The effect of servant leadership on the organizational citizenship behavior of instructors in public universities in Amhara region

Abstract

The purpose of this study was to examine the effect of servant leadership on the organizational citizenship behavior (OCB) of instructors in public universities in Amhara region. An explanatory sequential mixed research design was employed. The size of the population was 2170. Of these, 338 instructors were taken into the sample using Yamane’s (1973) formula. Questionnaires and interviews were used to collect the data. Both quantitative and qualitative analysis techniques were used to analyze the data. The results obtained through the questionnaire and interview revealed that servant leadership and OCB were practiced moderately in the universities. There was a positive and statistically significant relationship between servant leadership and OCB with a standardized correlation coefficient (r = .578) by the critical ratio test greater than ± 1.96 at p < .05.Thus, 33.4 % of the variance in OCB was predicted by servant leadership with standardized regression coefficient (β =.578) at p < .05.The result of one way ANOVA indicated that significant differences were observed among instructors in their perception of servant leadership and OCB. Based on the finding of this study, it is concluded that servant leadership had a significant effect on the OCB of instructors. Therefore, it is suggested to improve the implementation of servant leadership and OCB.

Keywords:

• instructors
• organizational citizenship behavior
• public universities
• servant leadership

PUBLIC INTEREST STATEMENT

Though some scholars criticize servant leadership as idealized which cannot be implemented in the real world, it becomes a popular issue in the literature when organizations increasingly demand genuine leaders. That is, service-oriented leadership styles such as servant leadership can enhance the extra-role behaviors of employees in the workplace. The purpose of this study was to assess the effect of servant leadership on the organizational citizenship behavior of instructors in the public universities in Amhara region. An explanatory sequential mixed design was employed. The size of the population was 2170. Of these, 338 instructors were taken into the sample using Yamane’s (1973) formula. Questionnaires and interviews were used to collect the data. Both quantitative and qualitative techniques were used to analyze the data. Based on the finding of this study, it is concluded that servant leadership had a significant effect on OCB. Finally, important recommendations have been forwarded at the end.

1. Introduction

In this competitive environment, organizations highly depend on the quality of leadership to facilitate the changes required to maintain equitable advantage through optimizing human capital. Leadership is a highly complex concept to define in which most of the early definitions focus on the ability of a person to influence others (Leithwood & Jantzi, 1999). But now a days many scholars have shifted their attention from self-centered and individualized forms of leadership to follower-oriented types of leadership (Peterson et al., 2012). For leadership to be effective, it must be built on a solid foundation consisting of a clear mission, vision, specific strategies, and culture instrumental for success. Effective leaders exhibit specific skills and attribute to achieve the goal of the organization by creating a conducive environment. Most importantly, strong communication skills, self-confidence, the ability to manage others, and the willingness to embrace change are also qualities of good leaders (Investopedia, 2016).

Servant leadership is one of the theories of leadership initiated by Greenleaf in Greenleaf (1970) which gives high emphasis to human needs (Ehrhart, 2004). A research done by Jiménez-Estévez et al. (2023) indicated that servant leaders enhanced the emotional well-being of hotel employees in Spanish. This implies that employees who have servant leaders increased their affective well-being. For dynamic employees, both having a servant leader and experiencing personal development are imperative to feel safety (Jiménez-Estévez et al., 2023). Similarly, Ruiz-Palomino et al. (2022) found that employees who have servant leaders were less likely affected by depression. This means that servant leadership has the potential to reduce the level of employee depression in time of difficulty because servant leaders have the quality to give high attention to meet the needs of employees. Employees with servant behaviors can develop trust on others, in turn, leaders could support these employees get higher levels of personal social capital by addressing the existing challenges successfully (Ruiz-Palomino et al., 2022). Central to servant leadership idea is prioritizing the wellbeing of followers, valuing and developing people, building of community and the practice of authenticity.

Servant leadership manifests itself in the care taken by the servant first to make sure that other people’s highest priority needs are being served (Greenleaf, 1970; Ruiz-Palomino et al., 2022). He suggests that the highest test of servant leadership is whether those served grow as people, become healthier, wise, freer and more autonomous. Unlike traditional leadership where the needs of followers are left behind the interest of the leaders, servant leadership puts the needs of employees over the leader (Greenleaf, 1977). This type of leadership is not only focused on the benefits of others but also organizational stewardship instructions concentrated on providing service and enhancing the community (Bobbio et al., 2012). It can also develop and empower followers to reach their highest potential (Lord et al., 1999; Sendjaya & Sarros, 2002). These can be achieved through developing self-confidence, inspiring trust and providing information (Liden et al., 2008).

Though some scholars criticize servant leadership as idealized which cannot be implemented in the workplace (Lloyd, 1996; Northouse, 2013), it has become a popular issue in the literature when organizations increasingly demand ethical and genuine leaders (Blanchard, 2002; Boyum, 2006; Covey, 2006). That is, organizations with servant leadership cultures tend to be healthy, efficient, and leader-employee relationships are smooth and synchronized (Doraiswamy, 2012; Korkmaz, 2007). A research done by Ruiz-Palomino et al. (2021) also indicated that servant leadership had a positive indirect effect on service differentiation and a culture of innovativeness. This shows that servant leadership can create conditions that support the successful implementation of strategic service differentiation in the workplace.

Servant leadership has the potential to offer organizations with leadership approach grounded in values, ethics, morals, and empowerment of others. According to Ruiz-Palomino et al. (2019), servant leadership had an indirect positive impact on the innovativeness of firms’ innovativeness. These results provide new human resource related insights regarding the encouragement of firm innovativeness. Strong ethics, belief, and values are seen as the core of servant leadership which bring long-term success to organizations. Servant leader produces a motivated workforce and eventually successful organizations through cultivating a participatory approach, empowering environments, and fostering the talents of followers.

Organizational citizenship behavior (OCB) is another organizational variable that is described as a discretionary behavior that is neither in the job description nor recognized in the formal reward system but it contributes to the success of the organizations (Hoffman et al., 2007). In other words, OCB is described as a special type of employee behavior beyond formal duties such as working for extra hours, helping colleagues, and giving suggestions to management to enhance performance and improvements (Organ, 1988). It is conceptualized in terms of altruism, conscientiousness, sportsmanship, courtesy, and civic virtue (Organ, 1988).

Employees who engage in positive work behaviors can enhance organizational effectiveness, improve work performance, and positively contributes towards organizational function and survival (Mahembe & Engelbrecht, 2014; Ozturk, 2010). OCB can also facilitate the success of organizational goals by improving the commitment of employees and the effectiveness of the organization (Ozturk, 2010; Podsakoff & MacKenzie, 1997). In addition, OCB is considered a precondition for organizational performance (Bolino & Turnley, 2003; Zellars et al., 2002). It encompasses anything positive and constructive that employees do of their own volition which supports the function of the organization. Studies have shown that OCB makes important contributions to individuals, groups, and the effectiveness of organizations (Organ et al., 2006).

Serving leadership is a leadership that begins with a sincere feeling arising from a heart that desires to serve, that is to be the first serving party (Haider et al., 2015). They considered servant leadership as an important factor that influences OCB. Numerous studies show that there is a significant relationship between servant leadership and OCB in the workplace (Dirks & Ferrin, 2002; Ehrhart, 2004; Hu & Liden, 2011). Smith et al. (1983) also found a positive relationship between servant-leadership and OCB. This shows that servant leaders strengthen individual-level OCB to build task focused community, it will also eventually help the whole team effectiveness (Ruiz‐Palomino et al., 2023).

Similarly, Lin and Peng (2010) found that the increased frequency of OCB by individual team members should enhance overall performance of employees. Similarly, servant leadership has an impact on employees’ relationship, which is fully mediated by social relations with colleagues inside and outside their groups respectively (Zoghbi-Manrique de Lara & Ruiz-Palomino, 2019). This shows that a better understanding of the mechanisms to improve the set of personal resources help employees to develop their work. Hence, leaders must understand that meeting the needs of their employees first and foremost is a positive for the successful of all involved (Ruiz‐Palomino et al., 2023).

According to Linuesa-Langreo et al. (2018), extra role behavior of employees partially mediates the positive effect of servant leadership on group social capital. These findings can help managers to lead their business and enhance the competitiveness of their work groups, and organizations.

Based on the discussions made so far about servant leadership and OCB as well as the relationship between these variables, the researcher has synthesized a new conceptual framework for this study as indicated in Figure 1.

Figure 1. The conceptual framework of the study.

Source: Author own survey, 2021.

Employees who are supported by their leaders and coworkers are willing to exert extra effort for the accomplishment of organizational goals beyond their formal duty. A range of research findings indicated that servant leadership has a significant effect on OCB (Haider et al., 2015; Ruiz‐Palomino et al., 2023; Zhao et al., 2014). Employees with good performance can produce the strongest link between performance and functional participation (Turnipseed & Rassuli, 2005). Indeed, as citizenship appears to consist of discretionary behaviors, how the employee perceives the organization would likely influence the employee either perform or withhold such performance (Grojean et al., 2004). Even if maximum performance within the limits of contract obligations is shown, no one can be successful unless they engage in citizenship activities that contribute more to the interest of the organization than their own need. Moreover, helping behaviors enhance the social environment in the organization, reduce the rate of absenteeism and turnover intentions, and increases employee well-being and productivity (Podsakoff et al., 2000).

Currently, there appears to be widespread evidence that senior faculty members hold leadership positions in higher education institutions without adequate preparation. As a result, staff in higher education are subjected to the feeling of unfair treatment and become less committed. Some authors tried to examine the specific challenges encountered by academic leaders, such as the complex and dynamic social, economic, and political contexts in most universities that are responsible for ineffective leadership in higher education (Smith & Hughey, 2006). Poor leadership is also responsible for the subordination of an organization’s interest for self-interest, violation of the professional code of ethics, undermining organizational missions and goals, and declining attentions to organizational growth and development (Hellsten, 2006).

Although some studies have been done on servant leadership and OCB, research has not been conducted on the causal relationship between these variables in the context of Ethiopian public universities in general and in the public universities of Amhara region in particular. Therefore, this study aimed to examine the effect of servant leadership on the OCB of instructors in public universities in Amhara region. Thus, the researcher formulated the following research questions to guide this study:

1. What is the perception of instructors towards servant leadership and OCB in the public universities in Amhara region?

2. Is there a significant relationship between servant leadership and OCB in the public universities in Amhara region?

3. What is the effect of servant leadership on OCB in the public universities in Amhara region?

4. Are there differences among instructors in their perception of servant leadership and OCB in the four generations of public universities in Amhara region?

2. Materials and methods

2.1. Research design

Research design is the plan and structure of the investigation used to obtain evidence to answer research questions. Based on the nature of the study, researchers may follow different research designs to obtain answers to the research questions. For the purpose of this study, explanatory sequential mixed design (QUAN qual) was employed to understand the quantitative results in depth using the qualitative data. Explanatory sequential mixed design has a strong quantitative orientation in which the researcher first gathers the quantitative data and analyzes the results, and then plans the qualitative phase of the study using the quantitative results (Creswell, 2014). This design will capitalize on the strengths and minimize the limitations of quantitative and qualitative research approaches (Creswell & Clark, 2007; Johnson & Onwuegbuzie, 2004). A simplified illustration of explanatory sequential mixed design is shown in Figure 2.

Figure 2. Explanatory sequential mixed design.

(source: Creswell, 2014, p. 270).

2.2. Research settings

This study was conducted in the public universities of Amhara region that were categorized into four generations based on the year of establishment. That is, two universities in the 1^st generation (from 1950–1962), three universities in the 2^nd generation (in 2007), two universities in the 3^rd generation (in 2011), and three universities in the 4^th generation (in 2017) were established. These generations were considered as which strata contained universities with nearly similar characteristics such as infrastructures, staff profiles, academic programs, and so on.

2.3. Population, sample, and sampling techniques

Initially, six universities were selected out of ten universities from the specified strata for manageability reasons. That is, the University of Gondar from the 1^st generation, Wollo and Debre Markos Universities from the 2^nd generation, Debre Tabor University from the 3^rd generation, and Injibara and Debarq universities from the 4^th generation were selected using a stratified random sampling technique. These stratifications allowed us to ensure the representation of universities from each generation and make comparisons among respondents in their perception of servant leadership and OCB.

Then, 21 colleges were selected from the six universities to determine the size of the population and subpopulations of the study. Specifically, five colleges from the University of Gondar, four colleges from Wollo University, four colleges from Debre Markos University, four colleges from Debre Tabor University, two colleges from Injibara University, and two colleges from Debark University were selected using a simple random sampling technique mainly lottery method. In this regard, a total of 2170 instructors that were found in the selected universities were taken as the population of the study. Of the total academic staff; 731 from the 1^st generation, 710 from the 2^nd generation, 427 from the 3^rd generation, and 302 from the 4^th generation were considered as the subpopulations of the study.

Since the study population was found large, employing appropriate sampling techniques has become necessary to select the required sample size without jeopardizing its representatives. This helps to minimize challenges related to time and resource constraints. Thus, the sample size of this study was determined using Yamane’s (1973) formula given:

$\mathrm{n}=\frac{\mathrm{N}}{\hspace{0.17em}1+\mathrm{N}{\left(\mathrm{e}\right)}^2}$

Where;

N = total population

n = required sample size

e = level of precision = 0.05 (5%), and assuming 95% confidence level at 0.5 variability

Using the above formula, 338 respondents were drawn from the total of 2170 instructors as the sample of this study using the given formula. The calculation of the sample size is as follows:

$\mathrm{n}=\frac{2170}{1+2170{\left(.05\right)}^2}=338$

After determining the total sample size of the study, the sample size of each stratum was calculated based on the respective size of their population using the proportional sample allocation method developed by Pandey and Verma (2008). This allowed me to select a representative sample from each stratum since the number of instructors in the selected universities was significantly varied. Hence, the researcher determined the sample size of each stratum using the Pandey and Verma (2008) formula as given below. i.e.

$n_k={\left(\frac{Nk}{N}\right)}_{\hspace{0.17em}}\ast n$

Where;

n_k = Sample size of k^th strata

N_k = Population size of the k^th strata

N = Total population size

n = Total sample size

Of the total sample of 338 instructors; 114 from the 1^st generation, 111 from the 2^nd generation, 67 from the 3^rd generation, and 66 from the 4^th generation were taken into the sample using proportional stratified random sampling technique with the help of the above formula. Then, these 338 samples were selected using simple random sampling technique, mainly lottery method with the intent to give equal chance to all teachers to be selected in the sample. This helped to avoid the misrepresentation of some members of the population in the study. The summary of the population and sample of the study are shown in Table 1 below.

Table 1. Summary of population and sample of the study

Strata	Name of universities	Population size of each university (Nk)	Sample size of each university (nk)
1^st generation	University of Gondar	731	114
	Wollo University	280	44
2^nd generation	Debre Markos University	430	67
3^rd generation	Debre Tabor University	427	67
	Injibara University	145	22
4^th generation	Debark University	157	24
	Total	N= 2170	n= 338

Source: Author own survey 2021.

With regard to the qualitative phase of the study, five respondents were purposively selected to collect the qualitative data through interviews to further explain the quantitative results.

2.4. Data collection tools

Both questionnaires and interviews were used to collect the data. The questionnaire has three parts containing close-ended items. The first part of the questionnaire dealt with respondents’ characteristics related to sex, work experience, educational qualification, and generations of universities. The second part of the questionnaire assessed respondents’ perception of servant leadership in the workplace using the latest version of Liden and his colleagues (Liden et al., 2008). It was measured using 28 items organized in seven dimensions—emotional healing, conceptual skill, creating value for the community, empowering, helping subordinates grow and succeed, putting the interest of subordinates first, and behaving ethically. The third part of the questionnaire consisted of 25 items developed by Lee and Allen (2002) to measure respondents’ perception of OCB in terms of altruism, conscientiousness, sportsmanship, courtesy, and civic virtue.

Finally, respondents were asked to rate items using five points Likert items ranging from 1 (not at all) to 5 (a very great extent) to measure servant leadership and OCB in the workplace.

With regard to qualitative data, a semi-structured interview was conducted on the status of servant leadership and OCB in the study area.

2.5. Validity and reliability of the questionnaire and interview

The face and content validity of the questionnaire were checked through technical evaluation by research experts. The face validity of the questionnaire related to the feasibility, readability, clarity of language, consistency of style, and formatting was improved based on the comments of my colleagues. With regard to content validity, the organization of the questionnaire related to relevance, simplicity, and wording of items was adjusted based on the feedback obtained from the research advisors. After items were identified to be included in the final questionnaire, the content validity index was calculated for the mean of the content validity ratio values of the retained items. Thus, the content validity indices of items are (.87 and .82) for servant leadership and OCB respectively showing that items were valid to measure their respective dimensions.

The pilot test was also conducted on 90 selected respondents out of the main sample of the study to assess the reliability of the questionnaire. The distribution of the sample for the pilot test followed the same procedures as in the main sample of the study. Cronbach Alpha was used to test the internal consistency of items. The reliability coefficients of the instrument with Cronbach Alpha (α) = (.84) and (.85) for items concerning servant leadership and OCB respectively. This indicates that items in the respective dimensions with reliability coefficients ≥ .80 are considered internally consistent to measure servant leadership and OCB constructs (George & Mallery, 2010).

2.6. Data collection procedures

Initially, a permission letter was sought from Bahir Dar University to collect the data from the research sites. Based on the given permission, the list of target respondents was accessed, and identified the required samples. Participants were informed about the absence of potential risks and benefits due to participation in the study. Moreover, respondents were informed that their involvement in the study was voluntary, so they would withdraw from the research at any time. More importantly, any communication with the concerned bodies was not carried out without the consent of the research respondents. Then, the researcher distributed the questionnaire to the sample respondents. After the questionnaire was returned to the researcher and analyzed the data, the interview was conducted with each interviewee through face to face approach inside the universities to further explain the quantitative results.

2.7. Data analysis techniques

Both descriptive and inferential statistical analysis techniques were used to analyze the data using SPSS and AMOS software. Specifically, one sample t-test was used to measure the status of servant leadership and OCB in the workplace (Research question #1). Structural equation modelling was used to assess the relationship between servant leadership and OCB (Research question #2). Structural equation modelling was also used to analyze the effect of servant leadership dimensions on OCB (Research question #3). In addition, ANOVA was employed to determine whether there are significant differences among instructors in their perception of servant leadership and OCB in the four generations of public universities in Amhara region (Research question #4).

Finally, the data collected through a semi-structured interview on the status of servant leadership OCB were analyzed qualitatively.

3. Results

This section presents the results and discussion of the study according to the themes of the research questions. It began with testing the construct validity of the respective dimensions of the latent variables and measurement model fit through conducting exploratory and confirmatory factor analyses (Leech et al., 2005).

3.1. Testing the construct validity and measurement model

Exploratory and confirmatory factor analyses were carried out with the respective dimensions of servant leadership and OCB to assess whether the proposed measurement model fit to the data. This aims to establish construct validity of the measurement model after checking the internal consistency of items (Temesgen, 2017). The validation of the respective subscales of servant leadership, and OCB was carried out for factor analysis that is presented in the next subsections.

Even though there is no clear criterion to decide what is large or small, items with factor loadings ± .33 and above are considered to satisfy the minimum level of threshold based on the suggestion of Ho (2006). The result of factor analysis indicated that the items in the dimensions of servant leadership construct had factor loadings higher than|.50|with practical significance. As indicated in Table 2, 26 items were highly loaded on the dimensions of servant leadership construct with factor loading values ranging from .601 to .886 and thus they accounted for more than 53 % of the variance in servant leadership construct. But, two items with low factor loadings were excluded from the analysis for they suppressed the accuracy of the results. Since factor analysis of items in each dimension resulted in a one factor solution, the results of the principal component analysis showed that the factor loadings of servant leadership dimensions with eigenvalues ranging from 1.592 to 2.976 were greater than the minimum threshold of 1.

The results of factor analysis also indicated that the items in the respective dimensions of OCB construct had high factor loadings higher than |.50|with a .05 significance level. To state it in detail, 23 items were heavily loaded on the dimensions of OCB with factor loading values ranging from .681 to .914, and they accounted for more than 52 % of the variance in OCB construct. On the contrary, two items with low factor loadings were excluded from the analysis for they suppressed the accuracy of the results. The results of the principal component analysis indicated that the respective dimension of OCB construct had factor loadings with an eigenvalue greater than the minimum threshold of 1. That is, the factor loadings of OCB dimensions with eigenvalues ranging from 1.949 to 3.175 as shown in Table 3.

Based on the results of the construct validity, the dimensions of servant leadership and OCB were used as parcels to run AMOS analysis. Each dimension (parcel) consists of items ranging from 4 to 5 questions. Specifically, the researcher identified the seven dimensions of servant leadership (emotional healing, creating value for the community, conceptual skill, empowering, helping subordinates grow and succeed, putting subordinates first, and behaving ethically) and five dimensions of OCB (altruism, conscientiousness, sportsmanship, courtesy, and civic virtue), as indicators. Taking the identified dimensions of the two latent variables, the researcher constructed the measurement model using AMOS version 23 as shown in Figure 3.

Figure 3. The measurement model of the study.

Source: Author own survey 2021.

After testing the internal reliability of the questionnaire, confirmatory factor analysis was carried out with the respective dimensions of servant leadership and OCB to assess the parameter estimates and the overall fit of the measurement model to the data. Although there is little agreement among scholars on the type of fit indices and their cutoff points as to what constitutes good fit, the researcher used relative chi-square (CMIN/DF), goodness fit index (GFI), adjusted goodness of fit index (AGFI), normed fit index (NFI), incremental fit index (IFI), Tucker-Lewis index (TLI), comparative fit index (CFI) and root mean square error of approximation (RMSEA) to assess the degree to which the measurement model fits the observed covariance matrix using AMOS version 23. The goodness of fit indices used to assess the measurement model fit are indicated in Table 4.

Table 2. Factor component matrix estimate of items in servant leadership dimensions

Number of items in the subscales	Factor loadings
	Emotional healing	Creating value for the community	Conceptual skill	Empowering	Helping subordinates grow	Putting subordinates first	Behaving ethically
1	.713	.800	.803	.882	.761	.843	.848
2	.815	.807	.838	.837	.818	.854	.886
3	.649	.754	.869	.692	.601	.809	.871
4	.	.874	.840	.770	.	.784	.844
Eigenvalues	1.592	2.622	2.809	1.957	1.609	2.710	2.976
% of variance	53.174	65.557	70.228	65.247	53.452	67.743	74.400

As indicated in Table 4, the measurement model satisfied the fit indices. That is, the results indicated that the measurement model fits the data by the chi-square test, χ2 (N = 338, df = 53) = 225.884, p < .05.In addition, the fit indices of GFI (.947), AGFI (.917), NFI (.929), IFI (.942), TLI (.925) and CFI (.942) were higher than the threshold of .90 and a RMSEA (.045) was lower than the cutoff of point .05 with p = .000.This indicates that the possible improvement of the measurement model ranging from .053 to .083. The maximum likelihood estimates of the unstandardized and standardized regression weights also confirmed that all the path coefficients in the model are significant at p < .05 as shown in Table 5.

Table 3. Factor component matrix estimate of items in OCB dimensions

Number of items in the subscales	Factor loadings
	Altruism	Conscientiousness	Sportsmanship	Courtesy	Civic virtue
1	.681	.816	.903	.808	.754
2	.814	.811	.914	.810	.800
3	.738	.792	.891	.825	.861
4	.753	.795	.872	.844	.833
5	.	.768	.850	.687	.
Eigenvalues	2.112	3.173	2.443	3.175	1.949
% of variance	52.810	63.463	81.443	63.459	64.953

Table 4. AMOS outputs on the fitness indices against the criteria of the measurement model

Criteria	Obtained values	Threshold
Relative chi-square (CMIN/DF)	4.828	<5
Goodness of fit index (GFI)	.947	>.90
Adjusted goodness of fit index (AGFI)	.917	>.90
Normed fit index (NFI)	.929	>.90
Incremental fit index (IFI)	.942	>.90
Tucker-Lewis index (TLI)	.925	>.90
Comparative fit index (CFI)	.942	>.90
Root mean square error of approximation (RMSEA)	.045	<.05

Source: Author own survey 2021.

As indicated in Table 5, the unstandardized regression weights of all the dimensions of servant leadership and OCB are significant with the critical ratio test greater than ± 1.96 at p < .05.This indicates that the critical ratio tests of all dimensions are extremely far from the threshold of ± 1.96 which indicates a significant path at p < .05.The standard regression estimates of 12 dimensions in the measurement model were significantly represented by their respective latent variables. Specifically, the standardized regression weights of the observed variables in the measurement model range from .329 (behaving ethically) to .815 (creating value for the community). This implies that the observed variables explained the respective latent constructs ranging from 10.82 to 66.42 %. These values indicate that servant leadership and OCB were significantly measured by their respective dimensions.

Table 5. Unstandardized and standardized regression weights of the measurement model

Parameters/dimensions			Unstandardized				Standardized
			Estimate	S.E.	C.R.	P	Estimate
Emotional healing	<—	SL	1.000				.752
Creating value	<—	SL	1.117	.052	21.468	***	.815
Conceptual skill	<—	SL	1.077	.056	19.364	***	.727
Empowering	<—	SL	.959	.051	18.900	***	.710
Helping subordinates	<—	SL	.737	.055	13.466	***	.510
Putting subordinates first	<—	SL	.642	.055	11.657	***	.442
Behaving ethically	<—	SL	.472	.072	6.529	***	.329
Altruism	<—	OCB	1.000				.738
Conscientiousness	<—	OCB	1.109	.053	21.084	***	.784
Sportsmanship	<—	OCB	1.076	.051	21.205	***	.789
Courtesy	<—	OCB	1.024	.048	21.117	***	.786
Civic virtue	<—	OCB	1.048	.052	20.284	***	.754

Source: Author own survey 2021.

Thus, it is concluded that all dimensions in the measurement model were internally consistent and structurally valid to measure their respective constructs.

3.2. Status of servant leadership and organizational citizenship behavior

Mean scores were used to assess the extent servant leadership and OCB including their respective dimensions were practiced in the workplace. Hence, the results are indicated in Table 6.

Table 6. A one sample t-test for the dimensions of servant leadership and OCB

Dimensions	Mean	Std.	Mean difference	t-value	df	Sig.(2-tailed)
Servant leadership
Emotional healing	3.54	.777	.53945	12.755	337	.000
Creating value for the community	3.61	.701	.60725	15.917	337	.000
Conceptual skills	3.76	.676	.76479	20.788	337	.000
Empowering	3.38	.766	.38240	9.177	337	.000
Helping subordinates grow and succeed	3.51	.723	.50690	12.886	337	.000
Putting subordinates first	2.78	.578	−.21967	−6.978	337	.000
Behaving ethically	3.17	.985	.17308	3.230	337	.001
Grand mean score	3.39	.514	.39347	14.072	337	.000
Organizational citizenship behavior
Altruism	3.74	.683	.74112	19.925	337	.000
Conscientiousness	3.17	1.131	.16982	2.759	337	.000
Sportsmanship	2.97	.875	−.020	−.369	337	.713
Courtesy	2.39	.852	−.61006	−13.159	337	.000
Civic virtue	3.73	.579	.73422	23.274	337	.000
Grand mean score	3.21	.537	.21448	7.339	337	.000

N=338, *Sig. <.05

Source: Author own survey 2021.

The results of one sample t-test indicated that the mean scores of emotional healing (3.54), creating value for the community (3.61), conceptual skill (3.76), empowering (3.38) and helping subordinates grow and succeed (3.51), and behaving ethically (3.17) were higher than the expected mean value (3) at (t = 27.435), (t = 18.126), (t = 13.626), (t = 9.177), (t = 21.608), and (t = 3.230) respectively at p < .05, df = 337. The respective positive mean differences and t-values of these dimensions also ensured that emotional healing, creating value for the community, conceptual skill, and helping subordinates grow and succeed were showed by the department heads as perceived by instructors. On the contrary, the mean score of putting subordinates first (2.78) was less than the expected mean value (3) at (t = −6.978) respectively at p < .05, df = 337. The negative mean difference and t-value of this dimension proved that putting subordinates first was exercised to a little extent by the department heads. In sum, the grand mean score of servant leadership (3.39) was higher than the expected mean value (3) at (t = 14.072) at p < .05, df = 337.

Similar result is also obtained through the interview on the status of servant leadership. Almost all interviewees reported that department heads showed servant leadership behavior in relation to emotional healing, creating value for the community, conceptual skills, empowering, helping subordinates grow and succeed and behaving ethically. Specifically, Participant 1 reported that department heads showed servant leadership behavior in terms of emotional healing, creating value for the community, and helping subordinates in achieving their goals. Similarly, Participant 2 and 4 said that their department heads were interested in helping people who faced different problems in the community. In the same way, Participant 3 and 5 reported that their department heads were competent enough to solve work-related problems effectively.

On the contrary, the majority of respondents stated that department heads were less likely to exhibit servant leadership behavior in serving subordinates first. For example, Participant 1, 2, 3 and 5 reported that their department heads made decisions about their job without collecting accurate information. They also reported that their department heads did not give high emphasis to satisfy the needs of instructors.

With regard to OCB, the results of one sample t-test showed that the mean scores of altruism (3.74), conscientiousness (3.17), and civic virtue (3.73) were higher than the expected mean value (3) at (t = 19.925), (t = 2.759) and (t = 23.274) respectively at p < .05, df = 337. This implies that academic staff was to some extent volunteered to support their coworkers and universities by offering important suggestions. On the contrary, the mean scores of sportsmanship (2.97), and courtesy (2.39) were lower than the expected mean value (3) at (t = −.369), and (t = −13.159) respectively at p < .05, df = 337. This means that instructors were a little extent willing to perform the function of their universities voluntarily. Thus, the grand mean score of OCB (3.21) was higher than the expected mean value (3) at (t = 7.339), df = 337.

The responses obtained through the interview also indicated that instructors were engaged in OCB yet it was not sufficient to support the function of the universities as well as instructors. Participant 1 reported that he provided professional support for fellow workers who faced difficulty related to their tasks under unusual circumstances. He also gave professional support to the newly employed instructors and reoriented them to adapt to the work environment and feel good about their job. In the same way, Participant 3 and 4 said that they actively participated to improve the effectiveness of the universities by attending important meetings to provide constructive comments and suggestions with the intent to improve the performance of the universities. They also voluntarily participated in the affairs of the universities by offering constructive comments to solve work-related problems. However, other participants reported that they showed a low level of engagement in OCB related to conscientiousness, sportsmanship, and courtesy (Participant 2 & 5). They also reported that they did not exert extra effort beyond the formal duties to achieve the goals of the universities. This means that they were not much interested to support the universities willingly.

3.3. The relationship between servant leadership and Organizational citizenship behavior

As can be seen from Table 7, a positive and statistically significant relationship was observed between servant leadership and perceived OCB with a standardized correlation coefficient (r = .578) by the critical ratio test greater than ± 1.96 at p < .05.This implies that respondents’ scores of servant leadership and perceived OCB were increased concurrently.

Table 7. Correlation coefficient of servant leadership and OCB

Latent variables in the structural model			Unstandardized				Standardized
			Estimate	S.E.	C.R.	P	Estimate
Servant leadership	<–>	OCB	.808	.057	14.068	***	.578

Source: Author own survey 2021.

3.4. The effect of servant leadership on organizational citizenship behavior

As shown in Figure 3 and Table 7 & 8 the path that links servant leadership and OCB with a standardized regression coefficient (β =.578, p < .05) indicates that servant leadership is significantly predicted instructors’ OCB. This means that instructors are more likely to engage in OCB when leaders manifest servant leadership behavior in the workplace. Thus, the value of R² indicated that 33.4 % of the variance in OCB was predicted by servant leadership.

Table 8. Regression of OCB on servant leadership

Latent variables in the structural model			Unstandardized				Standardized
			Estimate	S.E.	C.R.	P	Estimate
OCB	<—	Servant leadership	.555	.028	19.904	***	.578
R-squared			.334

Source: Author own survey 2021.

3.5. One way ANOVA on the perception of instructors in servant leadership and OCB

As revealed in Table 9, the result of one-way ANOVA indicated that a significant difference was found among instructors in their perception of servant leadership at F(3, 736) = 4.204, p = .006.However, there was no significant difference among instructors in their perception of OCB at F(3,334) = .506, p = .678 in the four generations of universities. The mean scores of the 1^st, 2^nd, 3^rd, and 4^th generation universities also confirmed that instructors differed in their perception of servant leadership.

Table 9. One way ANOVA on differences in perception of servant leadership and OCB among instructors in the four generations of universities

Variables	Generations of university	Sum of Squares	df	Mean Square	F	Sig.
Servant leadership	Between Groups	3.875	3	1.292	4.204	.006
	Within Groups	102.622	334	.307
	Total	106.497	337
OCB	Between Groups	.978	3	.326	.506	.678
	Within Groups	215.023	334	.644
	Total	216.001	337

Source: Author own survey 2021.

While a significant difference was observed among instructors in their perception of servant leadership in the four generations of universities, the location of the difference was not identified. As a result, post hoc analysis was conducted to identify which generation universities differed from others at .05 level of significance using the Scheffé test.

Table 10 indicates that there was a significant difference between instructors in the 1^st and 3^rd generation universities in their perception of servant leadership. It also shows that the first generation universities exercise servant leadership than second generation universities. On the other hand, instructors in the 2^nd generation universities did not differ from instructors of the 1^st, 3^rd, and 4^th generation universities in their perception of servant leadership. Similarly, a significant difference was not found between instructors in the 1^st and 4^th as well as between the 3^rd and 4^th generation universities in the perception of servant leadership.

Table 10. Post hoc tests of multiple comparisons among instructors in their perception of servant leadership

Variables	(I) Generations of universities	(J) Generations of universities	Mean Difference (I-J)	Std. Error	Sig.	95% Confidence Interval
						Lower Bound	Upper Bound
Servant leadership	1^st generation	Second generation	.1907	.07370	.084	−.0164	.3978
		Third generation	.2778*	.08271	.011	.0454	.5102
		Fourth generation	.2548	.14991	.411	−.1665	.6760
	2^nd generation	First generation	−.1907	.07370	.084	−.3978	.0164
		Third generation	.0871	.07574	.724	−.1258	.2999
		Fourth generation	.0640	.14617	.979	−.3467	.4747
	3^rd generation	First generation	−.2778*	.08271	.011	−.5102	−.0454
		Second generation	−.0871	.07574	.724	−.2999	.1258
		Fourth generation	−.0231	.15092	.999	−.4471	.4010
	4^th generation	First generation	−.2548	.14991	.411	−.6760	.1665
		Second generation	−.0640	.14617	.979	−.4747	.3467
		Third generation	.0231	.15092	.999	−.4010	.4471

Source: Author own survey 2021.

4. Discussion

4.1. Status servant leadership and organizational citizenship behavior

The finding of this study is similar to the work of Fentahun and Matebe (2020) in that they found servant leadership moderately implemented by the leaders of public universities. Fikre Lobago (2017) also proved the successful implementation of servant leadership in the Compassion International organization worked in Ethiopia, which is consistent with the finding of this study. Similarly, Boume found that servant leadership is an important approach to facilitating the performance of higher education institutions. The result of this study is also similar to the works of Alamri (2011), Kassahun (2015), and Lerra (2015) in that that higher education institutions were led by unethical, authoritative, bureaucratic leaders who want to meet their own needs over the needs of others. In the same way, other studies confirmed that leaders in higher education institution is ineffective, and holds individual needs over those of the institutional goal (Frew et al., 2016).

The finding of this study also indicated that OCB was found to a little or some extent in the workplace. This result is quite similar with the findings of (Akyuz, 2012; Buluç, 2008; Oguz, 2011; Polat & Celep, 2008) on the presence of moderate levels of OCB in workplaces. The finding of this study also supported by Turnipseed and Murkison (2000) that OCB contributes to the organization by creating positive workplace environments. Extra-role behaviors are also essential for an organization, as they are likely to promote more effective communication, which allows best practices to be shared amongst employees or fosters increased coordination among employees (Ren-Tao & Heung-Gil, 2009). Employees who engage in OCB can enhance organizational effectiveness and efficiency by exerting extra effort beyond one’s duties.

4.2. The relationship between servant leadership and organizational citizenship behavior

The result of this study is similar with the findings of Abdu (2014) and Mostafa (2014) that servant leadership is positively related to OCB. Other studies further proved the presence of a positive relationship between servant leadership and OCB (Ehrhart, 2004; Taylor et al., 2007). Research conducted by Setyaningrum (2017) found that servant leadership was positively related to citizenship behavior. In addition, several studies show that there is a significant relationship between servant leadership and OCB in the workplace (Dirks & Ferrin, 2002; Hu & Liden, 2011). Servant leadership can create a servant mindset in employees so that they are encouraged to help colleagues find creative solutions for a given problem.

4.3. The Effect of servant leadership on organizational citizenship behavior

The result of this study is supported by the findings of other researchers concerning the effect of servant leadership on the OCB of employees (Ehrhart & Naumann, 2004; Ruiz‐Palomino et al., 2023; Schlechter & Engelbrecht, 2006). Consistent with this result, Van Dierendonck and Nuijten (2011) and Abid et al. (2015) also found that servant leadership had a significant effect on the OCB of employees. Consistent with this result, Mirshekar and Haddadi (2017) found that the servant leadership approach of leaders significantly predicted employees’ OCB. Moreover, Reed (2016) and Prabowo and Setiawan (2013) reported that servant leadership significantly predicted employees’ OCB. Employees tend to exhibit citizenship behavior when they are cared for and fairly treated by their leaders. The findings of this study were similar to Mira and Margaretha (2012) that hold servant leadership has a positive influence on OCB.

5. Conclusion and implications

5.1. Conclusion

The purpose of this study was to investigate the effect of servant leadership on the OCB of instructors in public universities in Amhara region. Based on the finding of the study, the researcher concluded that the OCB of instructors was influenced by servant leadership practices. Specifically, servant leadership and OCB were practiced moderately in the public universities in Amhara region. This shows that a positive and significant relationship was found between servant leadership and OCB with a standardized correlation coefficient.

5.2. Implications of the study

This research can provide valuable information for higher education officials in encouraging employees to engage in OCB. Specifically, it tries to generate empirical data that indicate the importance of servant leadership in supporting the functions of the universities. This implies that human resource management practices can be instituted to affect efficient and equitable improvement in the behavior of employees. The results of this study generally emphasize the central role played by team commitment in the quest to understand factors that help employees to perform their roles. Servant leadership is also recommended as one of the people-oriented leadership approaches that foster employee development and has a significant effect on team effectiveness when employees feel committed to their work.

As far as my reading, servant leadership and OCB have received little attention in the context of Ethiopia, this research may initiate other researchers to conduct further studies by considering all public and private universities found in different parts of Ethiopia.

6. Limitations of the study

Since the study used a new conceptual framework, sufficient literature has not been found in the context of Ethiopian higher education institutions related to servant leadership and OCB. Because of this, the literature reviewed related to these variables in the context of other organizations is used for this study. In addition, this study consisted of only academic staff in the public universities in Amhara region due to time and resource constraints. Thus, the findings of this study may not be generalized to all Ethiopian public universities.

Acknowledgments

I would like to extend my gratitude to all instructors of Gondar, Wollo, Debre Markos, Debre Tabor, Debark, and Injibara Universities who participated in providing the necessary data for this study. I am also grateful to those who coordinated the data collection process at the research sites. My special acknowledgment also extends to Mohammed Beshir (Ph.D.) for proofreading this article by devoting his precious time.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplemental material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/23311975.2023.2288711

Additional information

Funding

The author did not get any fund to conduct this study.

Notes on contributors

Shimelis Mesfin

Shimelis Mesfin (Ph.D.) is an Assistant professor of Educational Policy and Leadership at the University of Gondar in the Department of Educational Planning and Management. He has been working as a lecturer in higher education institutions since 2009. He has researched leadership, job satisfaction, motivation, organizational commitment, supervision, ICT, etc. And he has published six articles in reputable journals and presented most of his works at national and international conferences. He also wants to conduct further studies in the areas of school improvement, incentive strategies, teachers’ professional competencies, and ethics, organizational justice, and commitment. Therefore, he has planned to provide training for leaders and instructors to improve the implementation of servant leadership and organizational citizenship behavior in the study areas.