An Exploration of Supervisory Satisfaction as a Mediator Variable in Organisational Citizenship Behaviour

This paper seeks to explore the mediating construct of supervisory satisfaction on the relationship between leadership styles and organisational citizenship behaviour in companies across Malaysia. While there have been past research that investigated the relationship between leadership styles and citizenship behaviour, until now, there is very little research investigating the mediating effects of supervisory satisfaction on such relationship especially in the Malaysian context. Data was collected from 280 respondents that represent major industries such as services, manufacturing, mining and construction. SmartPLS was used to test the proposed framework. The results show that the transformational leadership style has significant positive relationship with subordinates’ organisational citizenship behaviour. This result found that supervisory satisfaction mediates the relationship between transformational leadership and citizenship behaviour. The conclusion of the research emphasizes the effects of the research for future research and practice.


Introduction and Objective of the Study
The purpose of this research is to determine the transformational leadership styles on subordinates' organisational citizenship behaviour (OCB).Supervisory satisfaction is explored closely as a mediator on OCB.This research rationale is to explore the relationship between these variables and not on specific organisational citizenship problems that Malaysian organisations are facing.Extensive literature has shown that leadership styles has the greatest impact on subordinate's response to work condition, thus, it is an important predictor.Although leadership style and organisational citizenship behaviour has been examined in the past (Wang, Law, Hackett, Wang & Chen, 2005;Ehrhart, 2004;MacKenzie, Podsakoff & Rich, 2001;Schlechter & Engelbrecht, 2006;Boerner, Eisenbeiss, & Griesser, 2007;Bass, Avolio, Jung, & Berson, 2003;Podsakoff, MacKenzie, & Bommer, 1996;Goodwin, Wofford, & Whittington, 2001), there has been little research done in determining the role of supervisory satisfaction as a mediator.
Malaysia is the choice of location for the study due to the shortage of empirical work which explores supervisory satisfaction on the relationship between leadership styles and OCB.This research attempts to add knowledge in the Malaysian context with regards to the literature surrounding organisational behavior.
This study seeks to answer the following research questions: • What is the direct and indirect effect of leadership styles on subordinates' organisational citizenship behaviour?

•
How does supervisory satisfaction mediate the relationship between leadership styles and organisational citizenship behaviour?

Theoretical Framework and Literature Review
An examination of the literature was done to outline and elaborate the interactions among the three variables: leadership styles, supervisory satisfaction and OCB.

Leadership Styles
Past researches have extensively studied transactional leadership as the core component of effective leadership behaviour in organisations.This was prior to the introduction of transformational leadership theory into the literature (Bass, 1985;Burns, 1978;House, 1977).Transactional leadership is based on exchange relationship where subordinates agreed with, accepted, or complied with the superior in exchange for rewards, resources or the avoidance of disciplinary action (Podsakoff, Todor & Skov, 1982;Podsakoff, MacKenzie, Moorman, & Fetter, 1990).More recently, much empirical work has focused on transformational leadership, in particular on the extent to which transformational leadership augments the effect of transactional leadership in explaining various outcomes such as leader effectiveness (Hater & Bass, 1988), subordinate satisfaction (Seltzer & Bass, 1990) and subordinate effort (Bass, 1985).These earlier studies are of particular relevant for this research as the existing literature suggests that (1) transactional and transformational leaders elicit different patterns of follower conformity (Kelman, 1958;Howell, 1988;Tichy & Devanna, 1986).This emergent genre of leadership study advocates that transformational leaders can motivate followers to perform beyond the normal call of duty.

Supervisory Satisfaction
Supervisory satisfaction is a collection of feelings or affective reactions of the organisational members which are associated with their immediate manager supervision.The aspect of supervising style is simply the amount of supervision and direction given to the subordinates.These being: how closely their job activities are structured, monitored, and directed.Justification for the need to investigate supervisory satisfaction is exemplified in the seemingly observed relationship between the levels of job dissatisfaction, absenteeism, grievance expression, tardiness, low morale and high turnover.Dissatisfaction with supervision is an antecedent of intention to leave the workplace.Unsatisfied workers will leave their jobs more than their satisfied colleagues (Gangadhraiah, Nardev, & Reddy, 1990;Martin, 1990;Padilla-Vellez, 1993).

Transformational Leadership Styles and OCB
Many studies have shown that transformational leadership can affect followers' OCB (Geyer & Steyrer, 1998;Howell & Avolio, 1993;Lowe, Kroeck & Sivasubramaniam, 1996;Podsakoff et al., 1990).Graham (1988) has suggested that the most important effect of transformational leadership behaviour is the ability to promote extra-role behaviours that exceed the requirements of in-role expectations.Podsakoff et al. (1990) support Graham's view and argues further that "the most important effects of transformational leaders should be on extra-role performance, rather than in-role performance" (p.109).Such leadership is proposed to "lift ordinary people to extraordinary heights" (Boal & Bryson, 1988, p. 11) and to cause followers to "do more than they are expected to do" (Yukl 1989, p. 272) and "perform beyond the level of expectations" (Bass, 1985, p. 32), andHouse, Landis &Umberson (1988, p. 100) claim that these leaders motivate their subordinates to "perform above and beyond the call of duty".In sum, this suggests that transformational leadership may have an important effect on extra-role or OCB that is of a discretionary nature which are not part of the employee's formal role requirements.
Hypothesis 1: Transformational leadership style is positively correlated with OCB.

Transformational Leadership Styles and Supervisory Satisfaction
Among the determinants of job satisfaction, leadership is viewed as an important predictor and plays a central role.Leadership is mostly directed towards people and social interaction, as well as to the process of influencing people so that they will achieve the goals of the organisation (Skansi, 2000).The link between transformational leadership and work-related attitudes and behaviours such as job satisfaction, is well established (Avolio, Bass, Walumbwa, & Zhu, 2004).
Under the circumstances when job satisfaction is examined using Bass's (1985) leadership model, it is suggested that managers who display more transformational leadership characteristics might intrinsically foster higher job satisfaction.In a similar fashion, the transformational leader is able to motivate his or her followers to take on more responsibility by granting them work autonomy.As such, work tasks provide employees with an increased level of accomplishment and satisfaction.Additionally, the transformational leader's ability in grooming the individual personal development of followers projects a sense of belonging among subordinates as they observe that someone is concerned for their needs.Hence, the transformational leader-employee interactions may be more balanced since the manager and satisfied employees both jointly and effectively work toward achieving the organisation's common goals.
Essentially, both empirical and meta-analytic studies suggest that followers working with transformational leaders are more involved, satisfied, empowered, motivated and committed to their organizations, and demonstrate fewer withdrawal behaviours (Barling, Weber & Kelloway, 1996;Bono & Judge, 2003;Walumbwa & Lawler, 2003;Walumbwa, Wang, Lawler, & Shi, 2004).Such leaders enhance followers' confidence, effectiveness and motivation by giving them personal attention and by learning their needs and aspirations (Walumbwa & Kuchinke, 1999).Thus, based on the foregoing discussion, the following hypothesis is put forward: Hypothesis 2: Transformational leadership and subordinates' supervisory satisfaction will be positively correlated
A meta-analytic review by Petty and his colleagues (1984) indicated that job satisfaction and performance have a positive relationship.These conclusions were supported by Judge, Thoresen, Bono & Patton (2001), who found that earlier studies sometimes were too restrictive in their conceptualizations of both job satisfaction and performance.Concerning the direction of the satisfaction-performance relationship, it was proposed that positive emotions such as a feeling of satisfaction would bring about higher performance, especially when performance is being associated with the outcome of the participant's extra role behaviour.Therefore, the next hypothesis is formulated as: Hypothesis 3: Subordinates' organisational citizenship behaviour will be increased as supervisory satisfaction increases.

Supervisory Satisfaction as Mediator
Job satisfaction is seen in some studies as a mediator to the effect of various antecedents on career commitment (Rhodes & Doering, 1983;Malik, Waheed, & Ur-Rehman Malik, 2010;Steers & Mowday, 1981;Oliver & Brief, 1977;Farrel & Rusbult, 1981;Yousef, 2002a,b;Chen, Myrtle, Lin, & Fahey, 2011;Malik et al., 2010;Cheung, Wa, Chan, & Wong, 2009).Farrell and Rusbult (1981) commented that job satisfaction is an intervening variable to be a function of rewards and costs associated with a specific job where job commitment is an outcome from job satisfaction, investments and alternatives.Other studies have also found that job satisfaction does not mediate the effects of other antecedents on commitment; rather these antecedents influence career commitment directly (Goulet & Singh, 2002).
The argument could be made that supervisory satisfaction as a mediating variable occurs because supervisory satisfaction represents a situational factor that is within a leader's sphere of influence.When supervisory satisfaction increases, the leader becomes more instrumental.Conversely, as supervisory satisfaction decreases, the leader becomes less instrumental.Thus, one could argue that the link between leadership and OCB would be stronger when supervisory satisfaction is high and weaker when supervisory satisfaction is low.On top of that, the transformational leader is said to be able to increase supervisory satisfaction and supervisory satisfaction has been found to be positively related to extra-role performance (Churchill, Ford, Hartley, & Walker, 1985;Jaworski & Kohli, 1991).Based on this rationale, the following is expected.
Hypothesis 4: Supervisory satisfaction mediates the relationship between transformational leadership style and organisational citizenship behaviour.

Sampling Design
The sample selection for this research includes executives, managers and professionals in companies that deal with services, manufacturing, mining and construction.These companies are mainly located in the Klang Valley, Malaysia.Companies that met the criteria above were carefully chosen from the Federation of Manufacturers, Malaysia, Service Directory, Construction Industry Development Board Directory, and Malaysian Trade and Commerce Directory published in 2008.

Leadership Styles
Six dimensions of the Transformational Leadership Behaviour Inventory (TLI: Podsakoff, MacKenzie, Morrman & Fetter, 1990) were used to measure the 14 items of leadership style.A 7-item scale was used to assess the transactional leadership from Leader Reward and Punishment Behaviour Questionnaire (LRPQ: Podsakoff, Todor, Grover, & Huber, 1984;Podsakoff, Todor, & Skov, 1982).To measure subordinates' perception of the superior leadership styles the instrument uses a 7-point Likert scale.

Organisational Citizenship Behaviour (OCB)
The research uses Smith, Organ and Near's (1983) 7-item scale to measure OCB.In particular the scale measures the dimension of OCB connected to altruism.This instrument uses 7-point Likert scale with ranges from 1 as strongly disagree to 7 as strongly agree as incumbent rated.

Supervisory Satisfaction
This study will focus on one facet of job satisfaction called satisfaction with supervision developed by Hackman and Oldman (1980).The satisfaction consists of three items assessing satisfaction with supervision, method of supervision and quality of supervision.The internal consistency reliability for this scale as reported by Warren (1998) is 0.93

Data Analysis Procedure
The SmartPLS was the primary statistical technique used and correlational analysis is the other statistical analysis employed.

Research Results
Compared to the covariance-based structural equation modelling technique, PLS provides a better explanation of variables relationships and it can accommodate a minimal sample size and sample distribution (Fornell & Larcker, 1981;Chin, 1998).The PLS analysis was conducted using the SmartPLS 2.0 software (Ringle, Wende & Will, 2005).

Measurement Reliability and Validation
The construct reliability, convergent validity, discriminant validity, and nomological validity test were applied to all constructs.The psychometric properties of all latent constructs are presented in Table 1.As seen in Table 1, the standardized loadings (λ) of each reflective measurement item on its corresponding latent construct was greater than the threshold levels of 0.50, were all significant at p < 0.01, and therefore convergent validity was achieved (Hair et al., 2010).The Cronbach's alpha values (α) and composite reliability (ρ) for all latent constructs were above 0.70, supporting adequate reliability of the multi-item scales (Bagozzi & Youjae, 1988).All average variance extracted (AVE) were greater than 0.50 threshold level, supporting convergent validity of the latent constructs (Fornell & Larcker, 1981;Chin, 1998;Hair, Black, Babin, & Andeson, 2010).As seen in Table 2, the discriminant validity was supported as the AVE for each construct was greater than the squared correlation with other constructs (Fornell andLarcker, 1981, Malhotra, 2010).The nomological validity was assessed using the measure of explained variance.The R 2 value of supervisory satisfaction and organizational citizenship behaviours were 0.63 and 0.55 respectively, which were large effect sizes (~26%; Cohen, 1988).The goodness-of-fit (GOF), which was calculated as the geometric mean of average R 2 (0.59) and average communality (0.65) was 0.62, and suggested satisfactory fit to the model (Tenehaus et al., 2005).

Structural Model
Blindfold approach was carried out to evaluate the predictive validity of the structural model.The means of the cross-validated (CV) communality (H 2 ) and redundancy (F 2 ) indicate that the quality of structural model is well above the recommended standard of 0.30 (Tenenhaus et al., 2005).In fact, the average CV-communality (H 2 ) was 0.65 and average CV-redundancy (F 2 ) was 0.49.The average variance accounted (AVA) for the model was 0.35, which was greater than the cut-off value of 0.10 (Falk & Miller, 1992).These results suggest that the model exhibits adequate explanatory power.
Table 3 presents the estimated path coefficients, t-values, and variance explained (R 2 ) of each endogenous construct.The results confirm the research hypotheses.As can be seen from Table 3, transformational leadership has a significant positive effect on supervisory satisfaction (β = 0.79, t = 30.33,p < 0.01).This provides support for H3.Similarly, supervisory satisfaction was found to have a positive effect on organizational citizenship behaviour (β = 0.61, t = 8.54, p < 0.01).However, the direct effect of transformational leadership on organizational citizenship behaviour was not significant (β = 0.15, t = 1.84, p = 0.07).

Mediation Analysis
A mediation analysis was carried out to examine the mediating role of supervisory satisfaction in the relationship between transformational leadership and organizational citizenship behaviour using Preacher and Hayes (2008) procedure.The latent variable scores obtained using SmartPLS was analysed using SPSS routine developed by Preacher and Hayes (2008) for generating the 95 percentile confidence intervals.Table 4 presents the results of the mediation analysis.As can be seen from the results, the total effect of transformational leadership on organizational citizenship behaviour is 0.64.However, the direct effect of transformational leadership on organizational citizenship behaviour is not significant (β = 0.15, t = 1.84, p = 0.07) and the total indirect effect is 0.49.These findings suggest full mediation of supervisory satisfaction in the relationship between transformational leadership and organizational citizenship behaviour.Note: *p < 0.05.Standardized beta coefficients are presented with biased corrected 95% confidence intervals.

Discussion
Hypothesis H1 predicts that transformational leadership style is positively correlated with organisational citizenship behaviour (Table 2: r = 0.64, p < 0.01).This result is quite similar to past studied (Schlechter & Engelbrecht, 2006;Ferres, Travaglione & Connell, 2002;Mackenzie, Podsakoff & Rich, 2001) that unambiguously indicate that there is a positive relationship between transformational leadership and OCB.The relationships between leadership and OCB have been empirically studied with the conclusion that transformational leadership was consistently linked to followers' higher level of OCB (Goodwin, Wofford, & Whittington, 2001;Wang, Law, Hackett, Wang, & Chen, 2005;Mackenzie, Podsakoff, & Rich, 2001).Hence, there is strong conceptual support for the notion that transformational leaders motivate their followers to exhibit extra-role behaviours.
Hypothesis H2 predicts the relationship between transformational leadership and subordinates' supervisory satisfaction will be postively correlated (Table 2: r = 0.69, p < 0.01) lend support to Hypothesis H2.This result reinforces the general speculation that transformational leadership produces desirable work-related attitudes and behaviours in the subordinates such as job satisfaction and organisational commitment (Avolio, Zhu, Koh, & Bhatia, 2004;Avolio, Bass, Walumbwa, & Zhu, 2004).Transformational leaders are also known for their ability to recognise, understand and attempt to address each follower's needs on a one-on-one basis (Bass, 1998) and continue to raise those needs to higher levels of inspiration (Avolio, 1999).Subordinates who perceive they receive the leader's special attention are more likely to exhibit greater satisfaction with supervision (Bartram & Casimir, 2007;Berson & Linton, 2005;Emery & Barker, 2007).
Hypothesis H3 posits that subordinates' organisational citizenship behaviour will be increased as supervisory satisfaction increases (Table 2: r = 0.73, p < 0.01) lend support for Hypothesis H3.Thus, the result seems to agree with the conventional reasoning that supervisory satisfaction provides the opportunity of a person to perform effectively and efficiently.Other research findings (Judge et al., 2001), have also demonstrated a positive relationship between supervisory satisfaction and performance.There is clearly more empirical evidence to suggest the simple conclusion that the supervisory satisfaction ensures better performance (including extra-role) as one is more certain about what is expected to be accomplished.Hopefully, the inccrease in supervisory satisfaction will lead to an increase in organisational commitment and task performance and ultimately, increase the employees' OCB level (MacKenzie, Podsakoff, & Ahearne, 1998;MacKenzie, et al., 2001;Tubre & Collins, 2000).
Hypothesis H4 states that supervisory satisfaction mediates the relationship between transformational leadership style and organisational citizenship behaviour as shown in Table 4.The result asserts that transformational leadership has a significant and direct relationship with subordinates' OCB and that this relationship is mediated by supervisory satisfaction.In relation to the direction of effect, it would appear that supervisory satisfaction serves to increase the strength of the relationship between transformational leadership and OCB.This can be explained from the fact that transformational leadership will generally increase the supervisory satisfaction and this will lead to an increased in OCB (MacKenzie, Podsakoff, & Rich, 2001).Transformational leaders by their attempt to provide followers with a clearer perspective on their work may actually increase employees' supervisory satisfaction.The present result seems to acknowledge the contention made by several researchers that leadership style effectiveness is very much dependent on the supervisory satisfaction (Yousef, 2002a,b;Jaworski & Kohli, 1991).

Conclusion
This research sought to determine how a superior is able to attain a better understanding of the appropriate leadership styles when faced with supervisory satisfaction.As predicted the superior is incline to adopt a transformational leadership style as compared to a transactional leadership style when greater subordinate OCB is pursued.It is well suggested that the superior must adopt the appropriate leadership style to ensure that subordinates surpass their call of duties.The supervisory satisfaction as a mediating effect was purposely researched so as to investigate how the variable is able to strengthen or weaken the relationship between leadership styles and OCB.OCB is intended as an outcome variable in this research so as to solicit subordinate superior effort in the organisation.The presence of such a correlation shows the significance of leadership style in promoting subordinates' OCB in organisation.Such behaviour should have great practical significance and thus should be promoted in organisation.The research also supports the contention that subordinates' supervisory satisfaction has a direct and indirect effect on the outcome, serving to increase the strength of transformational leadership on OCB.This result seems to support a widely held assertion that the effectiveness of leadership depends very much on the situation (Fiedler, 1967;House & Dessler, 1974;Weed, Mitchell, & Moffitt, 1976).In short, this research provides the theoretical grounding for exploring a greater variety of outcomes vis-à-vis managers' leadership styles and their effect via supervisory satisfaction.The present research could assist to foster further research efforts to apply the findings of organisational research to industrial settings.

Managerial Implications
From the present research there are numerous managerial implications.Such as, in terms of a practical standpoint, the research findings recommend that given a choice of leadership styles, a superior should show greater importance in adopting a transformational leadership style so as to achieve greater OCB.By adopting the transformational leadership style the harmful influencing network created is changed by shifting the superior-subordinate power difference.As such, it is important to foster organisational cultures that are transformationally-oriented through the encouragement of job and organisational design, human capital decisions and training and development.The behaviour of the transformational leader is fostered by training through mentoring and recognising the ever-changing development needs of the employees.Rather than promoting the win-lose relationship which are inherently damaging the intellectual stimulation of the transformational leadership style found in the integrative problem-solving relationship should be promoted.Subordinated can learn to adopt the transformational leadership style through the learning of scenarios such as role play and even videotapes of actual cases in organisation.With the appropriate feedback, work productivity will increase.Likewise, organisations that are facing rapid environmental change will profit at all levels from the flexibility adopted by transformational leadership.With the help of supervisory satisfaction as mediator the variable is able to uncover how leadership styles provide indirect effects on its outcome.As such, the transformational leader deals with supervisory satisfaction so as to induce higher OCB levels.

Contribution and Direction for Future Research
This research contributes to the existing literature of leadership and OCB research in several ways.Firstly, the model offers a base for incorporating a large and diverse body of research.By appealing to the managerial leadership, the model is able to provide the theoretical grounding for research to distinguish the effectiveness of different leadership styles and OCB.In line with past research study (Podsakoff, Mackenzie, Paine, & Bachrach, 2000;Mackenzie, Podsakoff, & Rich, 2001;Wang, Law, Hackett, Wang, & Chen, 2005;Walumbwa, Wu, & Orwa, 2008), this research concurs that subordinates tendency to adopt a transformational leadership style as compared to a transactional leader in engaging OCB is more interesting and encouraging.In future, research should attempt to include other endogenous variables such as motivation, compliance and performance of subordinate that are more indicative of organisational outcomes.

Table 1 .
Psychometric properties of the measures

Table 2 .
Correlation and discriminant validity Note: Diagonal values represent square root of average variance extracted.*p < 0.01

Table 4 .
Mediation analysis results