1986) it's a normal result because most areas of psychology, including social, treat phenomena that have multiple causes.
Young (2011) pointed out engagement as psychological state make employees believe in their work, proud of their roles, feel enthusiastic about daily work, feel meaningfulness of their work roles, feel satisfaction about their roles, feel energized when do their roles, and value their daily accomplishments.These features can be justified the mediated role of PE.Therefore, the study results confirm that employees seek a balance in exchange relationship with their organization (Wayne et al., 1997), this balance happens throw create a psychological state commensurate with their perception about the degree of organizational support.
The main limitation of this study is its cross-sectional nature.Another limitation is using self-reports to measure POS as a predictor variable and WI as a criterion variable.The partially mediated role of PE indicates to another mediating variables intervene the relationship between POS and both WB and WI.So that, the study proposes investigate the mediating roles of another psychological variables, such as empowerment and/or involvement, in these relationships.

Conclusion
The study of employees' perceptions is gained much importance to determine employee's behaviors in the work place (Ali et al., 2010).The results suggested the following: negative impact of POS on both WB and WI; negative impact of PE on both WB and WI; Positive impact of POS on PE; and PE partially mediated the relationship between POS and both WB and WI.Social Exchange and organizational support theories provide a meaningful theoretical basis to interpret these results.Therefore, the study recommends managers and organizations to support their employees throw practices such as: care about their opinion, care about their well-being, consider their objectives, introduce help if they have a problem, forgive a trivial mistakes, and value their contributions.These practices will back to the organization in shape of positive outcomes.
the work I do every day makes a difference.0.69 I get a feeling of personal satisfaction from the work I do at my factory.0.78 When I am performing my role, I feel energized. (R) 0.81 When I finished a day's work, I feel like I have accomplished something worthwhile.0.80 searching for an alternative to the factory.0.74 As soon as it is possible, I will leave the factory.0.58 R : Reverse item.

Table 1 .
Factor analysis and reliability of perceived organizational support scale R: Reverse Item.

Table 5 .
A test of the mediated role of psychological engagement in the relationship between perceived organizational support and both Withdrawal behavior and intensions