Development and Validation of a Spanish Short Servant Leadership Survey (SSLS6-3F) among Spanish Workers in Religious Non-Profit Organizations

Religious non-profit organizations are becoming increasingly important in the third sector in a wide range of countries, where they are currently leading players in different areas, such as education, healthcare, and social work. These organizations have the peculiarity of providing a service to their users while transmitting them the values of their mission statement. An usually employed and effective management strategy for these institutions is a servant leadership style. This article seeks to introduce a theoretical discussion of this leadership approach by providing a Spanish version of an instrument for measuring servant leadership in Spanish religious non-profit institutions. To this end, workers of different Spanish faith-based non-profit organizations of the third sector, a relatively unexplored context, were analyzed after obtaining 463 valid questionnaires. This study used the Spanish translation of a seven-item and three-factor servant leadership scale. An exploratory and confirmatory factor analysis was performed. The results confirm that the six-item and three-factor servant leadership scale was the most effective scale to measure this construct. In conclusion, this research covers a notable research gap by providing a reliable and valid Spanish short version of the servant leadership scale for workers of Spanish religious non-profit organizations.


Introduction
Recent literature has focused on achieving a consensual definition of the concept of servant leadership and finding an instrument to measure it [1][2][3]. Due to the importance of analyzing servant leadership in religious organizations, and the lack of a Spanish short scale for measuring servant leadership in these institutions, the main aim of this research was to develop and validate this instrument. To this end, we tested the reliability and validity of the Spanish translation of a short version (seven items) of a servant leadership scale for workers [4].
Nowadays, religious organizations are significant players in the global economy. They represent a considerable part of the third sector in areas such as education, social services, and health. They contribute toward maintaining the welfare state, representing an essential part of the European economic and social context [5,6]. These organizations have some peculiarities that distinguish them To test the reliability and validity of the Spanish translation of a seven-item version of the servant leadership scale for workers, we utilized an extensive database consisting of workers of different Spanish religious non-profit organizations of the third sector. This investigation covers a primary gap in the literature; to our knowledge, it is the shortest Spanish version of a servant leadership scale for workers. Short scales have a wide range of benefits, such as increasing the response rate and the quality of the responses, and allowing for embedding into more extensive surveys. Furthermore, this scale was tested on religious non-profit organizations, a relatively unexplored context.
The structure of this paper continues with a theoretical discussion on servant leadership, involving defining the term and the selected scale. The methodology section describes the methodology used in detail and the results section presents the model and the test results, verifying the validity and reliability of the scale. Finally, the article evaluates the most relevant empirical outcomes in the discussion section, and ends by summarizing the main conclusions, implications, and limitations of the research.

Servant Leadership Theory
Although several researchers have shown interest in investigating servant leadership [33,34], a generally accepted definition is still lacking [20,23]. Consequently, this study aimed to collect the most accepted definitions [21]. The concept of servant leadership was introduced five decades ago by Greenleaf [19]. According to this author, among the broad range of leadership frameworks, the idea of servant leadership condenses the emotional, moral, and relational areas presented in the management literature. Servant leaders manage organizational challenges by prioritizing organizational stakeholders' interests over personal ones. Developing servant leaders is one of the main goals of servant leadership. Moreover, servant leaders, unlike other leadership styles, see their role as a vehicle for serving workers, the organization, and the community. Greenleaf [19] and Spears [43] conceptualize this desire to help and serve the development of people and groups as receptively listening to others, developing a high level of empathy, trusting more in persuasion than coercion, and committing to building community in the work environment. Spears [17], inspired by Greenleaf [19], described the ten key characteristics of servant leadership: empathy; listening; commitment to the growth of people; building community; stewardship; healing; foresight; conceptualization; persuasion; and awareness of others, situations, and oneself. Sendjaya et al. [44] argue that spirituality is also an important source of motivation for servant leaders.
Servant leadership is a management strategy that prioritizes the needs of employees and turning those needs into goals, placing the employees' good over the self-interest of the leader, and showing concern toward others [22,23,33]. Therefore, servant leadership is based on promoting the value and development of people, the construction of a community, sharing power and status for the common good, and the exercise of authenticity [45]. From this point of view, Reinke [4] states that servant leadership is a relationship, not a position, and defines it as "leadership that puts the needs of others and the organization first, is characterized by openness, vision and stewardship, and results in building community within the organization". This author considers that a servant leader is "committed to the growth of both the individual and the organization, and works to build community within organizations".
Servant leadership shares many of its principles with other leadership theories. Servant leadership and authentic leadership have the idea of the ethical component and the development of followers in common. They also agree on the use of positive modeling and support self-determination as mechanisms to influence the collaborators [19,46]. Servant leadership also shares several aspects with transformational and transactional leadership [47]. Some authors argue that transformational and transactional forms of leadership have so many similarities that it is difficult to differentiate them, while others state that, at least, they have the explicit component of individual consideration in common. Their leaders pay attention to the personal achievement and growth needs of their team members [48]. Hence, since the concept of this dimension is similar to some ideas of servant leadership, it is positively related to transactional leadership [49], as well as transformational leadership [50].
Servant leadership is also related to the concept of authenticity [45]. Van Dierendonck et al. [51] affirm that a servant leader is characterized by authenticity, integrity, courage, objectivity, and humility. Being authentic means acting in accordance with your values and beliefs [52], and servant leaders live their lives according to the values to which they have adhered [53]. As such, servant leaders encourage their followers to be transparent and demonstrate consistency between what they say and do, facilitating the development of the community [44].
Moreover, servant leadership leads to a wide range of positive outcomes. Several authors [24,25] explain that servant leadership has positive effects on work engagement. Consequently, servant leaders take care of their followers, creating an effective and productive working climate by providing the necessary job resources [25]. Servant leadership also focuses on workers' wellbeing [21][22][23] by trying to satisfy their needs [54], their development, and their empowerment [23], even above the interests of leaders [55]. To achieve long-term organizational goals, leaders must facilitate the growth, development, and general wellbeing of their teams [56].

Servant Leadership Scale
The survey employed for this research is the scale of Reinke [4], since this study considers that its concept of servant leadership involves the most relevant and common points of servant leadership theory. These authors built this survey by re-conceptualizing all the characteristics of servant leadership into a multidimensional construct with three dimensions: openness, vision, and stewardship. First, openness encompasses Spears [17] elements of listening, empathy, and awareness of others, in two items (e.g., "I feel comfortable telling my supervisor about departmental problems"). Second, the vision dimension refers to contextualizing situations and looking at them in perspective to plan for and anticipate future needs. This dimension is composed of two items (e.g., "my supervisor emphasizes doing the right thing for the long-term benefit of all"). Third, stewardship includes Spears [17] elements of persuasion, stewardship, and commitment to the growth of people. Stewardship refers to a participatory leadership style in which a servant leader puts the needs of the employees and the organization first and is committed to their growth. This last dimension consists of three items (e.g., "my supervisor puts the employees' needs first before looking out for him or herself"). The questionnaire is included in Appendix A.
Therefore, the scale selection is based both on the conceptualization of the servant leadership of Reinke [4] and on the wide range of benefits it offers. First, it is psychometrically sound and shorter than the three versions of multidimensional servant leadership recommended by Eva et al. [33], which have 28 items [57], 30 items [44], and 35 items [58]. The scale of Reinke [4] is also similar in length to the shortest versions of these scales, which are SL-7 (global servant leadership scale) [34], SLBS-6 (servant leadership behavioral scale) [59], and even less than half the length of SLS (servant leadership survey; 18 items) [60]; these scales have not been selected due to being global measures rather than measuring multidimensional concepts. The scale in Reinke [4] is also shorter than the existing Spanish versions: 14 items [61], 30 items [62], and 36 items [26]. This fact raises the response rate, as larger scales usually decrease the respondent's attention by producing fatigue or boredom. Larger scales may negatively influence the quality of answers, lowering their integrity and subsequent validity [63]. Moreover, short scales can be included in more extensive surveys since longer questionnaires use up time that could be employed for measuring other variables in the same study [63]. Second, the scale of Reinke [4] considers servant leadership as a multidimensional variable, as the theory explains, contrary to other measures recommended by Eva et al. [33]. According to Gefen et al. [64], researchers should evaluate whether each theory-based construct is better represented as a first-order or second-order construct. Thus, this scale, being a short version, allows for capturing the full domain of each dimension. Third, the questionnaire of Reinke [4] was validated using a considerable sample of 254 employees in Georgia. Finally, the original scale of Reinke [4] presented good reliability and validity.

Sample and Data Collection
The target samples were from the following Spanish Catholic religious non-profit organizations. The first sample (hereafter "social centers") was composed of 30 social centers in Andalucia, Canary Islands, and Extremadura. This sample was part of the social sector and included different social intervention projects, such as day services, socio-labor insertion, and support for immigrants. Data collection was carried out in July 2019. The second sample (hereafter "educational centers") was mainly composed of private religious schools in Andalucia. This sample was composed of nine educational centers (schools), three early childhood educational centers, one language school, one sports school, one full training cabinet, one music school, and one employment-training center. Data collection was carried out between May and November 2019. The research was conducted through a Google form survey sent to all workers of the target organizations. Loyola Andalucia University's ethics committee approved the investigation, which was performed according to the Declaration of Helsinki. All questionnaires were anonymous and each participant gave their informed consent.
Of the 1019 (social centers = 499; educational centers = 520) total workers, 514 (social centers = 283; educational centers = 231) answered the survey, resulting in a valid response rate of 52.1% (260 valid responses) for social centers and 39.0% (203 valid responses) for educational centers. The percentage of valid responses in the manager position was 23.8% for social centers and 18.8% for educational centers, where the rest were from non-manager workers. Most of those surveyed were women (72.4% in social centers and 68.8% in educational centers). The average age and seniority were 38.1 and 4.6 years, respectively, for social centers, and 41.0 and 11.1 years, respectively, for educational centers. Most of the workers had completed higher degree studies (university degree, master's degree, or PhD): 91.2% in social centers and 94.1% in educational centers. Table 1 displays the main demographic data.

Measures
This study applied the follower version of leadership questionnaires. This means that the employees evaluated their supervisors. The servant leadership scale employed in this research was the Spanish translation (using a standard back-translation procedure; the back translation matched the original items) of Reinke [4]. It was composed of seven items that measured three dimensions (openness: I1, I2; vision: I3, I4; and stewardship: I5, I6, I7) according to a Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). The questionnaire is included in Appendix B.
Furthermore, to obtain additional evidence regarding the servant leadership scale validity, on the one hand, authentic leadership was measured by Walumbwa et al. [65] using the Spanish translation of the ALQ (Authentic Leadership Questionnaire), consisting of 16 items. It was designed to evaluate four Sustainability 2020, 12, 3766 6 of 18 components (self-awareness, internalized morals, balanced processing, and relational transparency) using a Likert scale from 1 (nothing) to 5 (always or almost always). On the other hand, transactional leadership was measured by Molero et al. [48] using a questionnaire that evaluates two dimensions (individualized consideration and contingent reward), employing a Likert scale from 1 (never) to 5 (almost always). Moreover, authenticity was assessed using the IAM (Individual Authenticity Measure at work), developed by Van den Bosch and Taris [52]. It includes three dimensions (authentic living, accepting external influence, and self-alienation), which are evaluated using a Likert scale that ranges from 1 (totally disagree) to 7 (totally agree).
To corroborate the discriminant validity, the Spanish version (developed by Benevides-Pereira et al. [66]) of the UWES (Utrecht Work Engagement Scale) was used. This scale evaluates the three dimensions that constitute this construct (absorption, dedication, and vigor). The questionnaire uses a Likert scale ranging from 1 (never) to 7 (always). Additionally, a scale developed by Diener et al. [67] was employed to measure subjective wellbeing. Three dimensions were evaluated using a Likert scale: satisfaction with life (Satisfaction With Life Scale (SWLS)) from 1 (totally disagree) to 7 (totally agree); positive and negative experiences (Scale of Positive and Negative Experience (SPANE)) from 1 (very rarely or never) to 5 (very often or always); and flourishing (Flourishing Scale (FS)) from 1 (totally disagree) to 7 (totally agree), which is an adaptation to the work context that was developed by Mendonça et al. [68].

Data Analysis
The analyses were performed using the statistical software IBM SPSS 25 and STATA/SE 16.0. First, the adequacy of the seven items in the servant leadership scale (hereafter called the SSLS7) developed by Reinke [4] was analyzed using Pearson correlations calculated in SPSS. Inter-item correlations were analyzed. Later, the analysis performed for checking the internal reliability was executed using SPSS. The analyses revealed that the scale would work better after deleting one item. Second, an exploratory factor analysis (EFA) of the resulting six-item scale was completed using SPSS. Third, to test the validity of the scale, four empirical models of structural equations for servant leadership (SSL7-3F, SSL7-1F, SSL6-3F, and SSLS6-1F; please refer to Section 4.3) were built by employing STATA to allow for confirmatory factor analysis (CFA). Loadings and model fit tests showed that the most appropriate model was SSLS6-3F. Finally, the convergent and discriminant validities were analyzed by employing Pearson correlations calculated in SPSS, and criterion-related validity using t-tests performed in SPSS.

Adequacy of the Data and Internal Consistency of the SSLS7
First, the adequacy of the data, i.e., the items´significance, was analyzed by inspecting the Pearson correlation matrix. As seen in Table 2, all coefficients were significant and surpassed the recommended threshold of 0.3 for testing the internal consistency of the scale, where all coefficients were higher than 0.44, except those related to one item (I4), which were between 0.1 and 0.2. Additionally, as far as the questionnaire's reliability was concerned, the internal consistency of the scale and subscales was analyzed. Due to the low Cronbach's alpha (0.52) of the vision dimension presented in the original scale [4], we checked the internal consistency of the scale not only with Cronbach´s alpha but also with "Cronbach´s alpha if the item was deleted". The results revealed that vision was the only dimension that did not surpass (Cronbach's alpha = 0.30) the limit of internal reliability of 0.7 suggested by Carmines and Zeller [69]; the "Cronbach´s alpha if the item was deleted" corresponding to I4 was 0.91 (compared to a total Cronbach´s alpha of 0.86 when including the seven items).

Exploratory Factor Analysis (EFA)
EFA is performed to examine the dimensionality of a scale. Due to the results of the internal consistency tests, an EFA was performed for the Short Servant Leadership Six-Item Scale (SSLS6), with I4 removed. The six items were subjected to principal components factor analysis (PCFA). The results revealed that the data were adequate for the PCFA (determinant = 0.14; Kaiser-Meyer-Olkin (KMO) = 0.871; Bartlett´s sphericity test showed statistical significance: χ 2 (21) = 1902.7 with p-value = 0.000). Although the PCFA revealed one factor that explained 69.19% of the variance, which we did not consider enough, following the suggestions of Reinke [4], we also performed EFA analysis, extracting three factors (3F). The three factors model explained 88.15% of the total variance: factor 1 (69.51%), factor 2 (12.31%), and factor 3 (6.32%). Factor 1 was composed of I1 and I2, factor 2 was composed of I3 and I5, and factor 3 was composed of I6 and I7. As can be observed, I5 was part of the vision dimension in SSLS6-3F, together with I3, as revealed by the correlation´s matrix, components graph, and rotated components matrix.
Furthermore, after obtaining these results, the reliability of SSLS7 and SSLS6 was compared. The reliability analysis performed in SPSS consisted of calculating the internal consistencies of the scale and subscales using Cronbach's alpha coefficient. It revealed that SSLS6-3F presented a very good internal consistency (openness: 0.83; vision: 0.87; stewardship: 0.83; SSLS6 total scale: 0.91), and was better than the SSLS7-3F (openness: 0.83; vision: 0.30; stewardship: 0.87; SSLS7 total scale: 0.86). The split-half reliability was also satisfactory for SSLS6-3F, with a Spearman-Brown coefficient of 0.846.

Confirmatory Factor Analysis (CFA)
A CFA confirms if a model represents a construct well. Therefore, to test the construct validity of the short servant leadership scale and confirm its dimensionality, we followed a confirmatory approach. CFA allows models to be driven both statically and theoretically. As a consequence of the internal consistency tests and EFA, we performed a CFA by comparing the seven-item (the original version of the scale from Reinke [4]) and the six-item (suggested by the article results) versions, for three correlated factors, as proposed by the theory, and one factor, as indicated by the first step of the PCFA. Then, the CFA was carried out to test the following models: The variables of the model were not expected to be normally distributed after performing the Kolmogorov-Smirnov and Shapiro-Wilks tests. Hence, estimation was conducted using an asymptotically distribution-free method, as it is more sensitive to a non-normal distribution of scores [70,71]. The bootstrapping performed for the four models revealed that all items presented factor loadings greater than 0.707 [69], except for I4, which showed a factor loading equal to 0.2 for SSLS7-3F and SSLS7-1F. This circumstance suggested that the six-item models (both one-dimensional and three-dimensional) were more effective at measuring servant leadership. Finally, although both six-item models presented high factor loadings for all the items, SSLS6-3F achieved better results (see Table 3). During the analysis of the goodness of fit indices (see Table 4), the following tests were performed: χ 2 tests, the goodness of fit index (GFI), the comparative fit index (CFI), the root mean square error of approximation (RMSEA), the standardized root mean squared residual (SRMR), and the Tucker-Lewis index (TLI). All models obtained a coefficient of determination (CD) higher than 0.9 (SSLS7-1F: 0.926, SSLS7-3F: 0.982, SSLS6-1F: 0.926, and SSLS6-3F: 0.986). First, we considered the ratio χ 2 /df in this analysis since several authors [72,73] have recommended it for large samples. The literature suggests that this index should be in the range of up to 5 [74], although Hu and Bentler [75] and Kline [76] consider a limit of 3. As can be seen in Table 4, SSLS6-3F was the only model that fulfilled both conditions. GFI is a measure of the relative amount of variance accounted for by the model, while CFI is a population measure of a model's misspecification. Values higher than or equal to 0.9 indicate a good fit [75,77]. SSLS6-3F is the only model that surpassed the 0.9 criterion for both GFI and CFI. CFI, together with TLI, comprise a comparative fit index, such that a value close to 1 indicates a good fit [73]. SSLS6-3F was the model with CFI and TLI values that were closer to 1. On the other hand, RMSEA measures the discrepancy per degree of freedom and values smaller than 0.08 indicate an acceptable fit [75,78]. Out of the four models, only SSLS6-3F fulfilled this criterion. Lastly, SRMR is a measure of the average difference between the observed and predicted correlations in the model. Hu and Bentler [75] suggest values close to 0.08 or below. SSLS6-3F, together with SSLS7-3F, did not meet this threshold. In conclusion, the data showed a better fit for the SSLS6-3F model, presenting a good or acceptable validity with all fit indices, whereas SSLS6-1F, SSLS7-3F, and SSLS7-1F did not seem valid (see Table 4). Figure 1 summarizes the results of the CFA for SSLS6-3F. models, only SSLS6-3F fulfilled this criterion. Lastly, SRMR is a measure of the average difference between the observed and predicted correlations in the model. Hu and Bentler [75] suggest values close to 0.08 or below. SSLS6-3F, together with SSLS7-3F, did not meet this threshold. In conclusion, the data showed a better fit for the SSLS6-3F model, presenting a good or acceptable validity with all fit indices, whereas SSLS6-1F, SSLS7-3F, and SSLS7-1F did not seem valid (see Table 4). Figure 1 summarizes the results of the CFA for SSLS6-3F.

Convergent and Discriminant Validity
To check that the scale behaved correctly, convergent validity refers to the extent to which the SSLS6-3F factors were correlated with each other and with general related concepts. Convergent validity was assessed by checking the correlations of the SSLS6 total scale and its corresponding dimensions (openness, vision, and stewardship) with a similar scale (the original SSLS7 validated by Reinke [4]), and with validated measures of similar constructs (authentic leadership, transactional leadership, and authenticity). The SSLS6 total scale, together with its three dimensions, were strongly and significantly correlated between them, and as expected, with the original version SSLS7 complete scale (see Table 5). There were also positive and significant relationships between the SSLS6 total scale and its corresponding dimensions, with authentic and transactional leadership, as well as with authenticity, where the correlation was especially strong with both leaderships and a medium strength with authenticity (see Table 5). This indicates that the more a person perceived a higher level of servant leadership in their supervisors, the more they were inclined to report higher scores in perceiving authentic leadership and transactional leadership, as well as experiencing a higher level of authenticity at work. Hence, these relationships, which were in the same direction as the theory (see Section 2.1), proved that the scale behaved correctly.

Convergent and Discriminant Validity
To check that the scale behaved correctly, convergent validity refers to the extent to which the SSLS6-3F factors were correlated with each other and with general related concepts. Convergent validity was assessed by checking the correlations of the SSLS6 total scale and its corresponding dimensions (openness, vision, and stewardship) with a similar scale (the original SSLS7 validated by Reinke [4]), and with validated measures of similar constructs (authentic leadership, transactional leadership, and authenticity). The SSLS6 total scale, together with its three dimensions, were strongly and significantly correlated between them, and as expected, with the original version SSLS7 complete scale (see Table 5). There were also positive and significant relationships between the SSLS6 total scale and its corresponding dimensions, with authentic and transactional leadership, as well as with authenticity, where the correlation was especially strong with both leaderships and a medium strength with authenticity (see Table 5). This indicates that the more a person perceived a higher level of servant leadership in their supervisors, the more they were inclined to report higher scores in perceiving authentic leadership and transactional leadership, as well as experiencing a higher level of authenticity at work. Hence, these relationships, which were in the same direction as the theory (see Section 2.1), proved that the scale behaved correctly.
Discriminant validity refers to the extent to which a set of variables are correlated with their variables outcomes. Discriminant validity was assessed through the correlations (all of them significant at p < 0.01 and positive) between the SSLS6 total scale and its corresponding dimensions, with the outcomes, work engagement, and the three dimensions of subjective wellbeing (flourishing, satisfaction with life, and the presence or absence of positive and negative feelings). Correlations with the work engagement total scale and flourishing were of medium strength, while the correlations with satisfaction with life and positive and negative feelings were low. These correlations, matching with the literature (see Section 2.1), proved the discriminant validity of the SSLS6-3F. Authentic leadership total scale

Criterion-Related Validity
For further validity testing, additional relations with the constructs, such as demographics, can be analyzed. Criterion-related validity was assessed through several t-tests that were performed for openness, vision, and stewardship (the dimensions of SSLS6-3F). First, the t-tests that were performed with the sex and position variables revealed that there was no significant difference between the way that men and women, and managers and non-manager employees, perceived the servant leadership. However, if workers performed their activities in social centers rather than educational centres, they seemed to recognize a higher servant leadership in the vision dimension (t = 2.686, df = 457, p < 0.01) and stewardship dimension (t = 4.698, df = 455, p < 0.001); however, a significant difference in the openness dimension was not shown.

Discussion
Servant leadership is probably a very beneficial and valuable leadership style for religious organizations, which constitute a fundamental part of the third-sector's economy in Spain. Therefore, this research aimed to provide a Spanish instrument for measuring this leadership style in Spanish workers of religious non-profit organizations. This article has extended previous knowledge with an improved version of the servant leadership survey from Reinke [4] in Spanish, which has been tested in a sample of workers of religious non-profit organizations. Starting with a parsimonious model, the original version of Reinke [4] (composed of seven items and three dimensions), we performed a standard back-translation procedure from English to Spanish. The data from the Spanish SSLS7-3F showed proper adequacy and internal consistency, except for one item, which was deleted. Then, with the resulting six-item scale, the EFA and the CFA verified that the SSLS6-3F is an adequate scale and offered the best fit. The resulting SSLS6-3F not only had one less item but also changed one item from the stewardship dimension to the vision dimension, which is discussed below. Hence, the resulting SSLS6-3F fulfilled the requirements of convergent validity, where the scale behaved as expected since it was positively and significantly related to similar constructs (authenticity, authentic leadership, and transactional leadership). As far as the discriminant validity is concerned, the scale was also positively and significantly related to the outcomes of work engagement and subjective wellbeing, which is consistent with previous studies in this area. Lastly, the criterion-related validity was tested through several t-tests, demonstrating that the activity, level of studies, age, seniority, authenticity, work engagement, and subjective wellbeing did influence the way that workers perceived servant leadership, while sex and position did not.
Some advantages and improvements of the SSLS6-3F scale in comparison to the questionnaire of the scale from Reinke [4] are the following. First, it is shorter since one item was removed that did not work well in the scale and caused insufficient reliability in the vision dimension. Second, it improved the reliability of the total scale, as well as the reliability of the vision dimension. Third, it was tested using a larger sample of workers and different types of organizations. Fourth, it presented a more thorough scale validation process. Fifth, the validity was tested through correlations with a larger number of similar concepts and outcomes.
To our knowledge, this article also provides a wide range of advantages, concerning the servant leadership research. First, SSLS6-3F is the shortest Spanish version of a servant leadership scale. Second, with just six items, it is a multidimensional construct that integrates the essential components of servant leadership (openness, vision, and stewardship). Third, it was tested on a large sample of workers (unlike other Spanish leadership scales that have been tested in other populations, such as students), as well as in a relatively unexplored context, namely religious non-profit organizations that operate in the third sector.
An interesting and significant result of the present study is the shift of an item (I5: "My supervisor is committed to helping employees grow and progress") from the stewardship dimension to the vision dimension. From a theoretical point of view, a possible explanation could be that helping employees' growth and progress is a way of planning future needs and keeping situations in perspective, which is more the aim of the vision dimension than the stewardship dimension [4]. I5 is also more related to acting toward the benefit of all (I3: "My supervisor emphasizes doing the right thing for the long-term benefit of all") than to put the needs of others before oneself, which is the idea behind I6 and I7 (the items that conform to the stewardship dimension). In fact, other authors also do not include the idea of employee growth and progress as part of the stewardship dimension, understanding it as the pursuit of the common good, beyond the leader's self-interest [79,80]. Future studies should test the dimensionality of the scale again and inquire into this reflection.
This investigation supports the idea that servant leadership is not opposed to other management styles, such as authentic leadership and transactional leadership, by showing a significant and strong positive correlation between them. This idea is similar to research by other authors who affirm that servant leadership and authentic [19,46] and transactional [49] leaderships are constructs with many similarities. This means that they are complementary leadership styles in religious organizations, not only because they share similar ideas but also because their workers perceive the three of them in a positive and valued way. This fact leads us to think that the three of them should be linked in the management style of religious organizations due to the characteristics that precede them. Religious organizations are concerned not only with the service they provide, but also with other important aspects, such as the way they provide their service, or how their employees live and feel their work. Therefore, their leaders need to be committed at three levels. First, they need to be committed to the service of their teams, seeking the growth of their employees. Second, they need to be committed to the mission and cause of why they perform their activity. The leaders must have a behavior that is consistent with their beliefs and speech. Third, they must act with justice. The leaders must show equanimity in recognizing and rewarding team members. Indeed, it is an advantage if the three leadership styles coexist in religious organizations, while there would be something lacking in the management strategy if the leader is not perceived as a servant displaying authentic and transactional characteristics.
The positive and significant correlations between servant leadership and authenticity, work engagement, and subjective wellbeing are also highly relevant. The objective of these institutions is based both on the provision of their service and on how they perform this work, transmitting their character and charisma. In this sense, religious organizations need to have engaged workers who feel authentic. They look for authentic workers who share the values of the organization and can transmit them while providing their service. Moreover, having disengaged workers with a low level of subjective wellbeing is a long-term survival risk for these institutions since religious organizations are usually less competitive on the market in terms of attracting and retaining talented workers [5,6]. According to Latif and Marimon [26], servant leadership and life satisfaction revealed a negative relationship in their recent study, in contrast with the theory [27,28]; therefore, they encouraged examining the relationship between them. Toward this end, this study showed a positive and significant correlation between servant leadership and satisfaction with life, confirming the previous studies of the theoretical framework and favoring the employment of this management style in institutions where the management of emotions attains a particular meaning.
In many organizations, there is still a long way to go before implementing servant leadership. The point here is that while other types of organizations might look for different management styles depending on their purposes, faith-based entities might need to develop a servant leadership strategy to fulfill their mission. For instance, nowadays, some religious organizations still need to change their structure, from hierarchical to horizontal and participatory to be able to implement servant leadership [81]. Moreover, religious institutions are not the only ones having difficulties implementing this strategy since other third-sector entities are also struggling. Servant leadership is a management approach that is capable of dealing with the changing environment [82]. Some entities of the third sector are still not aware of the fast changes that the economy and world are suffering; therefore, it is important to have an instrument that allows them to analyze servant leadership. With this measure, organizations will be able to evaluate how their employees perceive their servant leadership strategy and consequently improve their performance.

Conclusions
This article provides the shortest Spanish scale for measuring the multidimensional concept of servant leadership in workers. One of the main strengths of this questionnaire is that it is easy to administer and can be combined with other instruments, as well as used in longitudinal studies. Moreover, to our knowledge, it utilized one of the largest samples of workers for the validation of this servant leadership scale, which investigated the relatively unexplored context of religious non-profit organizations. We conclude that SSLS6-3F reported satisfactory reliability and validity, and was able to measure servant leadership very quickly and very accurately.

Limitations and Future Research Directions
This research, despite its strengths, has some limitations that should be highlighted. First, this is a cross-sectional study, and as such, the findings represent a snapshot scenario and the stability of the scale across time cannot be confirmed. Consequently, it could be worthwhile for future research to develop longitudinal studies for test-retest purposes. Second, the study employed a sample of Spanish workers of religious non-profit organizations within the third sector; further studies among different samples of employees would increase the external validity of the scale, such as employees of other industries, for-profit organizations, or other developed and developing countries. Moreover, although this model was tested in Christian religious organizations, future research could validate the scale in organizations of other religions, such as Islam, Judaism, Orthodox, etc. Future investigations could also corroborate the dimensionality of the scale. Third, the validity of the scale was tested through correlations between the servant leadership scale and authentic and transactional leadership, authenticity, work engagement, and subjective wellbeing; several t-tests were also performed on the SSLS6-3F results regarding activity (private schools or social centers), level of studies, age, seniority, sex, position, authenticity, work engagement, and subjective wellbeing. However, future research could study the relationship of the SSLS6-3F with other variables, such as trust, career satisfaction, career commitment, empowerment at work, job stress, or work-life enrichment.