The influence of servant leadership on psychological empowerment and organisational citizenship on a sample of teachers

Orientation: The exhibition of organisation citizenship behaviour (OCB) by teachers is of great benefit for human capital development. Teachers’ perceptions of school principals as servant leaders play a critical role in their psychological empowerment and exhibition of OCBs.Research purpose: The goal of the study was to analyse the relationship between servant leadership, psychological empowerment and OCB of teachers in the South African school system.Motivation for the study: Teachers are usually under pressure to attend to the academic needs of the learners, which may be required either during normal working hours or after hours. This has necessitated the need to investigate the role played by a school principal’s servant leadership style in influencing the exhibition of OCBs.Research approach/design and method: A non-probability sample of 203 teachers drawn from selected schools in the Western Cape Province of South Africa was utilised. Item and dimensional analysis as well as structural equation modelling were performed on the data.Main findings: The constructs of servant leadership, psychological empowerment and OCB displayed high levels of internal consistency. Generally, a reasonable model fit was found for the structural and measurement models of the latent variables through confirmatory factor analysis and structural equation modelling. Positive relationships were found between servant leadership, psychological empowerment and OCB.Practical/managerial implications: A principal’s servant leadership approach is likely to promote teacher empowerment and development as it enables the principal to develop teachers to their fullest potential through mentoring and coaching. The teachers are likely to respond with OCB, which benefits the learners and society at large.Contribution/value-add: Schools face the challenge of motivating teachers. One of the ways of achieving this objective is by encouraging servant leadership amongst principals. Servant leadership promotes positive outcomes and extra-role behaviours.


Introduction
School principals and teachers play a pivotal role in imparting the important skills required for successful learning performance and further education and training (Mahembe & Engelbrecht, 2013;Mestry, 2017;Subedi & Shrestha, 2020). According to the Personnel Administrative Measures, section 4 of the Employment of Educators Act (1998), the teacher's core duties and responsibilities during and outside the formal school day include scheduled teaching time, relief teaching, extra and co-curricular duties, pastoral duties, administration, supervisory and management functions, professional duties, planning, preparation and evaluation. In addition to these teaching responsibilities, teachers take on various voluntary support roles such as offering additional classes, thus exhibiting organisational citizenship behaviour (OCB) (Collett, 2013).
Although the offering of extra classes is not formally recognised or remunerated as part of a teacher's core duties, it is a vital component of a teacher's job. The variance in learner performance at the school level is thus by implication directly related to the extent to which teachers exhibit organisational citizenship behaviour (Mahembe & Engelbrecht, 2014).
Various studies have reported that employees who exhibit OCBs are more satisfied and committed to their organisations (Aslam, 2012;Foote & Tang, 2008;Hamman-Fisher, 2008;Mohammad, Habib, & Alias, 2011;Organ & Ryan, 1995;Veloen, 2016). Teachers perceive extra role, unrewarded and non-obligatory tasks to have a strong influence on themselves, their students and on their school (Oplatka, 2009). It is, therefore, important to identify the behaviours that promote the enactment of OCBs. Principals, as leaders, can play a critical role in helping teachers to realise their potential and engage in extra-role behaviour (Cerit, 2009). One of the leadership approaches that are likely to affect school team effectiveness is servant leadership. Servant leadership unlike traditional leadership styles, which were largely mechanical, bureaucratic and dictatorial focuses on developing employees' potential using persuasive mapping and one-on-one coaching to achieve goals (Greenleaf, 1977). Principals are expected to act as role models that inspire teachers to have self-confidence and the ability to foster good long-term relationships with parents and other important stakeholders through respectful engagement (Graham, 1991;Liden, Wayne, Zhao, & Henderson, 2008;Van Dierendonck & Nuitjen, 2011). Robert Greenleaf, the father of servant leadership, defined a servant leader as one whose mind and behaviour is centred on the development and empowerment of followers, whilst encouraging them to act as servant leaders themselves (Greenleaf, 1997;Newman, Scharwz, Cooper, & Sendjaya, 2017). Therefore, the power of the servant leader lies in the leader's ability to unleash the potential and thus the power of those around them (Van Winkle, Allen, Devore, & Winston, 2014). In a school setting, the principal as a servant leader formulates a conducive environment where teachers collaboratively work towards a joint vision to ultimately improve the school environment and its functioning (Cerit, 2009;Du Four, 2001). Servant leaders motivate followers to go beyond their prescribed job requirements to exhibit OCBs (Ehrhart, 2004;Mahembe & Engelbrecht, 2014). By its very nature, servant leadership is empowering (Barbuto & Wheeler, 2002;Greenleaf, 1997). As the leader is available to nurture and coach his or her subordinates, the recipients of the nurturance become 'wiser' and psychologically empowered (Barbuto & Wheeler, 2002;Greenleaf, 1997). In addition, when employees feel that leaders care for their employees' development and growth, they feel psychologically empowered and are highly likely to engage in OCB.
Individuals who are empowered experience a sense of selfdetermination because they feel they are not micro-managed and can act autonomously (Quinn & Spreitzer, 1997). It is, therefore, proposed that leaders provide employees the opportunity to exercise their freedom to choose where and how they can contribute. Leaders (principals) are responsible for the empowerment of their employees (teachers) (Ehrhart, 2004). Leadership empowerment behaviours include providing work that are meaningful, fostering participation in decision-making, expressing confidence in high performance and encouraging autonomy (Quinn & Spreitzer, 1997).
Although there has been an increase in studies on servant leadership and OCB (Mahembe & Engelbrecht, 2014;Vondey, 2010), most of the studies are conducted in corporate environments with the exception of a study conducted by Mahembe and Engelbrecht (2014) in the education sector. This study investigated the role that servant leadership played in influencing OCB and team effectiveness. There is, however, a dearth in literature on the role played by servant leadership in influencing psychological empowerment and OCB amongst teachers in the South African education sector.

Aim of study
The main goal of this research was to study the relationship between servant leadership, psychological empowerment and OCB in the South African school system at selected schools in the Western Cape Province. The secondary goal was to validate the theoretical model exploring the structural relationships amongst these variables.

Conceptualising organisational citizenship behaviour
Over the past three decades, a substantial increase of interest in OCB occurred. Organisational citizenship behaviour as a concept has gained considerable popularity, which is supported by a plethora of scholarly articles documenting OCB as an important work behaviour (Hendricks, 2017;Mahembe, Engelbrecht, Chinyamurindi, & Kandekande, 2015;Unal, 2013). Employees who engage in positive work behaviours are characterised as good citizens. This is because OCBs promote organisational effectiveness, enhance work performance and positively contribute towards overall organisational functioning and survival (Lavelle, 2010;Mahembe & Engelbrecht, 2014;Özturk, 2010). Organ (1990, pp. 43-72) as cited in Unal (2013) itemised positive OCBs and actions as follows: • expressing a personal interest in the work of others (helping others) • staying late or working weekends • providing suggestions for improvement • performing at levels that exceed enforceable standards • caring for organisational property • punctuality • willingness to endure occupational cost • refraining from expressing resentment and complaining about insignificant matters.
Since its inception, numerous definitions and conceptualisations of OCB have been postulated, with Organ's (1988) conceptualisation as the most widely accepted (Mahembe et al., 2015). Organ (1988) defined OCB as workrelated behaviours that are discretionary, not directly or explicitly acknowledged by the formal reward system, and summatively promote the effective and efficient functioning of the organisation. OCB is a multidimensional concept with overlapping dimensions and themes (see Table 1).
The five dimensions of OCB as proposed by Organ (1988) are widely accepted across the globe and are considered as one of the standard measures of this concept. The five dimensions that form the underlying factor structure of the organisational citizenship behaviour scale (OCBS), designed by Podsakoff, Mackenzie, Moorman and Fetter (1990) are altruism, conscientiousness, sportsmanship, courtesy and civic virtue (Organ, 1988).
Altruism refers to the degree of willingness an employee exhibits when helping another employee with task related matters (Unal, 2013). This can include tasks such as helping a colleague with a problem, voluntarily orientating a new employee, sharing strategies, showing employees how to accomplish challenging work-related tasks and teaching employees beneficial and useful knowledge or skills (Unal, 2013).
Conscientiousness refers to the degree to which the employee would go beyond the basic role requirements of the organisation, in terms of observing the rules and regulations, working diligently, not taking extended breaks, attendance, punctuality and working overtime with no extra pay (Farh, Zhong, & Organ, 2004;Unal, 2013).
Sportsmanship refers to the display of willingness to tolerate trivial and temporary shortcomings such as minor grievances and inconveniences by staff members. Sportsmanship involves the preserving of a positive attitude and good spirit amidst challenging work circumstances whilst accomplishing work-related tasks (Sha, 2007).
Courtesy refers to discretionary behaviour demonstrated by employees in the interest of preventing and avoiding future workplace problems and conflicts (Farh et al., 2004;Unal, 2013). A courteous employee is described as an individual who performs thoughtful and considerate gestures towards fellow colleagues and avoids creating problems for co-workers by reducing intergroup conflict (Ishak & Alam, 2009;Unal, 2013).
Civic virtue refers to the interest, constructive involvement and participation of an employee, in the activities of the organisation. This dimension consists of those behaviours such as attending and actively participating in meetings and managerial events, engaging in policy debates as well as expressing one's opinion on the implementation of policy changes (Ishak & Alam, 2009). An employee that displays civic virtue will show commitment to the organisation by looking out for the institution's best interest.

Conceptualising psychological empowerment
The concept of psychological empowerment is widely accepted in both management practice and theory (Cingoz & Kaplan, 2015;Conger & Kanungo, 1988;Gong, Wu, Huang, Yan, & Luo, 2020;Khan, Malik, & Saleem, 2020;Wang & Lee, 2009). A plethora of research has been conducted over the past two decades refining the concept of psychological empowerment and also exploring its antecedents and consequences (Wang & Lee, 2009). Psychological empowerment affects a broad range of outcomes in an organisational context, such as job satisfaction, innovative behaviour, intention to leave, organisational citizenship behaviour, employee creativity and innovation, job performance and organisational learning (Cingoz & Kaplan, 2015).
Psychological empowerment reflects an 'active, rather than a passive orientation towards work roles' (Spreitzer, 1995(Spreitzer, , p. 1444. Psychological empowerment refers to psychological states shaped by the work environment, making sense of an individual's work roles and the active engagement of individuals (Wang & Lee, 2009). Conger and Kanungo (1988) referred to psychological empowerment as a process whereby one's belief in himself or herself is enhanced. Therefore, the conceptualisation of psychological empowerment is viewed as a motivational concept of self-efficacy (Tastan, 2013). Thomas and Velthouse (1990) further defined psychological empowerment as increased intrinsic motivation manifested in a set of four cognitions. These cognitions reflect an individual's orientation

Researcher Themes
Smith, Organ and Near (1983) Altruism and generalised compliance Organ (1988) Altruism, conscientiousness, sportsmanship, civic virtue and courtesy Lin (1991) Identification with the organisation, assistance to colleagues, harmony, righteousness, discipline and self-improvement Williams and Anderson (1991) Organisational citizenship behaviour (OCBI) and organisational citizenship behaviour organisation (OCBO) Van Dyne, Graham and Dienesch (1994) Loyalty, obedience, social participation and functional participation Moorman and Blakely (1995) Interpersonal helping, individual initiative, personal industry and loyal boosterism Van Scotter and Motowidlo (1996) Interpersonal facilitation and job dedication Farh, Early and Lin (1997) Identification with the company, altruism towards colleagues, conscientiousness, interpersonal harmony and protecting company resources Podsakoff, MacKenzie, Paine and Bachrach (2000) Helping behaviours, sportsmanship and civic virtue, organisational loyalty, organisational compliance, individual initiative and self-development to the work role and include meaning, competence, self-determination and impact (Tastan, 2013).
Meaning refers to the value placed on work in relation to an individual's own ideals, beliefs, values and standards (Aksel, Serinkan, Kizilogulu, & Aksoy, 2013). Work is meaningful to individuals when personnel understand the importance of their job for themselves and the organisation, thus they pay more attention to their role and duties (Najafi, Noruzy, KhezriAzar, Nazari-Shirkouhi, & Dalvand, 2011). When employees find work meaningful, the outcome is that they work better and take pride in what they do and accomplish (Najafi et al., 2011).
According to Spreitzer (1995) and Thomas and Velthouse (1990), competence is the degree to which an individual believes they can perform tasks and activities skilfully. This is the self-confidence of individuals in their personal abilities to fulfil their assigned duties or tasks. In addition, there is also the belief that they can utilise the resources provided by the organisation to perform and fulfil their responsibilities (Najafi et al., 2011).
Self-determination refers to an individual's sense of control and autonomy over his or her work activities (Spreitzer, 1995). Employees feel empowered when they form part of decision-making, have an opportunity to exercise authority and have the freedom to choose the extent to which they can contribute (Theron, 2010).
Impact refers to the degree to which an individual may have an influence on work outcomes strategically, administratively and operationally (Ashforth, 1989). Impact refers to the amount of control an individual has over his or her jobs, the influential impact they have over pertinent organisational matters and the accomplishment felt in achieving the objective of a particular task (Theron, 2010).

Conceptualising servant leadership
Servant leadership has its roots in the 1970s work of Robert Greenleaf who realised that for optimal leadership effectiveness the leaders need to serve and care for their followers first through coaching and prioritising the personal skill development of the followers. Greenleaf (1977as cited in Spears, 2005 described servant leadership as follows: The servant-leader is servant first … It begins with the natural feeling that one wants to serve, to serve first. Then conscious choice brings one to aspire to lead. The best test is: do those served grow as persons; do they, while being served, become wealthier, wiser, freer, more autonomous, more likely themselves to become servants? (p. 1) According to Eva, Robin, Sendjaya, Van Dierendonck and Liden (2019) in their meta-theoretical article on servant leadership, the prominence of servant leadership is increasing as indicated by the fact that over 100 research articles have been published during the past 4 years. Servant leadership has been documented as essential for promoting gratitude, empowerment, innovativeness and organisational performance (Baykal, Zehir, & Hole, 2018;Cai, Lysova, Khapova, & Bossink, 2018); organisational citizenship behaviour and team effectiveness (Mahembe & Engelbrecht, 2014) and work engagement (Yang, Ming, Ma, & Huo, 2017). Servant leadership has a negative relationship with turnover intention (Brohi et al., 2018). In addition to the beneficial work outcomes, servant leadership has been reported to enhance student engagement in the classroom (Noland & Richards, 2015); teacher empowerment (Hammond, 2018) and teacher job satisfaction (Georgolopoulos, Papaloi, & Loukorou, 2018;Van der Hoven, 2016).
Despite the prominence and beneficial outcomes of servant leadership, there is no universally agreed on conceptual definition (see Table 2) (Focht & Ponton, 2015). Scholars are, however, convinced that any theory of servant leadership should allude to humility, empathy, altruism, empowerment, caring, vision, trustworthiness, interpersonal acceptance, integrity and hope amongst others as indicated in Table 2 (Van Dierendonck, 2011).

The relationships between servant leadership, psychological empowerment and organisational citizenship behaviour
Leadership behaviours were also found to be an important predictor of OCB. According to Sha (2007), an employee's willingness to engage in OCB is dependent on the leader's influence. Research has found that the quality of the relationship between the leader and the employee is thus of significant importance for OCB (Podsakoff et al., 2000). Organ, Podsakoff and MacKenzie (2005) stated that leaders who set a good example inspire followers to become servant leaders themselves. The servant leader encourages the follower to engage in OCB by serving them through the establishment of quality relationships and driving development and growth. A study conducted by Shahzad et al. (2013) reported that there is a significant positive relationship between servant leadership and OCB. At the school level, teachers are more likely to engage collaboratively with a principal who cares for their own personal development and serves them with humility and unconditional positive regard. The feeling created is likely to motivate the teachers to engage in extra-role behaviours as a form of appreciation. This will in return ensure that the school achieves its objectives (McCrimmon, 2010;Van Dierendonck, 2011).
Based on the given empirical findings and theoretical arguments, it was hypothesised that servant leadership positively affects OCB.
Fong and Snape (2015) posited that empowerment could be perceived as a state that comes from within an individual, demonstrating that the concept of empowerment and intrinsic motivation is similar. The results indicate that the role of empowering employees will result in employees exhibiting organisational citizenship behaviours (Raub & Robert, 2010). This is consistent with a study conducted by Chan, Taylor and Markham (2008) with a sample of 374 participants from a healthcare and educational environment. The results indicated that OCB is a significant outcome of psychological empowerment. In addition, a study conducted by Bhatnagar and Sandhu (2005) collected data from 111 managers from various IT organisations in India and found that managers who experienced psychological empowerment in their working environment exhibited OCBs.
Based on the given empirical findings and theoretical arguments, it was hypothesised that psychological empowerment positively affects organisational citizenship behaviour.
Empowerment is defined by Van Winkle et al. (2014, p. 72) as 'letting people do their jobs by enabling them to learn, grow, and progress' Furthermore, they explain that through leader empowering behaviours, servant leaders grant followers the liberty to progress towards their goals and empower them to make their dreams a reality. Principles of empowerment are at the heart of servant leadership (Patterson, 2003). The power of the servant leader lies in the leader's ability to unleash the potential and thus the power of those around them (Van Winkle et al., 2014). Organ et al. (2005) stated that instead of using power to control subordinates, the servant leader empowers and encourages subordinates to become responsible for their actions.
Based on the given empirical findings and theoretical arguments, it was hypothesised that servant leadership positively affects psychological empowerment.

Theoretical model
After an in-depth literature review, a theoretical model was developed. Figure 1 illustrates the theoretical model that depicts the specific hypothesised causal linkages between servant leadership, psychological empowerment and OCB.

Statistical hypotheses
Close fit hypothesis: The overarching research hypothesis for the close fit null hypothesis is: In order to test the validity of the proposed relationships in the structural model, the following specific research hypotheses were tested: Hypothesis 1: A significant positive relationship exists between servant leadership (ξ 1 ) and psychological empowerment (η 1 ) (H 02: γ 11 = 0; H a2: γ 11 > 0).

Research approach
In order to test the hypotheses postulated in this study, a quantitative research design was employed. Specifically, the theoretical model (see Figure 1) was tested using structural equation modelling (SEM).

Research procedure
Each participant received a questionnaire comprising of an information sheet, biographical section and three measuring instruments, namely servant leadership, psychological empowerment and organisational citizenship behaviour.

FIGURE 1:
The conceptual structural model.

Research method Sample
The sampling technique employed for this study was nonprobability sampling. The research participants comprised of teachers from selected primary and secondary schools in the Western Cape Province in South Africa.
The targeted Metro South District has a population of about 4104 teachers. A total of 400 questionnaires were distributed to selected schools. A total of 203 candidates with usable responses were included in the sample. This total represented a 51% response rate. The sample meets the minimum sample size of 200 for structural equation modeling as recommended by Diamantopoulos and Siguaw (2000).
The sample consisted of mostly female participants (71.9%).
The predominant age groups of the sample were between 51-65 (30%) and 36-50 (38.9%). Furthermore, 57.6% of the participants were married and 30.5% were single. The highest qualification obtained by most of the participants was a diploma (30.5%), a graduate degree (29.1%) and a postgraduate degree of 23.6%. Regarding years of service, the majority have been working for over 10 years (56.7%). Pertaining to position within the school structure, 86.2% of the sample were teachers and 10.8% were heads of departments (HODs). The majority of those who formed part of the sample were Intersen phase teachers (Grade 4-9) (60.5%) followed by Foundation Phase teachers (Grade R-3) (36.9%). With regard to the number of classes a teacher had, 48.8% of the sample had one class. In addition, the majority of the participants (62.6%) had a class size of 31-40 learners.

Measuring instrument
Three questionnaires were identified through the literature review as being reliable, valid and applicable when measuring the constructs under investigation.

Servant leadership
The servant leadership of the principal was measured using the Servant Leadership Questionnaire (SLQ) developed and conceptualised by Barbuto and Wheeler (2006), which measures five dimensions using 23 items. The reliability coefficients for the scale ranged from 0.82 to 0.92 (Mahembe & Engelbrecht, 2013).

Psychological empowerment
The Measuring Empowerment Questionnaire (MEQ) developed by Spreitzer (1995) was used to measure the psychological empowerment levels experienced by the teachers. The MEQ consists of 12 items measuring four dimensions of psychological empowerment (meaning, competence, self-determination and impact). Each subscale has three items. The candidates were asked to select responses from a seven-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree). Spreitzer (1995) reported sound psychometric attributes -an overall Cronbach's alpha coefficient of 0.92. The alpha coefficients for the subscales are 0.92 (meaning), 0.90 (competence), 0.85 (self-determination) and 0.84 (impact).

Organisational citizenship behaviour
The OCBS was used to measure teachers' organisational citizenship behaviours. The OCBS was developed by Podsakoff et al. (1990) and includes 24 items measuring five dimensions (altruism, conscientiousness, sportsmanship, courtesy and civic virtue). A seven-point Likert scale, ranging from 1 (strongly disagree) to 7 (strongly agree), was used. The Cronbach's alpha values for the subscales ranged from 0.70 for civic virtue to 0.85 for altruism. A research study conducted by Mahembe et al. (2015) reported reliabilities ranging from 0.80 to 0.98 on a South African sample.

Statistical analysis
Missing values: Missing data are usually a common occurrence when self-reporting instruments such as questionnaires are used. The missing values problem can have a significant effect on the conclusions drawn from the data. For this study, however, no missing values were found.
Test for multivariate normality: The multivariate normality was initially investigated through the maximum likelihood estimation method. However, because of the fact that some of the data were still not normalised after using this technique, Robust maximum likelihood (RML) estimation was used to estimate the parameters set free in the model (Jöreskog & Sorbom, 1996;Mels, 2003).

Structural equation modelling:
Structural equation modelling is a sophisticated approach to data analysis that dwells on assessing the patterns of covariances amongst the variables specified in the model (Diamantopoulos & Siguaw, 2000;Jöreskog & Sorbom, 1996). It incorporates confirmatory factor analysis, which helps to ensure that the psychometric quality of the measures used is sound. A distinction is made between the measurement and structural models. The measurement model looks at the extent to which the items are representing the manifest variables whilst the structural model goes one step further by depicting the nature of the hypothesised relationships in the model (Diamantopoulos & Siguaw, 2000).

Evaluating the measurement models
Item and dimensional analyses were conducted to determine the quality of the items and to ensure the uni-dimensionality of the scales. Items with corrected item loadings less than 0.30 were flagged as poor whilst the reliability Cronbach's alpha coefficients above 0.70 were accepted. Unidimensionality was performed as a prerequisite for item parcel creation (Holt, 2004). The scales were subsequently subjected to confirmatory factor analyses (CFA) using the LISREL 8.80 software (Jöreskog & Sörbom, 2006). Items with factor loadings less than 0.30 were excluded from inclusion in the model. No items were removed during item, dimensional and CFA.

Ethical considerations
Before commencing with the data collection and distributing the questionnaires, permission was obtained from the Western Cape Education Department: Directorate of Research and the university's research committee. Once permission was granted, the questionnaires were personally delivered to selected primary and secondary schools in the Western Cape Province. Convenience sampling was used and participation was strictly voluntary.

Missing values
Missing values are common in quantitative studies as they dwell on the use of questionnaires. In this study, missing values were not recorded. Hence, the sample size was 203.

Item and dimensional analysis
Item and exploratory factor analyses (EFAs) were performed to identify any poor items in the questionnaires. Exploratory factor analysis was also used to test for uni-dimensionality of the scales (see Table 3). Uni-dimensionality was achieved for all the subscales. High levels of reliability (α = 0.70) were found for each of the scales (Nunnally, 1967).

Goodness of fit of the measurement and structural models
The goodness-of-fit (GFI) indices reported in Table 4 generally indicate that the SLQ, MEQ and OCBS measurement and the structural models indicate a reasonable model fit with the data (Diamantopoulos & Siguaw, 2000;Hair, Black, Babin, & Anderson, 2010). The RMSEA indices for the measurement models fell in the reasonable model fit category (0.05-0.08), as values < between 0.05 and 0.08 represent a reasonable model fit.
The standardised root mean squared residual (SRMR) values of 0.056-0.065 fell above the good model fit (< 0.05). The GFI values for the SLQ, OCBS and the structural models fell below the 0.90 cut-off level whilst the GFI value for the MEQ was above the required cut-off level. The results of the incremental fit measures indicate that all the measurement and structural models achieved a normed fit index (NFI), non-normed fit index (NNFI), comparative fit index (CFI), incremental fit index (IFI) and relative fit indices > 0.90, which represents a good fit (Diamantopoulos & Siguaw, 2000;Hair et al., 2010;Kelloway, 1998). These relative indices, therefore, appear to portray a positive picture of model fit.
The LISREL program also tested the null hypothesis of a close fit (H 0: RMSEA ≤ 0.05). With the exceedance probability (0.032) being very small at the 0.05 level of significance, the close fit null hypothesis was rejected. The structural model, therefore, does not show a close fit. However, after interpreting all other fit indices, the conclusion was reached that the structural model fitted the data reasonably well (see Table 4).

The relationship between servant leadership and psychological empowerment
A positive relationship was found between servant leadership and psychological empowerment (t = 2.49, p < 0.05) (see Table 5). Thus, Hypothesis 1 was confirmed.

The relationship between servant leadership and organisational citizenship behaviour
A positive relationship was found between servant leadership and organisational citizenship behaviour (t = 2.17, p < 0.05) (See Table 5). Thus, Hypothesis 2 was confirmed.

The relationship between psychological empowerment and organisational citizenship behaviour
A positive relationship was found between psychological empowerment and organisational citizenship behaviour (t = 2.30, p < 0.05) (See Table 5). Thus, Hypothesis 3 was confirmed.

Discussion
The aim of the study was to test a theoretical model on the relationships between servant leadership, psychological empowerment and organisational citizenship behaviour on a sample of teachers working in the Western Cape. The specific objectives were to determine the manner in which servant leadership and psychological empowerment influence organisational citizenship behaviour. A further objective was to validate the theoretical model exploring the structural relationships amongst these variables in the South African school system at selected schools in the Western Cape province.
The validation of the theoretical model entails inspecting the GFI indices for the measurement and structural models to determine whether the implied covariance (theoretical model) is supported by the observed covariance (the data). An inspection of the GFI indices indicates that the measurement and structural models generally display a reasonable and acceptable fit. The positive support for the measurement models further demonstrates that the items used in the questionnaires measured the latent variables as postulated by the instrument developers: Hypothesis 1: Servant leadership affects psychological empowerment.
The t-value (t = 2.49) of the link between servant leadership and psychological empowerment is greater than 1.96. A significant (p < 0.05) positive relationship is, therefore, evident between these two constructs, which suggests that the proposed relationship between servant leadership and psychological empowerment was supported. Therefore, school principals who practice servant leadership are likely to empower their teachers. This is consistent with the findings reported by Van Winkle et al. (2014) in a study involving 116 employees of small businesses working in the United States of America. The confirmation of the relationship is not surprising given the understanding that servant leadership incorporates ideals of empowerment and is at the heart of servant leadership (Patterson, 2003): Hypothesis 2: Servant leadership affects organisational citizenship behaviour.
The t-value of the link between servant leadership and OCB is more than 1.96, indicating that there is a significant relationship between servant leadership and OCB (t = 2.17, p < 0.05). This result is consistent with previous research studies that have reported the positive influence and significant relationship of supportive and value-based leadership styles on citizenship behaviour (Mahembe & Engelbrecht, 2014;Podsakoff et al., 2000;Shahzad et al., 2013). One can infer that principals as servant leaders are likely to use their attributes in ways that motivate school teachers to go beyond their job descriptions by engaging in extra-role behaviours. Learning takes place all the time and, therefore, dedicated teachers are likely to assist their learners after class time: Hypothesis 3: Psychological empowerment affects organisational citizenship behaviour.
There is a statistically significant relationship between psychological empowerment and OCB (t = 2.30, p < 0.05). This finding suggests that the proposed relationship between these two variables exists. This is consistent with a study conducted by Chan et al. (2008) on a dyad sample of 374 participants from a health care and educational environment.
The results emanating from this study indicated that OCB is a significant outcome of psychological empowerment. In addition, a study conducted by Bhatnagar and Sandhu (2005) collected data from 111 managers from various information technology (IT) organisations in India and found that managers who perceive psychological empowerment in their working environment exhibited OCB. Furthermore, a research study conducted by Garg and Suri (2013) in the public banking sector in India revealed that psychological empowerment and OCB are highly and significantly correlated with each other. More recently, Bester, Stander and Van Zyl (2015) provided further support with results revealing that leader empowering behaviours and psychological empowerment significantly influences OCB.

Power assessment
The Rweb (1.03) syntax developed by Preacher and Coffman (2006) was used to determine the statistical power of the sample size in relation to the number of variables used in the study. In the present study, the significance level of 0.05, the sample size of 203 and the degrees of freedom 242 were entered into the Rweb syntax. The power value of 0.9982771 for the test of exact fit was obtained (see Table 6). This power value is reasonably large. In this case, the authors rejected the null hypothesis of exact fit. They calculated the power of the test of close fit as 0.9998195, which was high implying that, in the conditions that characterised this specific study, one would reject approximately 99.98195% of incorrect models. This shows a high statistical power and confidence in the model tested in this study.

Limitations of the study
This study was of a cross-sectional nature and the use of an availability sample merely provides a representation of teachers' opinions, feelings and perceptions at a particular point in time, and it does not capture maturational factors. This study collected data at a single point in time by using a single-point-in-time survey measurement. It is still widely accepted that measures employed in social sciences research are subject to a number of sources of error (MacKenzie & Podsakoff, 2012). Data collected at a single point in time, rather than having a longitudinal study, may exacerbate same-source or common method bias (Arnolds & Boshoff, 2004;Rylander, 2003).
The sample size that was used was too small, hence the study cannot be generalised to the entire population of teachers within the Western Cape province. In addition, only schools within the Western Cape were used and specifically Metro South District, thus the findings cannot be generalised to the entire population of teachers within the province or within South Africa.
Another shortcoming of the study is the measurement instruments used. The measuring instruments (questionnaires) were based on self-reporting methods. Self-report methods are influenced by how a teacher feels at a particular point in time and therefore, is subjective in nature and susceptible to social desirability.
This study only used a quantitative research approach and data were collected using only questionnaires and no other data collection methods such as interviews. Hence, the subjective experiences of the participants were not captured.
In this study, the sample was disproportionate as the sample consisted mostly of female participants (71.9%).

Managerial implications
This study reported positive relationships between servant leadership, psychological empowerment and organisational citizenship behaviour. The findings imply that school principals should practice servant leadership in order to create and develop a school environment that is conducive for teaching, learning and optimal functioning. A principal's servant leadership approach is likely to promote teacher empowerment and development within a conducive learning environment, as it enables the principal to develop teachers to their fullest potential through mentoring and coaching. Therefore, sufficient time should be made available to school leaders to build trusting relationships when empowering teachers. Through empowering others, institutions are discovering that servant leadership can truly improve how organisations are developed and function, whilst successfully meeting organisational objectives (Focht & Ponton, 2015).
This study will be of value to the human resource function specifically in the formulation of valid psychological explanations of the behaviour of teachers and the implementation of appropriate interventions to ensure an effective and efficient teaching workforce. It will also assist in the reviewing of the criteria used in the appointment of school principals. Imparting servant leadership skill to principals is likely to be met with motivation from the teachers to go beyond their usual duties as payback for the efforts and interest in the development of followers (teachers) thereby improving the service delivery of the teachers.

Conclusion
The principals who embrace the values of servant leadership are likely to propagate some feelings of psychological empowerment in their subordinates. Servant leaders are concerned with the growth of their subordinates at the expense of their own development through engaging in selfless coaching and mentoring interventions. These efforts from the servant leader principals will, however, not go unnoticed. As indicated in this study, teachers whose principals engage in servant leadership are likely to go the extra mile in terms of performance. This research has shown that the use of servant leadership has important implications for teacher development interventions that can be used to improve learning in schools.

Funding information
This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

Data availability
Data sharing is not applicable to this article as no new data were created or analysed in this study.