The Mediating Role of Organizational-Based Self-Esteem in Perceived Organizational Support and Counterproductive Work Behaviour Relationship

This study examined the direct influence of the perceived organizational support (POS) on counterproductive work behaviour (CWB). This study also analyzed the role of organizational-based self-esteem (OBSE) as a mediator to the above relationship. Data were obtained from 660 public service officers serving in the Malaysian ministries. The regression analysis showed that the POS and OBSE have a significant negative influence on both CWB-organizational (CWB-O) and CWB-interpersonal (CWB-I). The bootstrapping analysis carried out confirmed the role of OBSE as a mediator in the relationship between POS and CWB-O and CWB-I. From the theoretical implications, this study explains the role of OBSE as an intervening mechanism in linking the POS and CWB. The practical implication suggests the importance of enhancing POS and employees’ OBSE in the effort to reduce employees’ involvement in CWB.


Introduction
Counterproductive work behaviour (CWB) is an act that can harm or intended to harm an organization and their stakeholders such as co-workers, supervisors, clients and customers (Spector & Fox, 2005).CWB can also be defined as a voluntary behaviour that violates significant organizational norms, and in doing so, threatens the well-being of an organization, its members, or both (Robinson & Bennet, 1995).Generally, CWB can be categorized according to the target, specifically CWB-O directed towards the organization (e.g.sabotage, stealing, withdrawal, and production deviant) and CWB-I directed towards other people (interpersonal conflict, and abuse) (Spector et al., 2006).CWB often happens in any work organization (Lawrence & Robinson, 2007).For instance, it is estimated that around 33% to 75% employees are involved in various forms of CWB like vandalism, sabotage, absenteeism, and theft (Harper, 1990).The National Business Ethics Survey (2013, p. 41) performed on 6,579 respondents in the USA found that among the forms of CWB that are often reported include abusive behaviour, lying to employees, conflicts of interest, violating company policies, and discriminating against employees.
Employees' engagement in CWB has a very serious impact on the organization and other employees' well-being.The impact of CWB to the organization, for example, can be seen in financial loss amounting to billions of US dollars annually (Coffin, 2003;Greenberg, 1997;Chao et al., 2011), leading to the decline of productivity and work performance (Lawrence & Robinson, 2007;Hershcovis & Barling, 2010).Further, employees who are the victims of CWB are more likely to be involved in turn-over, has low morale, damaged self-esteem, increased fear and insecurity at work, and psychological and physical pain (Henle et al., 2005).CWB also affects interpersonal relationships, (Estes & Wang, 2008) and the quality of group experiences such as teamwork (Aube et al., 2011).The serious negative impact of CWB has resulted in an increasing level of interest among researchers from the fields of industrial and organizational psychology, and management, more specifically to study the contributing factors of CWB.
(OBSE) in contributing to CWB.This study gives a significant contribution to the literature of these two aspects.Firstly, a large portion of knowledge regarding CWB and the contributing factors are based on the western literature especially from North America (Smithikrai, 2008;Nasurdin et al., 2014).Therefore, this study contributes to the CWB literature in different contexts, especially in the eastern collectivistic cultures.Furthermore, this is the first research which studies the relationship between POS, OBSE, and CWB in the Malaysian context.Secondly, previous studies only evaluated the direct influence of POS on CWB (for instance Liao et al., 2004;Colbert et al., 2004;Liu & Ding, 2012).In this vein, the understanding of the process of how POS contributes to CWB is limited (Ferris et al., 2009).In relation to this, the current study puts forth the mediating role of OBSE in linking POS and CWB.Although Ferris et al., did study the mediating role of OBSE in the relationship between POS and CWB, it was restricted only to CWB-O.In this study, we extend their study by highlighting the mediating influence of OBSE in the relationship between POS and CWB-I.

Perceived Organizational Support and Counterproductive Work Behaviour
Perceived organizational support (POS) is the degree to which employees believe that their organization values their contributions and cares about their well-being and fulfils socioemotional needs (Eisenberger et al., 1986).The role of POS in influencing employees' behaviour can be understood by three processes (Baran et al., 2012).Firstly, based on norm of reciprocity, the relationship between the organization and employees should be reciprocal.When an organization displays high support, then employees feel obliged to return the kindness by demonstrating positive attitude and behaviour.Conversely, when the POS is seen to be low, employees have the tendency to reciprocate in forms of negative attitude and behaviour.Secondly, POS fulfils the employees' socio-emotional needs such as the needs for esteem, approval, and affiliation that lead to attachment to the organization and the formation of social identity.Other than that, the fulfilment of socio-emotional aspect also helps to mitigate stress at the workplace and enhance the well-being of each individual employee.Thirdly, POS also demonstrates the readiness of the organization in rewarding the efforts done by the employees.The three processes, as aforementioned are supported by previous studies.For instance, previous studies found that POS enhances the sense of felt obligation, trust in organization, job involvement, organizational commitment, organizational citizenship work behaviour, and duty orientation (Kurtessis et al., 2015;Karavardar, 2014;Baran et al., 2012;AlKerdawy, 2014).
Although previous studies tend to study the association between POS and positive outcomes, it is also linked with negative attitudes and behaviour.Rhoades and Eisenberger (2002) in their meta-analysis, for instance, have highlighted that POS has a negative relationship with strain, withdrawal, intention to leave, and turn-over behaviour.A study by Shusha (2013) also found that POS negatively contributes to both withdrawal behaviour and intentions.In the relationship between POS and CWB specifically, although previous studies are rather limited, there are evidence that both variables correlate with one another in a negative relationship (Jacobs et al., 2013;Liu & Ding, 2012;Akremi et al., 2010;Liao et al., 2004;Colbert et al., 2004).Based on these preceding studies, in this study, the following hypothesis is established: Hypothesis 1: POS has a negative influence on CWB-O and CWB-I.

Organizational-Based Self-Esteem and Counterproductive Work Behaviour
OBSE is defined as the degree to what extent an individual believes in himself or herself to be capable, significant, and worthy as a member of an organization (Pierce & Gardner, 2004).Literature has suggested that OBSE plays a significant role in influencing one's attitude and behaviour at the workplace (Bowling et al., 2010;Ferris et al., 2009;Pan, Qin & Gao 2014).This influence can be understood through the self-consistency theory (Pierce et al., 1989;Pierce & Gardner, 2004).According to this theory, individuals who possess high OBSE often involve themselves with positive behaviour that is consistent with his or her own positive evaluation of himself or herself.In turn, individuals with low OBSE have the tendency to engage in negative behaviour that relates to the image of himself or herself.The role of OBSE in determining the employee work-related attitudes and behaviours is supported by a number of previous studies.For instance, Bowling et al. (2010) in their meta-analysis study, found that OBSE has a positive relationship with job satisfaction, organizational commitment, job involvement, in role job performance, and organizational citizenship behaviour.Their study indicated that OBSE also negatively relates to turn-over intention, depression, and physical symptoms.Closest to this study, Ferris et al. (2009) found that OBSE contributes negatively to the CWB-O.However, unlike the study by Ferris et al., this study extends their conceptual model by also examining the influence of OBSE on CWB-I.Thus, the following hypothesis is established: Hypothesis 2: OBSE has a negative influence on CWB-O and CWB-I.

Perceived Organizational Support and Organizational-Based Self-Esteem
The belongingness theory states that one of the human needs is the desire to be accepted or to have a positive interpersonal relationship with the group (Ferris et al., 2009;Baumeister & Leary, 1995).One's satisfaction on the group acceptance can be measured through the level of individual's self-esteem, and the level of self-esteem depends on the acceptance or rejection by the group (Ferris et al., 2009).In an organizational context, the level of self-esteem can be determined by various interpersonal and management practice factors (Pierce & Gardner, 2004;Lee, 2003;Heck et al., 2005).One of the interpersonal factors that were identified which contribute to the level of self-esteem is social support, be it from the organization, supervisors, or co-workers (Lee, 2003;Pierce & Gardner, 2004;Ferris et al., 2009).POS in particular has the capability to influence the OBSE as it fulfils the employees' socio-emotional needs to be accepted and appreciated (Eisenberger et al., 2004;Rhoades & Eisenberger, 2002).A study by Lee and Peccei, (2007) found that POS has a positive impact on OBSE among the bank employees in South Korea.Meanwhile, Chen et al. (2005) also discovered that POS has a strong influence on OBSE among a sample of employees in China.Recently, Ferris et al. (2009) also highlighted that POS predicts the OBSE measured four weeks later.Based on previous findings, the following hypothesis is established: Hypothesis 3: POS has a positive influence on OBSE.

The Mediating Role Organizational-Based Self-Esteem
In this study, OBSE is assumed to have a role as the mediator in the relationship between POS and CWB.Previous studies have supported the role of OBSE as the mediator in the relationship between the organizational and behavioural factors (Pierce & Gardner, 2004).For instance, Heck et al., (2005) found that OBSE is an intervening mechanism between job satisfaction, affective commitment, procedural justice and leader-members exchange quality, and workplace complaints.The closest to this study is the study conducted by Ferris et al., (2009) who found that OBSE is the mediator of the relationship between POS and CWB-O.Nonetheless, this study expands the model of the study by Ferris et al., by testing the role of the mediator of OBSE in the relationship between POS and the dimension of the CWB-I that has not been studied yet.In this vein, the following hypothesis is formed: Hypothesis 4: OBSE mediates the relationships between POS and CWB-O and CWB-I.

Participants and Procedures
The study sample comprised of federal public service officers serving in the Malaysian ministries.Approximately 2400 sets of questionnaire were distributed to all ministries (24 ministries in total) located at the Federal Administrative Centre of Malaysia in Putrajaya.The distribution and administration of the questionnaires was assisted by the personnel in Human Resource Management Department.From this number, 781 sets of forms were returned with the response rate of 32.0%.However, 121 questionnaires were put aside as they were found to be incomplete, had poor quality responses, and the presence of outliers was detected.Finally, the actual number of questionnaires used for further analysis was 660.Going through the literature reviews regarding CWB, the response rate for the survey conducted in previous studies was often in low percentage between 20% and 50% (such as Bennet & Robinson, 2000;Sackett et al., 2006;Penney et al., 2011;Bolton et al., 2012).To test the non-response bias, the wave analysis method (Armstrong & Overton, 1977;Rogelberg & Stanton, 2007) was employed.Respondents were divided into two groups; those who returned the questionnaires two weeks earlier and those who did so after two weeks.The demographic mean score (age and tenure of service) and the study variable mean (POS, OBSE, and CWB) between both groups were compared using the t-test.The results of the t-test illustrated that there was no difference in the demographic and the studied variables mean between the groups tested, and this suggests that there is no occurrence of non-response bias.
The study sample was heterogeneous encompassing various levels of positions, education, gender, tenure of service, and organizational functions.In terms of position, 62.4% were support staff and the remaining 37.6% were management and professionals.Most of the participants were female (64.5%).The average age of respondents was 36 years old (standard deviation = 9.10), whereas the average tenure of service was 11.7 years (standard deviation = 9.57).Most of the respondents (64.20%) have a university degree.

Measures
The measurement for this study utilized a questionnaire that was adapted from previous studies.POS was measured using 8 items from the Perceived Organizational Support scale (Eisenberger et al., 1986).The scale assesses employees' perceptions that their organizations value their contribution and care about their well-being.Sample items are "The organization values my contribution to its well-being" and "The organization really cares about my well-being".The responses were rated using a five-point Likert scale ranged from "1=strongly disagree" to "5 = strongly agree".High scores means high POS and vice versa.The Cronbach Alpha was .83.OBSE was measured using 10 items from the Organizational-based Self-Esteem Scale (Pierce et al., 1989).The scale specifically evaluates self-esteem in an organizational context, or how an individual sees himself or herself as capable, significant, successful, and worthwhile within their organization.Sample items are "I am important around here" and "I make a difference around here".The scoring was also carried out using the Likert scale ranged from "1=strongly disagree" to "5=strongly agree".High scores show high OBSE and vice versa.The Cronbach Alpha was .92.
The CWB measurement was based on the Counterproductive Work Behaviour Checklist (Spector et al., 2006).The checklists measure the involvement in volitional acts that harm or intend to harm the organization (CWB-O) or other person in the organization (CWB-I).Accordingly, 21 items assessed CWB-O, whereas 22 items assessed CWB-I.Sample items of CWB-O are "Purposely did your work incorrectly" and "Took supplies or tools home without permission".Meanwhile, sample items of CWB-I are "Made fun of someone's personal life" and "Refused to help someone at work".The response options ranged from "1=never" to "5=every day".High scores reflect high frequency of involvement in CWB and vice versa.The Cronbach Alpha for CWB-O dimension was .78;while for CWB-I was .61.The CWB-checklist are causal indicator scale or also known as formative scale (Spector et al. 2006;Edward & Bagozzi, 2000).For formative scale, the internal consistency is not a relevant indicator that reflects the reliability of the measurement (Diamantoplous & Winklhofer, 2001).Apart from that, the items in the formative scale also often do not correlate with one another or they might have a low correlation (McKenzie et al., 2011), and this leads to low internal consistency reliability.Previous studies have also demonstrated that the internal consistency reliability value for the CWB measurement was sometimes low (Gruys & Sackett, 2003;Spector et al. 2006;Bruursema et al., 2011).

Descriptive and Correlation Analyses
Table 1 shows the mean, standard deviation, and the correlation between the variables of the study.It was found that the POS has a significant correlation with CWB-O (r = -.28,p < .001)and CWB-I (r = -.29,p < .001).OBSE also has a significant correlation with CWB-O (r = -.26,p<.001) and CWB-I (r = -.17,p < .001).Finally, POS and OBSE showed a positive correlation (r = .37,p < .001).

Hypothesis Testing
The hypotheses for this study were tested using PROCESS for SPSS (Hayes, 2013).The regression coefficient generated by PROCESS came in the form of unstandardized coefficients.To get the standardized coefficients, the scores of all variables were converted into z-scores beforehand (Hayes, 2013;Tsang et al., 2014).The analysis results are shown in Figure 1.It was found that POS showed a negative influence on CWB-O (β=-.22,p<.01) and CWB-I (β=-.26,p<.01.OBSE also showed a negative influence on CWB-O (β=-.18,p<.01) CWB-I (β=-.08,p<.05).However POS showed a positive influence on OBSE (β=-.37,p<.01).The results fully support hypotheses 1, 2, and 3.The final hypothesis depends on whether OBSE mediates the relationship between POS, and CWB-O, and CWB-I.
For that purpose, the bootstrapping non-parametric analysis was conducted.The bootstrapping non-parametric analysis has its own strength in evaluating the mediating effects, as compared to other methods such as causal steps approach (Baron & Kenny, 1986) and the product-of-coefficients strategy (e.g Sobel test) as it does not require the normal distribution assumption in the data, and potentially minimize the Type I errors (Hayes, 2013;Hayes & Scharkow, 2013).The evidence of mediation is ascertained through the 95% bias-corrected bootstrap confidence interval with 10,000 samples which is when the indirect effects is said to be significant when the value between the lower limit confidence interval (LLCI) and upper limit confidence interval (ULCI) do not include zero (Hayes, 2013).
The results are shown in Table 2.The analysis showed that the indirect effect of POS on CWB-O through OBSE was significant, as revealed by 95% bias-corrected bootstrap confidence interval which has no zero (estimate = -.070;LLCI = -.101,ULCI = -.035).This finding suggested that OBSE acts as a mediator in the relationship between POS and CWB-O.Next, the analysis showed that the indirect effect of POS on CWB-I through OBSE was also significant (estimate = -.030;LLCI = -.020,ULCI = -.004;zero is not contained within the CI).This means that OBSE also plays a role as the mediator in the relationship between the POS and CWB-I.This result supports hypothesis 4.

Discussion
Other than looking into the direct influence of POS on CWB, this study also looked into the psychological mechanism which links the two variables.Specifically, this study puts forth the mediating role of OBSE in the relationship between POS and CWB.Consistent with Jacobs et al., (2013) and Liu and Ding (2012), this study discovered that POS has contributed negatively towards CWB-O and CWB-I.This means that POS reduces workers involvement in negative behaviours.This finding can be interpreted through the social exchange theory especially based on the norm of reciprocity (Gouldner, 1960).The norm of reciprocity model emphasized the exchange process between two parties namely the organization and employees, whereby when one party gives the benefit, the other party needs to reciprocate in return.Thus, in the context of this study, organizations that are perceived supportive in the form of appreciating and taking care of the well-being of its employees will produce positive reciprocity from the employees (for example, reducing the involvement in CWB).
Next, parallel with Ferris et al. (2009) this study also discovered that OBSE has a negative influence on CWB-O.This means that employees with higher OBSE will have less involvement in CWB-O.Further, the findings of this study also expanded the study by Ferris et al., where it revealed that the OBSE also has a significant relationship with CWB-I which has yet to be studied before.This result supports the self-consistency theory which states that individuals have the tendency to get involved in a certain behaviour that is consistent with their self-image.The negative relationship between OBSE and CWB suggested that individuals who have high level of self-esteem are less involved in CWB as behaviour is not consistent with the positive evaluation of oneself (Ferris, Brown, Lian, Keeping, 2009).
Consistent with the study by Lee and Peccei (2007), this study also found that POS has a positive influence on OBSE.These findings suggested POS is able to fulfil the employees' socio-emotional needs in being accepted and respected (Eisenberger et al., 2004).When one feels that he or she is accepted and respected, the individual will make a positive internalisation in relation to himself or herself such as feeling that he or she is appreciated and important (Lee & Peccei, 2007).Finally, consistent with Ferris et al. (2009), this study verified that OBSE serves as a psychological mechanism that links the POS with CWB-O.Ferris et al., in their study, proposed for future studies to test the mediating role of OBSE in the relationship between the POS and the CWB-I.As what has been suggested, this study found that OBSE also mediated the relationship between POS and CWB-I.In conclusion, this study found that supportive organization will increase OBSE, and subsequently reduces the involvement in CWB.

Implications, Limitations and Recomendations
This study carried several theoretical and practical implications.In terms of the theoretical implications, this study demonstrated that the theory constructed in the West (e.g.social exchange theory; self-consistency theory; belongingness theory) can be applied to explain employees' behaviour in various contexts (for example in eastern culture).This study also contributed to the knowledge of psychological mechanism that links the POS and workplace behaviour.For instance, other than confirming the mediating role of OBSE in the relationship between POS and CWB-O as studied by Ferris et al. (2009), this study also showed that the OBSE also plays a role as a mediator in the relationship between POS and CWB-I that has not been studied before.For practical implications, this study showed the importance of increasing POS to reduce employees' negative behaviour.Organizations must be creative in formulating the policies and programs which come in the form of appreciating the contributions of the employees and taking care of their well-being.To mitigate CWB, an organization also has to improve the organizational characteristics that can enhance OBSE, for instance such as job empowerment, providing feedback about performance, and coaching towards development and progress (Gardner & Pierce, 2013).
This study has several limitations that may influence the study's findings.Relying on self-report may lead to the occurrence of common method bias (Podsakoff et al., 2003).However, several measures have been adopted to reduce the possibility that this issue may arise.For instance, controlling the procedure in administrating the questionnaire has been prioritised (among which, is ensuring that aspects such as guarantee of confidentiality, the construct segregated into certain sections and arranged at random, and specific instructions given to different sections).A statistical evaluation using the Harman's single-factor test also becomes the indicator which proposes that the common method bias did not occur.There are scholars who opine that the common method bias is not a big issue in the self-report measurement (Spector, 2006;Spector & Brannick, 2009).Additionally, the meta-analysis study showed that self-reporting appeared to be an appropriate method to scrutinize CWB (Berry et al., 2012).However, future researchers who are interested in looking into employees' engagement in CWB are encouraged to obtain some kind of evaluation from various sources (for example colleagues, supervisor, internal customers, and disciplinary reports).Measurements that come from these various sources can consolidate the findings of this study.
As this study is cross-sectional, the link between variables cannot be interpreted as a cause and effect.There is a possibility that there would be an inverse relationship between the variables involved.However, the research framework has been formed through a thorough analysis of the related theories and literature.It is easy to be convinced that the relationship illustrated by the variables is in the right direction.Future researchers who are interested in confirming this study are recommended to use the longitudinal study, so that the results would be more reliable, especially in testing the mediating hypothesis (Balducci et al., 2011).Finally, moderate magnitude of the relationship between variables in this study suggests that future research should take a look into moderators that are potentially influencing the relationship.One of the variables that may influence the perception, motivation and orientation of behaviour at the workplace is personality (Spector, 2011).For instance, an individual with high internal locus of control may not be affected by the organizational factors and subsequently may be less tendency to commit CWB.

Table 1 .
Means, standard deviations and zero-order correlation

Table 2 .
Indirect effect of POS on CWB-O and CWB-I through OBSE