ArticlePDF Available
Servant leadership: Validation of a short form of the SL-28
Robert C. Liden
, Sandy J. Wayne
, Jia Hu
, Junfeng Wu
, Chenwei Liao
University of Illinois at Chicago, USA
University of Notre Dame, USA
Michigan State University, USA
article info abstract
Article history:
Received 15 November 2013
Received in revised form 23 August 2014
Accepted 18 December 2014
Available online 15 January 2015
Handling Editor: William Gardner
Although research on servant leadership has been expanding over the past several years,
a concise, valid scale for assessing global servant leadership has been lacking. In the current
investigation a 7-item measure of global servant leadership (SL-7), based on Liden, Wayne,
Zhao, and Henderson's (2008) 28-item servant leadership measure (SL-28), is introduced.
Psychometric properties of the SL-7 were assessed at the individual level with data collected
from 729 undergraduate students, 218 graduate students, and 552 leaderfollower dyads
from 11 organizations, and at the team level with a study consisting of a total of 71 ongoing
intact work teams. Results across three independent studies with six samples showed
correlations between the SL-7 and SL-28 scales ranging from .78 to .97, internal consistency
reliabilities over .80 in all samples, and signicant criterion-related validities for the SL-7
that parallel those found with the SL-28.
© 2014 Elsevier Inc. All rights reserved.
Servant leadership
Scale development
Transformational leadership
Leadermember exchange
In response to the increasing need for employee engagement, creativity, and sharing among coworkers, as well as societal
demands for higher levels of ethical behavior in organizations, servant leadership has emerged as a desirable approach to leadership,
because it promotes integrity, focuses on helping others, and prioritizes bringing out the full potential of followers. Indeed, servant
leadership offers the promise of combatting negative outcomes associated with promoting one's self-interest (O'Reilly, Doerr,
Caldwell, & Chatman, 2014), which appears to underlie many incidents of unethical behavior (Hoogervorst, De Cremer, & van
Dijke, 2010). Servant leadership, although alluded to in ancient philosophy, was introduced in the contemporary vernacularby Robert
Greenleaf (1970) in his classic essay. Greenleaf stressed the importance of leaders prioritizing the support and development of
followers, accomplishing this by setting an example through demonstrating honesty, compassion, and hard work. Servant leadership
is unique relative to other approaches to leadership for its prioritization of serving followers before attending to one's own needs,
acting as a servant leader in all realms of life work, home and community and developing followers into servant leaders.
Although Greenleaf's (1970) essay appeared prior to the introduction of the most widely studied approaches to leadership (Dinh
et al., 2014), transformational leadership (TFL, Bass, 1985) and leadermember exchange (LMX; Dansereau, Graen, & Haga, 1975),
empirical scientic investigation of servant leadership was not actively pursued by researchers until after the publication of the
seminal work by Mark Ehrhart (2004). Over the past 10 years, empirical research has demonstrated the incremental value of servant
leadership as evidenced by the explanation of additional variance beyond TFL, LMX, and/or consideration/initiating structure
(Fleishman, 1998) in individual (Liden, Wayne, Zhao, & Henderson, 2008; Neubert, Kacmar, Carlson, Chonko, & Roberts, 2008; van
The Leadership Quarterly 26 (2015) 254269
Corresponding author at: Department of Managerial Studies MC 243 University of Illinois at Chicago 601 S. Morgan Chicago, IL 60607.
E-mail address: bobliden@uic .edu (R.C. Liden).
1048-9843/© 2014 Elsevier Inc. All rights reserved.
Contents lists available at ScienceDirect
The Leadership Quarterly
journal homepage:
Dierendonck, Stam, Boersma, de Windt, & Alkema, 2014), group (Ehrhart, 2004; Schaubroeck, Lam, & Peng, 2011), and organizational
(Peterson, Galvin, & Lange, 2012) outcomes. The strength and consistency of the incremental variance demonstrated in these studies
has served to legitimize servant leadership as a construct worthy of continued research attention.
Servant leadership has been presented as a multidimensional construct (Ehrhart, 2004; Liden et al., 2008; van Dierendonck &
Nuijten, 2011). The dimensions uncovered by Liden and colleagues are: 1) emotional healing, which involves the degree to which
the leader cares about followers' personal problems and well-being; 2) creating value for the community, which captures the leader's
involvement in helping the community surroundingthe organization as well as encouraging followers to be active in thecommunity;
3) conceptual skills,reecting the leader's competency in solving work problems and understanding the organization's goals;
4) empowering, assessing the degree to which the leader entrusts followers with responsibility, autonomy, and decision-making
inuence; 5) helping subordinates grow and succeed, capturing the extent to which the leader helps followers reach their full potential
and succeed in their careers; 6) putting subordinates rst, assessing the degree to which the leader prioritizes meeting the needs of
followers before tending to his or her own needs; and 7) behaving ethically, which includes being honest, trustworthy, and serving
as a model of integrity.
As with any multidimensional construct, it is important to specify the nature of relationships between each dimension
and the overall or global construct (Law,Wong,&Mobley,1998). Law et al. (1998, p. 743) specify three relational formsin
which dimensions relate to the overall construct: 1) latent, in which the construct exists at a deeper conceptual level than
its dimensions;2) aggregate, describing constructs that can be formed as an algebraic function of its dimensions;and 3) prole,
which are constructs that are based on levels of each of the dimensions.Servant leadership is best represented by the aggregate
model, as it is a construct that consists of the sum of its dimensions. Indeed, none of these dimensions alone, or even subsets
of the dimensions, adequately capture the complexity of global servant leadership. In contrast to the aggregate model where
the construct is formed from its dimensions, Law et al. (1998) describe the latent model as the focal construct leading to its
dimensions. We contend that servant leadership is not a higher-level construct that underlies its dimensions because the
dimensions of servant leadership are not representing the same construct with different degrees of accuracy but instead are
capturing different aspects of leader behavior. Servant leadership also does not tthetypologyforaprole multidimensional
construct. According to Law et al. (1998),aprole construct can only be interpreted via a set of proled characteristics of
the dimensions; there is not a single theoretical overall construct that summarizes and represents all of the dimensions
(p. 747). This is not the case for servant leadership as there is theory and empirical research that supports the overall construct.
Servant leadership is thus best described as the combination of its dimensions. It captures honest leaders who put the needs of
followers rst, promote helping in the larger community as well as at work, and possess the technical skills necessary to provide
meaningful help to followers.
Consistent with its identity as an aggregateconstruct (Johnson, Rosen, & Chang, 2011; Law et al., 1998), it is typically operation-
alized in its global form. This is appropriate given that servant leadership theory describes a comprehensive approach to leading. For
example, although it contains an ethical behavior component, it extends far beyond ethical leadership (Brown, Treviño, & Harrison,
2005), by specifying the multiple elements necessary for serving followers. Extending beyond TFL, leaders not only individualize
their consideration for followers but put the fulllment of follower needs ahead of satisfying their own personal needs. Similarly,
servant leadership transcends LMX by not only providing support to followers, but attending to the personal needs of followers
that may go beyond the work setting. While it is useful to explore specic research questions with the dimensions separately,
capturing the full essence of the construct requires the combination of all of its dimensions.
When focusing on overall servant leadership, each dimension must be captured in a global measure, but not as many items are
needed as when each dimension is analyzed separately. Indeed, at least three items per dimension are required when dimensions
are assessed individually for the purpose of accurately estimating internal consistency reliability of eachdimension.A global measure
requires a total of at least three items for the same reason, but does not require three or more items per dimension. In fact, Credé,
Harms, Niehorster, and Gaye-Valentine (2012) argued that long scales may have the unintended effect of reducing respondent
attention when reading items, thus lowering the integrity of responses and subsequent validity. In sum, there are both theoretical
and design factors that favor concise global measures.
Given the consistent support for relationships between servant leadership and important individual, team/unit, and organi-
zational outcomes, researchers have focused their attention on better understanding both the processes through which servant
leadership affects outcomes as well as the antecedents of servant leadership. Pursuing such an ambitious research agenda
requires sound measurement of servant leadership. Although a number of measures have been introduced, the scale developed
by Liden et al. (2008) has been frequently used due to the rigorous methods employed in its development (van Dierendonck,
2011). One limitation of this measure, however, is its 28-item length. The large number of items not only takes time that
could otherwise be used for measuring additional variables, but also may introduce fatigue or boredom among respondents,
which may negatively inuence the quality of the responses obtained (Credé et al., 2012).Atthesametime,Credé et al.
(2012) stress that short scales may compromise validity when not developed using rigorous methods. Thus, in introducing a 7-
item unidimensional version of the original SL-28 scale (Liden et al., 2008), labeled SL-7,we followed the recommendations of
Credé et al. (2012) for creatingshort versions of scales. These scholars emphasize the need for and assessment of several forms of va-
lidity, including criterion-related validity, as well as utilizing multiple diverse independent samples. Our ultimate goal was to develop
a global measure of overall servant leadership that is substantially shorterthan the original SL-28, but at the same time captures each
of the seven dimensions assessed in the full 28-item scale. A second purpose of the current investigation is to provide additional
validation of the full 28-item measure, which may be the measure of choice when research questions require assessing servant
leadership dimensions separately.
255R.C. Liden et al. / The Leadership Quarterly 26 (2015) 254269
Servant leadership measurement
Liden et al.'s (2008) scale development effort commenced with the identication of nine servant leadership dimensions based on
the servant leadership philosophy introduced by Greenleaf (1970), as well as the conceptual work by Barbuto and Wheeler (2006),
Graham (1991),Page and Wong (2000), and Spears and Lawrence (2002).Liden et al. (2008, p. 165) indicate that capturing the
domain of the construct was the key consideration when generating items. Specically, they state that Each member of our research
team independently reviewed the full list of potential items, selecting those items that would best capture each servant leadership
dimension.These researchers then subjected 85 items to exploratory factor analyses and seven factors emerged. After rotation,
the four items with the highest factor loadings for each factor were retained, resulting in a 28-item measure capturing seven servant
leadership dimensions.
At the individual level, research has demonstrated that servant leadership is positively related directly or indirectly to a wide
range of outcomes, including individual self-efcacy, job performance, engagement, organizational citizenship behaviors, community
citizenship behaviors, organizational commitment, commitment to the supervisor, creativity, customer service behaviors, and turn-
over intentions (Liden, Wayne, Liao, & Meuser, 2014; Liden et al., 2008; Neubert et al., 2008; van Dierendonck et al., 2014;
Walumbwa, Hartnell, & Oke, 2010). Team- and unit-level research also has demonstrated relationships between servant leadership
and team outcomes, such as team potency, team/unit performance, customer satisfaction, and team-level organizational citizenship
behaviors (OCBs; Ehrhart, 2004; Hu & Liden, 2011; Hunter et al., 2013; Liden, Panaccio, Meuser, Hu, & Wayne, 2014; Liden, Wayne,
Liao, & Meuser, 2014; Schaubroeck et al., 2011). Peterson et al. (2012) have also found CEOs' servant leadership to be positively related
to organizational performance as assessed by return on investment.
Although Liden et al.'s (2008) research presenting the psychometric development of the scale assessed each of the seven
dimensions separately, virtually all subsequent research using the scale has summed the items for use as a global measure of servant
leadership. Support for summing all 28 items to create a global servant leadership measure was provided through higher-order
conrmatory factor analyses (Hu & Liden, 2011). Studies employing Ehrhart's 14-item servant leadership scale have similarly
summed all items to form a global measure (e.g., Hunter et al., 2013; Walumbwa et al., 2010). Given the interest among researchers
to investigate global servant leadership, it would be desirable for researchers to have a scale that is psychometrically sound, yet
shorter than the existing 14- (Ehrhart), 28- (Liden et al., 2008), and 30-item (van Dierendonck & Nuijten, 2011) servant leadership
scales. Our main purpose in the current investigation was to provide such a scale.
Overview of the studies
In our effort to develop a short version of the servant leadership scale, we were sensitive to evidence demonstrating that addingor
deleting items from established scalescan severely compromise validity (Bono & McNamara, 2011; Keller & Dansereau, 2001). In con-
structing the abbreviated scale of servant leadership, we were guided by the advice of Smith, McCarthy, and Anderson (2000),who
identied common errors made by researchers when constructing shorter scales from longer scale parents. The rst step was to select
a subset of items from Liden et al.'s (2008) 28-item multidimensional servant leadership scale, abbreviated as SL-28.In order to
maximize domain coverage of the servant leadership construct, one item was selected from each of the seven dimensions based
upon the exploratory factor analysis loadings reported in the 2008 article. In cases where there was a tie for the highest loading on
a particular dimension, the item from that dimension with lower cross-loadings on the other 6 dimensions (factors) was selected.
We also compared the factor loadings from one previously published study (Hu & Liden, 2011) that used the full 28-item scale. The
selected items had the highest average factor loading on their respective dimensions among the three datasets. Given the robust
Table 1
Comparison of SL-7 and SL-28 psychometrics in Samples 13.
Item number sItem CFA standardized factor loadings
SL-28 SL-7 Sample 1 Sample 2 Sample 3
9 1 My leader can tell if something work-related is going wrong. .51 .41 .63
17 2 My leader makes my career development a priority. .70 .74 .82
1 3 I would seek help from my leader if I had a personal problem. .63 .72 .80
5 4 My leader emphasizes the importance of giving back to the community .55 .59 .65
22 5 My leader puts my best interests ahead of his/her own. .71 .83 .92
15 6 My leader gives me the freedom to handle difcult situations in the way that I feel is best. .61 .53 .72
27 7 My leader would NOT compromise ethical principles in order to achieve success. .54 .52 .64
.95 .80 Sample 1 Cronbach alpha CFI .97 .99 .96
.95 .81 Sample 2 Cronbach alpha SRMR .03 .03 .04
.97 .89 Sample 3 Cronbach alpha RMSEA .06 .04 .10
Chi-square (df) 45.36 (14) 20.03 (14) 32.21 (14)
Chi-square signicance b.01 .13 b.01
Pearson correlation between SL-7 and SL-28 .95 .95 .97
Note. SL-7 item 1 has been revised, adding work-relatedto more closely align it with the original meaning as a conceptual skills item rather than an emotional or
personal item. Sample 1 = undergraduate student sample (N = 598); Sample 2 = graduate student sample (N = 218); and Sample 3 = undergraduate students
(N =131). Pearson correlati on calculated between SL-28 and SL-7 within sample. Response scale: strongly disagree = 1; disagree = 2; slightly disagree=3;
neutral = 4; slightly agree = 5; agree = 6; and strongly agree = 7.
256 R.C. Liden et al. / The Leadership Quarterly 26 (2015) 254269
scale development procedures used in Liden et al.'s (2008) original multidimensional scale development as highlighted by van
Dierendonck (2011), parsimony dictated that a short servant leadership scale should be drawn from these preexisting items rather
than by crafting seven novel items. The resulting items contained in the SL-7 are presented in Table 1.
In order to ensure that the short version maintains the psychometric integrity of the full 28-item version (Liden et al., 2008), we
utilized rigorous scale development procedures in multiple studies (as per Credé et al., 2012). Specically, we compared the SL-7 and
SL-28 in terms of their reliability and validity in three separatestudies containing six independent samples. As with the original scale
development for the SL-28 (Liden et al., 2008), a 7-point strongly disagree to strongly agree response scale was used in all samples.
Given the focus of the developers (Lidenet al., 2008) of the original 28-item scale on ensuring that each item captured the intended
domain, we anticipate that reliability and validity results for the SL-7 should mirror ndings for the full 28-item scale. First, concerning
the psychometric properties of the SL-7, reliability is expected to exceed the minimum accepted cutoff of .70, as is true for the SL-28
when used ason overall measure of servant leadership. Second, just as support has been found for a higher order conrmatory factor
analysis showing that the seven dimensions, measured with four items for each dimension,are separate but all load on a higher order
factor capturing global servant leadership (Hu & Liden, 2011), we hypothesize that the seven items making up the SL-7 load on a
single factor that reects overall servant leadership. Third, in addition to being reliable, variance in the SL-7 should parallel the
variance captured by the SL-28, as reected in a high simple correlation between the SL-7 and SL-28. Finally, because the creators
of the frequently used servant leadership measures (Ehrhart, 2004; Liden et al., 2008; van Dierendonck & Nuijten, 2011)were
drawing largely from the work of Greenleaf (1970) and identied similar components of servant leadership, these measures are
expected to correlate with one another, demonstrating convergent validity.
Hypothesis 1. SL-7 is a psychometrically sound unidimensional representation of SL-28, such that it: a) demonstrates an internal
consistency reliability exceeding .70; b) represents a single factor reecting global servant leadership; c) positively correlates with
SL-28; and d) shows convergent validity as reected in positive correlations with Liden et al.'s SL-28 composite, Ehrhart's 14-item
composite, and van Dierendonck and Nuijten's 30-item composite.
In addition to construct and convergent validity, as captured in Hypothesis 1, criterion-related validity is essential in establishing
the overall validity of any measure (Credé et al., 2012; Society for Industrial & Organizational Psychology (SIOP), 2003). Because
servant leadership is relevant at multiple levels of analysis (van Dierendonck, 2011), we assessed criterion-related validity at
both the individual and group levels. At both levels of analysis, although a short version of a scale does not have the depth in
covering the domain of the construct that is possible with a longer version of the measure (Credé et al., 2012), we contend
that because all seven dimensions of the SL-28 are included in the SL-7, the SL-7 captures the essence of the global servant
leadership construct. Indeed, the hypotheses that follow, which are characteristic of hypotheses tested in the servant leadership
literature, are specically focused on overall or global servant leadership. Three main theoretical explanations, which comple-
ment one another, have been offered to understand why servant leadership relates to work outcomes at the individual level.
First, the support and encouragement provided to followers by servant leaders in the form of empowerment, prioritization of
fullling follower needs, and striving to bring out the full potential in followers serve to enhance follower job performance
and engagement in helping and other OCBs. This occurs due to increased self-efcacy that comes from the combination of
servant leaders providing developmental support through training and growth opportunities, and autonomy and inuence
through empowering followers. Indeed, research has demonstrated that empowerment (Seibert, Wang, & Courtright, 2011)
and self-efcacy (Stajkovic & Luthans, 1998) are positively related to job performance.
Second, accordingto social exchange theory, when leaders prioritize the needs of followers above their own self-interest and show
concern for followers' personal ambitions and potential, followers reciprocate the good deeds of their leaders by developing work
attitudes and engaging in work behaviors that benet their leaders (Blau, 1964; Gouldner, 1960). In an attempt to reciprocate for
what the leader has provided, followers not only perform their required job duties well, but also engage in discretionary behaviors
in the form of OCBs (Kamdar & Van Dyne, 2007; Walumbwa et al., 2010; Wayne, Shore, & Liden, 1997). And because servant leaders
provide a climate of psychological safety (Schaubroeck et al., 2011), risks associated when employing creative solutions to problems
are greatly reduced. As a result, we contend that servant leadership is positively related to follower creativity (Liden, Wayne, Liao, &
Meuser, 2014; Neubert et al., 2008).
Third, based on Bandura (1977) and Greenleaf (1970), in striving to encourage followers to alsoengage in helping/serving behav-
iors, servant leaders act as role models by engaging in behaviors that help others, both at work and in the larger community. Primarily
due to the servant leader's strong sense of ethics and integrity, followers admire and trust the leader. When followers form such
positive views of their leaders, they often strive to emulate the behaviors of the leader (Hunter et al., 2013) and, thus, engage in
such behaviors including OCBs (Yaffe & Kark, 2011).
Hypothesis 2. Servant leadership measured with the SL-7 and SL-28 both show positive relationships with follower in-role job
performance, creativity, helping, and OCBs.
Servant leadership has also been proposed to promote desirable group-level outcomes (Ehrhart, 2004). Because servant leaders
are ethical and strive to bring out the full potential in all followers, group members develop trust in the leader (Schaubroeck et al.,
2011). This collective trust can also be enhanced by the way that the servant leader empowers the group and provides support
designed to assist the team in meeting its goals. Coupled with servant leaders' encouragement of followers helping one another,
followers develop a sense of condence in their abilities, or potency beliefs of their teams (Hu & Liden, 2011). When followers feel
that the leader can be trusted and is there to assist the team, team members respond not only by performing well as a team, but
257R.C. Liden et al. / The Leadership Quarterly 26 (2015) 254269
also by engaging as a teamin discretionary behaviors (Hunter et al., 2013; Liden, Panaccio, Meuser, Hu, & Wayne, 2014; Liden, Wayne,
Liao, & Meuser, 2014; Schaubroeck et al., 2011). One of the explanations for the inuence of servant leadership at the group level is
that servant leaders positively affectwork climates and cultures. Specically, servantleadership appears to promote proceduraljustice
climate (Ehrhart, 2004; Walumbwa et al., 2010), service climate (Hunter et al., 2013; Walumbwa et al., 2010), as well as serving
culture (Liden, Panaccio, Meuser, Hu, & Wayne, 2014; Liden, Wayne, Liao, & Meuser, 2014). Because strong climates and cultures
are pervasive, they are shared by followers and thus inuence both team process and outcomes (Naumann & Bennett, 2000). The
more followers perceive that procedural fairness and an orientation toward helping/serving others is promoted, the more they
respond with high levels of team performance and team OCBs.
Hypothesis 3. At the team level, servant leadership measured with SL-7 and SL-28 both show positive relationships with team
potency and subsequent team performance and team OCBs.
Study 1: testing reliability, factor structure, and convergent validity
Method, Study 1
Study 1, Sample 1: working undergraduate students reliability and factor structure
In order to test the psychometric properties of the seven items identied for inclusion in the SL-7 scale, SL-28 data were collected
from 633 undergraduate students in business school courses at a major midwestern university. 598 (94%) provided usable data.
Demographics were reported as follows: 47.9% White/Caucasian, 19.5% Asian/Pacic Islander, 17.1% Hispanic/Latino/Latina,
9.3% Black/African American, 1.9% MiddleEastern, and 0.2% NativeAmerican. 53.7%were male. The average age was 22.96. The average
lifetime employment experience was 4.71 years. The average job tenure was 2.77 years. The average time knowing and working
for the leader was 2.77 and 2.40 years respectively.
Study 1, Sample 2: graduate students reliability and factor structure
In order to further increase condence in the psychometric properties of this scale, we also administered the SL-28 scale to
252 graduate students in the same college of business. 218 provided usable data (87%). Demographics were reported as follows:
43.3% White/Caucasian; 22.1% Hispanic/Latino/Latina, 16.6% Asian/Pacic Islander, 10.1% Black/African American, 1.4% Middle
Eastern, 0.5% Native American. 54.0% were male. The average age was 23.48. The average lifetime employment experience
was 5.45 years. The average job tenure was 3 years. The average time knowing and working for the leader was 3.46 and 2.20
years, respectively.
Study 1, Sample 3: university students with work experience convergent validity
In order to address convergent validity of the SL-7 scale, we collected data using three common servant leadership scales: 28 items
from Liden et al. (2008),α=.97;14itemsfromEhrhart (2004),α= .96; and 30 items from van Dierendonck and Nuijten (2011),
α= .95. As described earlier, the 28 items from Liden et al.'s (2008) measure capture 7 dimensions. Example items unique to the
28-item scale are My manager is always interested in helping people in our communityand My manager is interested in making
sure that I achieve my career goals.Example items from the Ehrhart (2004) measure include My department manager spends the
time to form quality relationships with department employeesand My department manager creates a sense of community among
department employees.The van Dierendonck and Nuijten (2011) measure consists of items capturing eight dimensions including
empowerment, standing back, accountability, forgiveness, courage, authenticity, humility, and stewardship. Example items include
My manager keeps himself/herself in the background and gives credits to othersand My manager is open about his/her limitations
and weaknesses.
Data were collected from a sample of 214 undergraduate students from 2 large midwestern universities. 131 provided usable
data (61%) after eliminating those who had not worked recently (N1yearsincetheirlastjob)andthosewhoreported
organization or leader dyad tenure of less than 2 months. We did this to ensure sufcient experience with and memory of the
leader. ANOVA revealed no signicant differences between the currently working and recently working samples on the four
study variables, so we pooled the two groups for analysis. Demographics were reported as follows: 55.7% White/Caucasian;
17.6% Hispanic/Latino/Latina, 18.3% Asian/Pacic Islander, 3.1% Black/African American, and 5.3% reported Other.44.3%
were male. Average tenure with leader was 1.5 years. The average age was 21.5; 6 elected not to report their age.
Results, Study 1
Results, Study 1, Sample 1: working undergraduate students
Supporting Hypothesis 1a, the Cronbach's alpha reliability estimates for the short (SL-7) and long (SL-28)servant leadership com-
posite scales in this sample were .80 and .95, respectively. The lower internal consistency reliability for the SL-7 was not unexpected,
as Cronbach's alpha is subject to ination due to the increased number of items present in the full 28-item scale (Cortina, 1993).
Providing support for Hypothesis 1b, a conrmatory factory analysis (CFA) conducted in Mplus suggests that a one factor SL-7 scale
shows acceptable t to the data (CFI = .97, SRMR = .03, RMSEA = .06; χ
2(14, N = 598)
= 45.36, pb.01; CFA standardized factor load-
ings are available in Table 1). Consistent with Hypothesis 1c, the Pearson correlation between the SL-7 and SL-28 scales in this sample
258 R.C. Liden et al. / The Leadership Quarterly 26 (2015) 254269
is .95. These results (Table 1) provide evidence for thereliability and single factor structure of SL-7 and its convergent validity with the
SL-28 composite measure.
Results, Study 1, Sample 2: graduate students
Providing additional support for Hypothesis 1a, Cronbach's alphas for the SL-7 and SL-28 scales in this sample were .81 and .95,
respectively. Hypothesis 1b was further supported via a conrmatory factory analysis (CFA) conducted in Mplus (Muthén &
Muthén, 2014) suggesting that a one factor SL-7 scale shows acceptable t to these data as well (CFI = .99, SRMR = .03,
RMSEA = .04; χ
2(14, N = 218)
= 20.03, p= .13; CFA standardized factor loadings are available in Table 1). The Pearson correlation
between SL-7 and SL-28 is .95, providing additional support for Hypothesis 1c. Table 1 contains the results of these analyses. Finally,
an ANOVA was conducted between the undergraduate and graduate student samples. Results do not show differences in means, t
(814) = .06, p= .95, or variances, F (1, 814) = 1.75, p= .19, of the SL-7 scale between these two samples. The graduate student
results therefore mirror the undergraduate results, showing the reliability and single factor structure of SL-7 and its convergent
validity with the SL-28 composite measure.
Results, Study 1, Sample 3: working undergraduate students
Supporting Hypotheses 1cand1d, Pearson correlations, reported in Table 2, show strong convergent validity among all four ser-
vant leadership scales (SL-7; SL-28 from Liden et al., 2008; Ehrhart, 2004; van Dierendonck & Nuijten, 2011) with correlations ranging
from .89 to .97. While the correlations with the Ehrhart and van Dierendonck and Nuijten scales are lower for SL-7 versus SL-28, they
are only slightly smaller. CFA results (Table 1) in this sample also show support for the integrity of the SL-7 scale (CFI = .96, SRMR =
.04, RMSEA = .10; χ
2(14, N = 131)
= 32.21, pb.01; CFA standardized factor loadings appear in Table 1), providing additional support for
Hypothesis 1b. As with Sample 1 and Sample 2, the Cronbach alpha of the SL-7 dropped when compared to the SL-28, but remained
strong at .89, providing additional support for Hypothesis 1a. Taken together, these(working) student samples support Hypothesis 1,
showing SL-7 to be a reliable measure that offers strong convergent validity with other, much longer, scales of servant leadership.
Study 2: convergent and criterion-related validity individual-level evidence
Although Study 1 results demonstrated internal consistency reliability, construct validity through conrmatory factor analyses,
and convergent validity with other servant leadership measures, validity is complex and needs to be evaluated in multiple ways
(Society for Industrial & Organizational Psychology (SIOP), 2003), including criterion-related validity.
Method, Study 2
Study 2, Sample 1: real estate employees
Data were collected online from employees of a large real estate company in the U.S. Followers (N = 499) and their managers
(N = 222) were invited to complete surveys. We received usable data from 174 dyads (effective response rate of 38.89%). The
majority of employees were White/Caucasian (77.0%); 10.3% were Black/African American, 5.2% were Hispanic/Latino/Latina, 3.4%
were Asian/Pacic Islander, 2.3% reported other, and three individuals elected not to report race. Most of the employees were female
(58.6%), 40.8% were male, and one individual elected not to report sex. The majority of the sample hada 4-year college degree(52.9%),
17.2% had a graduate education, 24.7% reported an associate's degree or some college, 3.4% had a high school diploma, and three
individuals elected not to report their education level. Their average organizational tenure was 5.14 years (SD = 5.63), average job
tenure 2.58 years (SD = 3.30), and average tenure with their manager was 2.00 years (SD = 2.13).
Table 2
Study 1, Sample 3: Means, standard deviations, bivariate correlations, standard error and 95% condence intervals around the correlations between the
Liden et al. (2008) SL-28, SL-7, Ehrhart (2004),andvan Dierendonck and Nuijten (2011) servant leadership scales.
Variables Mean SD 1234
1. van Dierendonck & Nuijten, 2011 4.76 1.01 (.95)
2. Ehrhart, 2004 4.74 1.32 Correlation .93 (.96)
Std. error .01
95% CI lower .90
95% CI upper .95
3. Liden et al., 2008 (SL-28) 4.75 1.24 Correlation .92 .94 (.97)
Std. error .01 .01
95% CI lower .90 .91
95% CI upper .94 .96
4. SL-7 4.67 1.35 Correlation .89 .90 .97 (.89)
Std. error .02 .02 .01
95% CI lower .85 .86 .96
95% CI upper .92 .94 .98
N = 131. All correlations are signicant at the pb.01 level (2-tailed). Cronbach alpha is reported in parentheses on thediagonal. Bootstrap results are based on 1000
samples. SD = standard deviation. CI = condence interval. SL-28 and SL-7 represent the long and short form servant leadership scales, respectively.
259R.C. Liden et al. / The Leadership Quarterly 26 (2015) 254269
Measures, Study 2, Sample 1. The response scale for OCB-O and OCB-I captured the frequency of citizenship behaviors with a scale ranging
from 1 = never to 7 = always. A strongly disagree = 1 to strongly agree = 7 response scale was used for all other measures.
OCB-O. Managers completed ratings of their followers' OCBs directed toward the organization using Podsakoff, McKenzie,
Moorman, and Fetter's (1990) 14-item measure. A sample item is attendance at work is above the norm.Cronbach's alpha for
this measure was .87.
OCB-I. Managers completed ratings of their followers' OCBs directed toward individuals in the workgroup using a 5-item measure
adapted from Rupp and Cropanzano (2002). A sample item is This follower helps you when you have a heavy work load.Cronbach's
alpha for this measure was .89.
In-role performance. Managers also completed ratings of their follower's in-role performance using Williams and Anderson's (1991)
7-item measure. A sample item is This follower adequately completes assigned duties.Cronbach's alpha for this measure was .90.
Servant leadership. Followers completed the SL-28 (Liden et al., 2008) of which the short form is a subset (SL-7; see Table 1 for the
complete list of items). Cronbach's alphas for SL-28 and SL-7 are .96 and .86 respectively.
Study 2, Sample 2: Singapore employees
Data were collected on-site from 10 organizations in Singapore. A variety of industries were represented, including education,
healthcare, consulting, and non-prot etc. In total, 409 employees (response rate = 88%) and their 78 managers (response
rate = 94%) participated in the survey during their work hours in the presence of a member of our research team. Employees
reported their managers' servant leadership. Managers rated the extent to which each subordinate engaged in creativity and
helping behaviors. Due to missing data, the nal sample consisted of 378 employeemanager dyads, leading to an effective
response rate of 81%.
Among the 378 employees, 35% were male. The majority of them had a graduate degree (45.0%), 28.0% had a bachelor's degree,
2.4% had an associate's degree or some college, 14.3% had a high school diploma, 7.7% did not have a degree, and ten individuals
did not indicate their education. The average organizational tenure was 6.80 years (SD = 6.04), average position tenure was
5.43 years (SD = 5.83), and average tenure with their manager was 3.25 years (SD =3.56).
Measures, Study 2, Sample 2. Employees reported their managers' servant leadership behaviors and managers rated their employees'
helping behaviors and creativity, both using scales ranging from 1 = strongly disagree to 7 = strongly agree.
Helping. Helping was measured by 5 items on altruism from Podsakoff et al. (1990). A sample item is This employee helps others
who have been absent.Cronbach's alpha for this measure was .89.
Creativity. We used a 4-item scale from Baer and Oldham (2006) to measure employee creativity. A sample item is This employee
often comes up with creative solutions to problems at work.Cronbach's alpha for this measure was .93.
Servant leadership. We assessed servant leadership using the same SL-28 scale (Liden et al., 2008) of which SL-7 is a subset.
Cronbach's alphas for SL-28 and SL-7 were .96 and .87 respectively.
Table 3
Study 2, Sample 1: Means, standard deviations,bivariate correlations, standard errorand 95% condence intervals around the correlationsbetween SL-28, SL-7, OCB-O,
OCB-I, and in-role performance.
Variables Mean SD 12345
1. SL-28 4.87 1.05 (.96)
2. SL-7 4.69 1.17 Correlation .96 (.86)
Std. error .01
95% CI lower .95
95% CI upper .97
3. OCB-O 5.56 .87 Correlation .39 .37 (.87)
Std. error .07 .07
95% CI lower .23 .22
95% CI upper .53 .50
4. OCB-I 5.40 1.34 Correlation .29 .28 .63 (.89)
Std. error .08 .08 .06
95% CI lower .13 .13 .49
95% CI upper .45 .43 .75
5. In-role performance 6.05 .83 Correlation .34 .28 .67 .53 (.90)
Std. error .09 .08 .05 .08
95% CI lower .16 .10 .57 .37
95% CI upper .50 .43 .76 .68
N = 174. All correlations signicant (pb.01). Bootstrapresults are basedon 1000 samples. Cronbach alpha isreported in parentheses on the diagonal. SD = standard
deviation. CI = condence interval. SL-28 and SL-7 represent the long and short form servant leadership scales, respectively. OCB-O = organizational citizenship
behavior directed toward the organization. OCB-I = organizational citizenship directed toward the leader.
260 R.C. Liden et al. / The Leadership Quarterly 26 (2015) 254269
Results: Study 2
Results Study 2, Sample 1
Table 3 contains the bivariate correlations between study variables as well as condence intervals around those estimates. The
correlation between SL-7 and SL-28is .96. Further, as is shown in Table 3, the bootstrapped 95% condence intervals (1000 samples)
around the correlations of SL-7 and SL-28 with the three study outcomevariables overlap substantially; the condence interval on the
difference between the correlations included zero. This indicates no signicant differences in the correlations of SL-7 and SL-28 and
our outcome variables. Together, these results provide additional evidence of convergent validity for the SL-7 scale.
In order to compare the criterion-related validity of SL-7 with that of SL-28, we ran two structural equation path models in Mplus
(Muthén & Muthén, 2014) using a single-indicator approach. To do this, we specied the path between the observed and latent variable
to be the square root of the Cronbach alpha reliability estimate and the error associated with the observed variable to be one minus the
Cronbach alpha reliability estimate multiplied by the observed variable's variance. We chose this analysis technique in order to simulta-
neously test the relationship between servant leadership and our three outcome variables and to compare the parameter weights and the
condence intervals around those estimates using both SL-7 and SL-28. The structural model results are presented in Fig. 1. These models
are fully saturated (i.e., just identied, df = 0), therefore we cannot evaluate (comparative) model t, as the traditional t statistics for
structural models are not meaningful for just identied models. We proceeded with this analysis, because the purpose of our statistical
test is not model comparison, for which comparison of t statistics is necessary, but rather a test of Hypothesis 2, which the path
weights and condence intervals around them provide. This method has three advantages for us over OLS regression: 1)we can eval-
uate all outcomes simultaneously; 2) measurement error is accounted for; and 3) we use the same statisticalmethod throughout this
paper on subsequent samples. The condence intervals and unstandardized parameter estimates are presentedin Table 4.ForbothSL-
7 and SL-28, all paths were signicant and positive. Specically, the unstandardized parameter estimates for SL-7 and OCB-O, OCB-I,
and in-role performance were .32, .37, and .26 and the estimates for SL-28 and these outcomes were .36, .40, and .34, respectively.
Consistent with Hypothesis 2, Study 2 provides evidence for the criterion-related validity of the SL-7 scale at the individual level.
Results, Study 2, Sample 2
The bivariate correlations between the variables as well as condence intervals around those estimates are displayed in Table 5.
Consistent with Study 2 Sample 1, the correlation between SL-7 and SL-28 is .96. Similar to the results in Study 2 Sample 1, the
bootstrapped 95% condence intervals (1000 samples) around the correlations of SL-7 and SL-28 with the two study outcome
variables overlap substantially, providing support for the convergent validity of the SL-7 scale.
Due to the nested structure of the data in this sample (employees nested within managers), we ran two multilevel struc-
tural equation models using the single indicator approach in Mplus (Muthén & Muthén, 2014) to compare the criterion-
related validity of SL-7 with that of SL-28 and account for rater effects. The multilevel structural model results are presented
in Fig. 2. Both models showed satisfactory t (SL-28: CFI = 1.00, RMSEA = .00, SRMR [within] = .00, SRMR [between] = .04,
AIC =2581.57, χ
2(2, N = 378)
= .19; SL-7: CFI = 1.00, RMSEA = .00, SRMR [within] = .00, SRMR [between] = .04, AIC = 2660.50,
2(2, N = 378)
= .12). The unstandardized parameter estimates for the relationship between SL-7 and helping and SL-7 and creativity
were .21 (p.01) and .14 (p.05), while the estimates for SL-28 and these outcomes were .23 (p.01) and .13 (p.05),
Supporting Hypothesis 2, these results provide evidence for the criterion-related validity of the SL-7 scale at
the individual level with employees from diverse organizations and in an eastern cultural context.
Study 3: convergent and criterion-related validity group-level evidence
Satised with the psychometric properties of the SL-7 subset items selected from the 28 item multidimensional servant leadership
scale and the criterion-related validity of the instrument at the individual level, we proceeded to test the convergent and criterion-
Fig. 1. Study 2, Sample 1: Structural models with unstandardized parameter estimates for both SL-7 and SL-28.
We did not present the bootstrapping results as multilevel bootstrapping techniques have yet to be developed and are not available in Mplus.
261R.C. Liden et al. / The Leadership Quarterly 26 (2015) 254269
related validity of the proposed SL-7 scale versus results obtained with the SL-28 scale at the group level. We used data from a
published group-level study that employed the original Liden et al. (2008) SL-28 scale as a composite measure of servant leadership
(Hu & Liden, 2011). Analyses described and executed with the SL-28 in the published article was re-run with the SL-7 scale described
above. Results using the SL-7 were then compared to the ndings from the same analyses reported in the published article that was
based on the SL-28. These ndings serve to establish a comparative baseline between SL-7 and SL-28, while at the same time provide
evidence for the convergent and criterion-related validity of the SL-7 at the group level of analysis.
Method, Study 3: Hu and Liden (2011)
The initial sample consisted of 95 teams operating in ve banks inChina. Theseteams worked in different functional areas, such as
accounting, personal nance, corporate nance, credit card, and trust. Two versions of surveys (one for team members and one for
upper-level managers) were distributed. Team members provided ratings on their perceptions of servant leadership (28-item
scale from Liden et al., 2008; 1 = strongly disagree to 7 = strongly agree; α= .96), team potency (7 items from Riggs & Knight,
1994; e.g., The team I work with has above average ability; 1 = strongly disagree to 7 = strongly agree; α= .78), goal clarity
(ve items from Sawyer, 1992;e.g.,please rate the degree of clarity felt about my duties and responsibilities; 1 = very uncertain to
7 = very certain; α= .87), and process clarity (ve items from Sawyer, 1992; e.g., please rate thedegree of clarity felt about how to
divide my time among the tasks required of my job; 1 = very uncertain to 7 = very certain; α= .84). Upper-level managers eval-
uated the performance (four items adaptedfrom Liden, Wayne, & Stilwell, 1993;e.g.,please rate the overall level of performance that
you observe for this team; 1 = unacceptable to 7 = outstanding; α= .96) and OCB (seven items adapted from Smith, Organ, & Near,
1983; e.g., In general, members of this team help others who have been absent; 1 = never to 7 = always; α=.89)oftheteamsunder
their jurisdiction. Participants were allowed to complete their surveys anywhere they preferred and were asked to mail the completed
surveys back to the researchers in self-addressed stamped envelopes. In order to ensure a stable membership (Hackman, 2002), only
teams with more than six months tenure were included in the analysis. The nal effective sample with matched data from upper-
level managers and team members comprised 71 teams, with 304 employees and 60 upper-level managers.
Results, Study 3: Hu and Liden (2011)
The means, standard deviations, internal consistency reliabilities, and correlations with correction for attenuation
(i.e., disattenuated correlations) are presented in Table 5. Cronbach's alphas were .96 and .90 for the SL-28 and SL-7, respectively.
The pattern of correlations remained the same for SL-28 and SL-7.
Table 4
Study 2, Sample 1: Comparison of unstandardized parameter estimates, standard errors, and 95% condence intervals for SL-7 and SL-28 criterion-related validity.
Variables Estimate Std. err. pval Lower Upper
OCB-O SL-28 .36 .07 b.001 .22 .49
SL-7 .32 .06 b.001 .20 .45
OCB-I SL-28 .40 .11 b.001 .16 .61
SL-7 .37 .11 b.001 .16 .56
In-role performance SL-28 .34 .08 b.001 .15 .48
SL-7 .26 .08 =.001 .10 .40
Note. N = 174. 95% condence intervals computed using 1000 bootstrap samples.
Table 5
Study 2, Sample2: Means, standard deviations, bivariate correlations, standard error and95% condence intervalsaround the correlationsbetween SL-28, SL-7, helping,
and creativity.
Variables Mean SD 12345
1. SL-28 4.94 .93 (.96)
2. SL-7 4.90 1.03 Correlation .96⁎⁎ (.87)
Std. error .01
95% CI lower .95
95% CI upper .97
3. Helping 5.39 .87 Correlation .19⁎⁎ .18⁎⁎ (.89)
Std. error .07 .07
95% CI lower .07 .06
95% CI upper .32 .31
4. Creativity 4.88 1.08 Correlation .11.11.70⁎⁎ (.93)
Std. error .06 .06 .04
95% CI lower .003 .001 .62
95% CI upper .22 .22 .76
Note. N = 378. Bootstrap results are based on 1000 samples. Cronbach alpha is reported in parentheses on the diagonal. SD = standard deviation. CI = condence
interval. SL-28 and SL-7 represent the long and short form servant leadership scales, respectively.
⁎⁎ pb.01.
262 R.C. Liden et al. / The Leadership Quarterly 26 (2015) 254269
As shown in Table 6, consistent with the original results using SL-28 reported by Hu and Liden (2011),conrmatory factor analysis
supports the hypothesized four-factor model wherein SL-7, team potency, goal clarity, and process clarity were treated as four inde-
pendent factors provided better t to the data (CFI = .95, SRMR = .08, RMSEA = .09; χ
= 382.81, pb.001) than the alternative
three-factor model wherein SL-7 and team potency were two factors and goal and process clarity were combined as a third factor
=43.43,pb.001) and the alternative one-factor model with all variables combined (Δχ
= 312.56, pb.001). In addition,
we followed Fornell and Larcker (1981, p. 46) to calculate the average variance extracted value, ρ
of SL-7 to determine if its con-
vergent and discriminant validity are adequate. According to Fornell and Larcker (1981),ifρ
is greater than .50, it indicates that
the variance explained by the construct is greater than the variance due to measurement error and the convergent validity is ade-
quate. At the same time, if ρ
is greaterthan the covariancebetween the focal constructand any other relatedconstructs, it suggests
adequate discriminant validity of the focal construct. The ρ
value of .63 for servant leadership measured by the SL-7 indicates that
the variance caused by measurement error is smaller than the variance explained by the servant leadership construct. These ndings
offer support for the convergent validity of the SL-7. Furthermore, the ρ
value is greater than the shared variance between servant
leadership and other studied constructs(i.e., .44 with goal clarity, .52 with processclarity, and .44 with team potency), meetingFornell
Note. * p≤ .05; ** p≤ .01.
Fig. 2. Study 2, Sample 2: Multilevel structural models with unstandardized parameter estimates for both SL-7 and SL-28.
Table 6
Study 2, Sample 2: Comparison of unstandardized parameter estimates and standard errors.
Variables Estimate Std. err. pval
Helping SL-28 .23 .07 .002
SL-7 .21 .07 .003
Creativity SL-28 .13 .07 .052
SL-7 .14 .06 .032
Note. N = 378.
Table 7
Study 3: Team construct means, standard deviations, and disattenuated correlations.
Variables Mean S.D. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
1. Organization 1 .25 .44
2. Organization 2 .24 .43 .31
3. Organization 3 .21 .41 .30 .28
4. Organization 4 .20 .40 .29 .27 .26
5. Organization 5 .10 .30 .18 .16 .16 .15
6. Team mean age 28.10 2.91 .33 .33 .23 .32 .12
7. Organizational tenure 4.78 2.96 .40 .30 .37 .35 .17 .92
8. Mean team tenure 2.99 1.67 .28 .36 .21 .31 .04 .73 .76
9. Goal clarity 5.74 .80 .11 .01 .19 .03 .24 .11 .24 .32 (.87)
10. Process clarity 5.68 .67 .15 .16 .18 .08 .33 .15 .05 .05 .84 (.84)
11. Team potency 5.23 .61 .55 .01 .07 .32 .36 .11 .11 .16 .39 .37 (.78)
12. SL-28 4.93 .51 .44 .01 .02 .18 .23 .29 .27 .08 .22 .13 .68 (.96)
13. SL-7 4.75 .69 .33 .04 .01 .10 .30 .31 .28 .08 .29 .16 .63 .84 (.90)
14. Team performance 5.31 .76 .58 .22 .11 .33 .23 .15 .10 .06 .46 .36 .68 .63 .53 (.96)
15. Team OCB 4.96 .68 .57 .04 .08 .27 .27 .20 .20 .03 .38 .38 .58 .63 .60 .78 (.89)
16. task interdependence 4.94 .50 .52 .01 .18 .23 .24 .05 .09 .01 .45 .45 .64 .53 .49 .67 .93 (.79)
Note. N = 71; Disattenuated correlations were reported. Cronbach alpha is reported in parentheses on the diagonal. SL-28 is the 28-item Liden et al. (2008) servant
leadership scale. SL-7 is the 7-item short form proposed in the present research. OCB is organizational citizenship behavior. The uncorrected correlation between
SL-7 and SL-28 is .78.
263R.C. Liden et al. / The Leadership Quarterly 26 (2015) 254269
and Larcker's (1981) discriminant validity criterion. In order to test criterion-related validity of the SL-7 versus the SL-28, hiera rchical
linear modeling (HLM; Raudenbush & Bryk, 2002) was used, as the upper level leaders who rated team performance and OCB
have oversight for multiple teams. Bauer, Preacher, and Gil's (2006) lower-level mediation method was used to test the medi-
ational role of team potency in the model. Details of the analysis method can be found in th e original article. Results of theanalysis
are presented in Tables 7 and 8.SupportingHypothesis 3, the pattern of signicance mirrors the original ndings and supports the Hu
and Liden (2011) study hypotheses in the same manner as presented in the original article. Specically, servant leadership measured
with the SL-7 scale was positively related to team potency (β=.26,pb.05), which in turn positively related to team performance
(β=.45,pb.01) and team OCB (β=.39,pb.01). At thesame time, servant leadership remained signicantly and positively related
to team performance (β= .59, pb.01) and team OCB (β=.43,pb.01), suggesting that team potency partially mediated the rela-
tionships between servant leadership and team effectiveness (Table 9). These results are consistent with the results based on
servant leadership measured with the SL-28 scale. In addition, the indirect effect was signicant for both team performance
Table 8
Study 3: Conrmatory factor analysis results for hypothesized variables.
Model χ
df Δχ
Model 1. Four-factor
382.81⁎⁎⁎ 246 .95 .08 .09
Model 2. Three-factor
426.24⁎⁎⁎ 249 43.43⁎⁎⁎ 3 .94 .08 .10
Model 3. One-factor
695.37⁎⁎⁎ 252 312.56⁎⁎⁎ 6 .89 .11 .16
CFI = comparative t index; SRMR = standardized root-mean-square residual; RMSEA = root-mean-square error of approximation.
Goal clarity, process clarity, team potency, and servant leadership as four independent factors.
Goal clarity and process clarity as one combined factor, team potency and servant leadership as two independent factors.
Goal clarity, process clarity, team potency, and servant leadership as one combined factor.
⁎⁎⁎ pb.001.
Table 9
Study 3: Hierarchical multilevel xed model analysis results of team potency as a mediator.
Total effects Fixed effects
Model 1a Model 1b Model 2 Model 3a Model 3b
Dependent variables Y: team performance Y: team OCB Y: team performance Y: team OCB
Step 1: control variables
Organization 1 .82⁎⁎⁎|.71⁎⁎⁎ .78⁎⁎⁎|.68⁎⁎⁎ .67⁎⁎⁎|.60⁎⁎ .52⁎⁎|.51⁎⁎ .73⁎⁎|.71⁎⁎
Organization 2 .12|.17 .37⁎⁎|.47⁎⁎ .13|.10 .02|.03 .12|.11
Organization 3 .41|.45 .36|.49⁎⁎ .22|.24 .32|.37 .35|.37
Organization 4 .06|.05 .16|.10 .09| .07 .08| .07 .17|.18
Team mean age .08|.11 .01|.06 .05|.05 .01|.02 .01|.02
Organizational tenure .05|.05 .01| .04 .02| .02 .04| .05 .01| .01
Team mean tenure .14|.16.03| .04 .13|.10 .08|.08 .03| .04
Step 2: independent variables
Goal clarity .29⁎⁎ (c)|.24.41⁎⁎ (c)|.32⁎⁎ .31⁎⁎ (a)|.28.14 (c)|.14 .29 (c)|.26
Process clarity .14(c)|.19.19⁎⁎ (c)|.19.25(a)|.24.01 (c)|.07 .09 (c)|.11
Servant leadership .77⁎⁎⁎ (c)|.63⁎⁎ .59⁎⁎ (c)|.54⁎⁎ .29⁎⁎ (a)|.26.63⁎⁎⁎ (c)|.59⁎⁎ .48⁎⁎ (c)|.43⁎⁎
Step 3: Mediator
Team potency .49⁎⁎ (b)|.45⁎⁎ .38⁎⁎ (b)|.39⁎⁎
X: goal clarity Estimated var (a
) = 0.03 Estimated var (a
) = 0.02
) = 0.16 E(a
) = 0.12
CI = 0.07, 0.24; S.E. = 0.04 CI = 0.03, 0.38; S.E. = 0.07
) = 0.30 E(a
) = 0.41
CI = 0.13, 0.46; S.E. = 0.09 CI = 0.14, 0.69; S.E. = 0.14
X: process clarity Estimated var (a
) = 0.02 Estimated var (a
) = 0.01
) = 0.10 E(a
) = 0.10
CI = 0.01, 0.21; S.E. = 0.06 CI = 0.01, 0.21; S.E. = 0.05
) = 0.19 E(a
) = 0.19
CI = 0.05, 0.43; S.E. = 0.12 CI = 0.04, 0.41; S.E. = 0.05
X: servant leadership Estimated var (a
) = 0.01 Estimated var (a
) = 0.01
) = 0.15 E(a
) = 0.12
CI = 0.01, 0.29; S.E. = 0.07 CI = 0.01, 0.25; S.E. = 0.07
) = 0.78 E(a
) = 0.60
CI = 0.48, 0.77; S.E. = 0.15 CI = 0.32, 0.87; S.E. = 0.14
Note.N = 71. X = goal clarity/process clarity; M = team potency; Y = team performance/team OCB; CI = condenceinterval; S.E. = standarderror. Numberon the
left of each cell is based on results with SL-28; numbers on the right of each cell is based on results with SL-7.
⁎⁎ pb.01.
⁎⁎⁎ pb.001.
264 R.C. Liden et al. / The Leadership Quarterly 26 (2015) 254269
as the outcome (95% CI = [.01, .29], excluding zero) and for team OCB as the outcome (95% CI = [.01, .25], excluding zero).
Finally, the nature of the interactions remains unchanged from Hu and Liden (Table 10). That is, in Fig. 3, we found that goal clar-
ity and team potency are positively related when servant leadership is higher (β= .43, pb.01) and became negative when ser-
vant leadership is lower (β=.22, pb.05). In Fig. 4, process clarity was more positively related to team potency when servant
leadership is higher (β=.29,pb.01) than when it is lower (β=.12, ns). This evidence supports the criterion-related validity
of the SL-7 scale and demonstrates it to be comparable to SL-28 at the group level of analysis.
The currentinvestigation provides strong support for the use of the SL-7 scale as an alternative to the SL-28 scale when researchers
are interested in investigating servant leadership as a composite or global variable. Three independent student samples demonstrated
SL-7's reliability, factor structure, and convergent validity to be commensurate with the SL-28 composite measure. Convergent
validity was further demonstrated with correlations ranging from .89 to .97 between the SL-7 and Ehrhart's (2004) 14-item and
van Dierendonck and Nuijten's (2011) 30-item servant leadership scales. In Study 2, two samples from 11 organizations were
Table 10
Study 3: Hierarchical multilevel analysis results of the moderating role of servant leadership in the relationship between goal and process clarity and team potency.
Variable Team potency
Model 1 Model 2 Model 3
Step 1: control variables
Organization 1, γ
.88⁎⁎⁎ .57⁎⁎|.60⁎⁎ .56⁎⁎|.59⁎⁎
Organization 2, γ
.39 .19|.20 .16|.16
Organization 3, γ
.52.33|.25 .35|.24
Organization 4, γ
.10 .16|.07 .15|.03
Team mean age, γ
.03 0|.05 0|.05
Organizational tenure, γ
.06 .05|.02 .06|.05
Team mean tenure, γ
.16⁎⁎ .11|.10 .14|.08
Step 2: independent variables
Goal clarity, γ
Process clarity, γ
.16|.14 .09|.08
Servant leadership, γ
.40⁎⁎|.36⁎⁎ .46⁎⁎|.38⁎⁎
Step 3: Moderator
Goal clarity servant leadership, γ
Process clarity servant leadership, γ
.15 .17|.17 .21|.22
.18 .58|.47 .74|.65
.13 .29|.26 .36|.34
.02|.02 .04|.05
.40|.29 .16|.18
.16|.13 .07|.08
Note. N = 71. Number on the left of each cell is based on results with SL-28; numbers on the right of each cell is based on results with SL-7.
(1 ICC1) + R
ICC1. ICC1 indicates the proportion of variance in the outcome variable that resides between groups. ICC1
for team potency as the outcome is .29.
⁎⁎ pb.01.
⁎⁎⁎ pb.001.
Low High
Team Potency
Goal Clarity
High servant
Low servant
Fig. 3. Study 3:Interaction between goal clarity and servant leadership on team potency. Note.For the interaction with SL7,the slope for the relationship between goal
clarity and team potency in high servant leadership condition is .43, pb.01; the slope for the low servant leadership condition is .22, pb.05.
265R.C. Liden et al. / The Leadership Quarterly 26 (2015) 254269
used to establish criterion-related (concurrent) validity at the individual level of analysis. In order to establish criterion-related
validity at the group level, data from a published servant leadership study were re-analyzed with SL-7 as the measure of servant
leadership and results were compared with those published in the original article using the SL-28 measure. In the analyzed datasets,
the patterns of signicance and results were unchanged with regard to study hypotheses and nearly unchanged overall between the
SL-7 and SL-28.
Implications for servant leadership theory
In order for a theory to be useful, it must be testable (Whetten, 1989), and in order for meaningful tests to be conducted, valid
measurement is essential. Although van Dierendonck identied the measures by Liden et al. (2008) and van Dierendonck and
Nuijten (2011) as the only servant leadership measures meeting this requirement due to the rigorous procedures that were
used in developing them, these scales have 28 and 30 items, respectively. Clearly, in order to be more widely usable, a concise
scale intended to assess global servant leadership is needed to expedite theory testing on servant leadership. We feel that the
SL-7 provides such a measurement tool for enhancing servant leadership theory testing.
Greenleaf's (1970) essay on servant leadership, while seminal, was an expression of a leadership philosophy rather than a leadership
theory. Indeed, it reads more akin to a Socratic dialogue than a scientic discussion of theory and hypotheses. Greenleaf was reecting on
his career, informed by his own experience and thoughtful reection on the condition of society as he saw it in the U.S. in the 1960's.
His original essay argued that serving others should not only coexist with leadership and productivity but is integral to them. While
the foundation of servant leadership theory was laid with Graham's (1991) insightful discussion of servant leadership and presentation
of case studies, her work can be seen as an appeal to develop and test theory, rather than a development of theory per se. One challenge
that may have prevented a more rapid adoption of the servant leadership construct and development of theory by researchers was
the lack of a clear denition provided by Greenleaf himself. As van Dierendonck (2011) notes, a plethora of servant leadership models
with various dimensions lled this vacuum, adding to the difculty that researchers face when developing servant leadership theory.
However, in recent years, scholars have made signicant progress in laying the groundwork for servant leadership theory (for a review
see Liden, Panaccio, Meuser, Hu, & Wayne, 2014). With the psychometrically sound SL-28 for assessing servant leadership dimensions
(Liden et al., 2008), as well as the concise SL-7 scale presented in the current study for efciently measuring global servant leadership,
servant leadership is poised to become a prominent theory of leadership.
Strengths, limitations, and suggestions for future research
A strength of the current investigation is the assessment of the new 7-item servant leadership scale (SL-7) employing six independent
samples to assess its reliability and validity at both the individual and group levels. Another strength of the present study is the use of
multicultural data, with data collected in the U.S., China, and Singapore. Despite this strength, it would be useful to expand the data col-
lection using the SL-7 to countries in other parts of the world.
A potential weakness is that the original SL-28 scale was developed using U.S. samples exclusively, which may limit its
generalizability. Although van Dierendonck and Nuijten's (2011) scale, which we demonstrated in the current investigation
to be highly correlated with the SL-7, was developed in the Netherlands, to increase the likelihood that all items capture global
relevance we recommend that future scale development efforts include samples from culturally diverse countries (Liden, 2012).
Specically, when the goal is to develop theories and corresponding measures that generalize across national borders, the standard
practice of developingscales in the U.S. and subsequently translating them for use in other countries should be abandoned. We should
note that although there is great value in developing theory and measures that transcend national borders, we are not arguing that
there is not a place for studies of indigenous practices or cultural artifacts. It is possible though, with the internet and dramatic
Low High
Team Potency
Process clarity
High servant
Low servant
Fig. 4. Study 3: Interaction between process clarity and servant leadership on team potency. Note. For the interaction with SL7, the slope for the relationship between
process clarity and team potency in high servant leadership condition is .29, pb.01; the slope for the low servant leadership condition is .12, pN.05.
266 R.C. Liden et al. / The Leadership Quarterly 26 (2015) 254269
increasesin the number of people traveling internationally, that such regionaldifferences over time may deem indigenous researchto
be less relevant.
Although the use of student samples is often seen as a limitation (Henry, 2008; Peterson, 2001; Sears, 1986)duetorange
restriction, experience base,and generalizability concerns, reduced external validity may not extend to all psychological phenomenon
(e.g., Clara, Coz, & Enns, 2003; Wiecko, 2010). The work experience present in our undergraduate and graduate student samples,
as well as the psychometric similarity of SL-7 between the undergraduate and graduate student populations, provide condence in
the veridicality of the student results. A limiting factor of the samples in Study 2 is that the sample sizes did not provide adequate
power to conduct the CFAs at the item level. A limitation of the group-level sample in Study3 is the restriction to the banking industry.
Further, as Smith et al. (2000) observe, correlating a subset of items with a larger set administered at the same time may upwardly
bias the correlation. Ideally, the short form scale would be collected at a different time or elsewhere in the survey. We do not view
this as a major limitation, especially given that instead of assuming the outcome-related validity of the SL-7 based on results of the
full 28-item scale, we directly tested the criterion-related validity of the SL-7 in conformance with the recommendations of Smith
et al. (2000). We also conformed with Smith et al.'s (2000) suggestion that short scales should retain coverage of the entire 7-
factor content domain (also see Credé et al., 2012 and Johnson et al., 2011).
Given the strong support we found for the SL-7 scale, we encourage researchers to develop comprehensive models on the
antecedents and/or outcomes of servant leadership. The shorter servant leadership scale allows for these larger models, which should
result in studies with greater incremental contributions to what we know about this form of leadership. The SL-7 scale may also allow
for additional leadership behaviors to be measured (such as transformational leadership) so that the unique contributions of servant
leadership in explaining outcomes are further demonstrated.
Practical implications
Researchers studying servant leadership can make use of SL-7, the shortest of servant leadership measures to date (van
Dierendonck, 2011). This reduces servant leadership measurement by 21 items when compared to the original SL-28 measure
(Liden et al., 2008) and by 7 items when compared to Ehrhart's (2004) 14-item measure, the heretofore shortest servant leadership
scale. This has obvious positive consequences for reducing respondent fatigue and recognizing that organizational participants' time
is at a premiumwithin organizations not interested in the dimensions of servant leadership exercised by their managers. Also, the use
of SL-7 has demonstrated strong psychometric properties versus ad hoc combinations of SL items that may severely compromise
validity (Bono & McNamara, 2011; Keller & Dansereau, 2001; Smith et al., 2000). Specically, the SL-7 approximates the domain of
the construct as all seven dimensions of the SL-28 are contained, with one item from each dimension. Finally, the small decreases
in reliability and validity from the SL-28 to the SL-7 are negligible, making the time savings associated with asking respondents to
complete 21 fewer items well worth it if the goal is to assess global servant leadership (Smith et al., 2000).
Although each of the 7 dimensions of the original SL-28 are represented in the SL-7, it is important to consider that as with any
short version of a scale, reducing the number of items limits the extent to which the domain of the construct is captured (Credé
et al., 2012; Smith et al., 2000). For example, the item included in the SL-7 to capture the emotional healing dimension, I would
seek help from my manager if I had a personal problem,does not directly address the manager's concern for the follower's
well-being, even though this is part of the denition of this dimension and is captured by another item that directly mentions the
leader's concern for the follower's well-being. Despite the fact that all 4 items making up the emotional healing dimension are
positively correlated, we can only assume that a follower would not seek a manager's help with a personal problem unless
that follower felt that the manager cared about her well-being. This example points to the limitation of any short version of a
scale in capturing the full domain of each dimension. Therefore, if researchers desire full coverage of the domain of each dimension,
the full scale should be used. Although the current investigation provides substantial evidence for the success of the SL-7 in capturing
the essence of the SL-28, the SL-7 does not serve as a replacement for the SL-28 for research in which the dimensions are to be
analyzed separately or when it is essential for the domain of the construct to be fully included in the measure.
Summary and conclusion
Substantial progress has been made toward developing servant leadership theory over the past decade. Research may be further
enhanced with the psychometrically sound SL-7 scale introduced in thecurrent investigation. Two undergraduate samples, a graduate
sample, and three organizational samples from the U.S., China, and Singapore, provide strong evidence for the validity of the SL-7.
Across these samples, the correlation between the SL-7 and SL-28 averaged .90, reliabilities for the SL-7 remained above .80 in all
samples, and criterion-related validities (tested only in the organizational samples) for the SL-7 were high and very similar to those
produced by the SL-28. In sum, the SL-7 has much to offer to future researchers interested in measuring global servant leadership.
We thank Samantha Balaton, Marie Halvorsen-Ganepola, David Henderson, Kent Keith, Ariel Levi, Courtney Masterson, Megan
Melton, Lois Siew, Kim Yang, Selena Yuan, and Hao Zhao for their research assistance.
267R.C. Liden et al. / The Leadership Quarterly 26 (2015) 254269
Baer, M., & Oldham, G.R. (2006). The curvilinear relation between experienced creative time pressure and creativity:Moderating effects of opennessto experience and
support for creativity. Journal of Applied Psychology,91(4), 963970.
Bandura, A. (1977). Social learning theory. Englewood Cliffs, NJ: Prentice Hall.
Barbuto, J., & Wheeler, D. (2006). Scale development and construct clarification of servant leadership. Group & Organization Management,31,300326.
Bass, B.M. (1985). Leadership and performance beyond expectations. New York: Free Press.
Bauer, D.J., Preacher, K.J., & Gil, K.M. (2006). Conceptualizing and testing random indirect effects and moderated mediation in multilevel models: New procedures and
recommendations. Psychological Methods,11,142163.
Blau, P.M. (1964). Exchange and power in social life. New York: John Wiley & Sons.
Bono, J.E., & McNamara, G. (2011). Publishing in AMJPart 2: Research methods. Academy of Management Journal,54,657660.
Brown, M.E.,Treviño, L.K., & Harrison,D.A. (2005). Ethical leadership: A social learning perspective for construct development and testing. OrganizationalBehavior and
Clara, I.P., Coz, B.J., & Enns, M.W. (2003). Hierarchical models of personality and psychopathology: The case of self-criticism, neuroticism, and depression. Personality
and Individual Differences,35,9199.
Cortina, J.M. (1993). What is coefficient alpha? An examination of theory and applications. Journal of Applied Psychology,78,98104.
Credé, M., Harms, P., Niehorster, S., & Gaye-Valentine, A. (2012). An evaluation of the consequences of using short measures of the big five personality traits. Journal of
Personality and Social Psychology,102,874888.
Dansereau,F., Graen, G.B., & Haga, W. (1975). A vertical dyad linkage approach to leadership in formal organizations. Organizational Behavior and Human Performance,
Dinh, J.E.,Lord, R.G., Gardner,W., Meuser, J.D., Liden,R.C., & Hu, J. (2014). Leadership theoryand research in the newmillennium: Current theoretical trendsand chang-
ing perspec tives. Leadership Quarterly,25,3662.
Ehrhart, M.G. (2004). Leadership and procedural justice climate as antecedents of unit-level organizational citizenship behavior. Personnel Psychology,57,6194.
Fleishman, E.A. (1998). Consideration and structure: Another look at their role in leadership research. In F. Dansereau, & F.J. Yammarino (Eds.), Leadership: The
multiple-level approaches (pp. 5160). Stamford, CT: JAI Press.
Fornell, C., & Larcker,D.F. (1981). Evaluating structural equation models with unobservable variables and measurementerror. Journal of Marketing Research,18,3950.
Gouldner, A.W. (1960). The norm of reciprocity: A preliminary statement. American Sociological Review,25,161177.
Graham, J.W. (1991). Servant-leadership in organizations: Inspirational and moral. Leadership Quarterly,2,105119.
Greenleaf, R.K. (1970). The servant as leader. Newton Centre, MA: The Robert K. Greenleaf Center.
Hackman, J.R. (2002). Leading teams: Setting the stage for great performances. Boston, CA: Harvard Business School Press.
Henry, P.J. (2008). College sophomores in the laboratory redux: Influences of a narrow data base on social psychology's view of the nature of prejudice. Psychological
Hoogervorst, N., De Cremer, D., & van Dijke, M. (2010).Why leaders not always disapproveof unethical followerbehavior: It depends on the leader's self-interest and
accountability. Journal of Business Ethics,95,2941.
Hu, J., & Liden, R.C. (2011). Antecedents of team potency and team effectiveness: An examination of goal and process clarity and servant leadership. Journal of Applied
Hunter, E.M., Mitchell, J., Neubert, M.J., Perry, S.J., Witt, L.A., Penney, L.M., et al. (2013). Servant leaders inspire servant followers: Antecedents and outcomes for em-
ployees and the organization. Leadership Quarterly,24,316331.
Johnson, R.E., Rosen, C.C., & Chang, C. (2011). To aggregate or not to aggregate: Steps for developing and validating higher-order multidimensional constructs. Journal
of Business and Psychology,26,241248.
Kamdar, & Van Dyne, L. (2007). The joint effects of personality and workplace social exchange relationships in predicting task performance and citizenship perfor-
mance. Journal of Applied Psychology,78,662674.
Keller, R., & Dansereau, F. (2001). The effect of adding items to scales: An illustrative case of LMX. Organizational Research Methods,4,131
Law, K.S., Wong, C.S., & Mobley, W.H. (1998). Toward a taxonomy of multidimensional constructs. Academy of Management Review,23,741755.
Liden, R.C. (2012). Leadership research in Asia: A brief assessment and suggestions for the future. Asia Pacific Journal of Management,29,205212.
Liden, R.C., Panaccio, A., Meuser, J.D., Hu, J., & Wayne, S.J. (2014). Servant leadership: Antecedents, processes, and outcomes. In D.V. Day (Ed.), The Oxford handbook of
leadership and organizations (pp. 357379). Oxford, England: Oxford University Press.
Liden, R.C., Wayne, S.J., Liao, C., & Meuser, J.D. (2014). Servant leadership and serving culture: Influence on individual and unit performance. Academy of Management
Liden, R.C., Wayne, S.J., & Stilwell, D. (1993). A longitudinal study on the early development of leadermember exchanges. Journal of Applied Psychology,78,662674.
Liden, R.C., Wayne, S.J., Zhao, H., & Henderson, D. (2008). Servant leadership: Development of a multidimensional measure and multilevel assessment. Leadership
Muthén, L.K., & Muthén, B.O. (2014). Mplus user's guide (Seventh Edition). Los Angeles, CA: Muthén & Muthén.
Naumann, S.E., & Bennett, N. (2000). A case for procedural justice climate: Development and test of a multilevel model. Academy of Management Journal,43,881889.
Neubert, M.J., Kacmar, K.M., Carlson, D.S., Chonko, L.B., & Roberts, J.A. (2008). Regulatory focus as a mediator of the influence of initiating structure and servant lead-
ership on employee behavior. Journal of Applied Psychology,93,12201233.
O'Reilly, C.A., III, Doerr, B., Caldwell, D.F., & Chatman, J.A. (2014). Narcissistic CEOs and executive compensation. Leadership Quarterly,25,218231.
Page, D., & Wong, P.T.P. (2000). A conceptual framework for measuring servant leadership.In S. Adjibolooso (Ed.), The human factorin shaping the course of history and
development (pp. 69110). Washington, DC: American University Press.
Peterson,R.A. (2001). On the use of college students in social science research: Insightsfrom a second-order meta-analysis. Journalof Consumer Research,28,450461.
Peterson, S.J., Galvin, B.M., & Lange, D. (2012). CEO servant leadership: Exploring executive characteristics and firm performance. Personnel Psychology,65,565596.
Podsakoff, P.M., McKenzie, S.B., Moorman, R.H., & Fetter, R. (1990). Transformational leader behaviors and their effects on followers' trust in leader, satisfaction, and
organizational citizenship behaviors. Leadership Quarterly,1,107142.
Raudenbush, S.W., & Bryk, A.S. (2002). Hierarchical linear models: Applications and data analysis methods (2nd ed.). Newbury Park, CA: Sage.
Riggs, M.L.,& Knight, P.A. (1994). The impact of perceived groupsuccessfailureon motivationalbeliefs and attitudes: A causal model. Journal of AppliedPsychology,79,
Rupp, D.E., & Cropanzano, R. (2002). The mediating effects of social exchange relationships in predicting workplace outcomes from multifoci organizational justice.
Organizational Behavior and Human Decision Processes,89,925946.
Sawyer, J.E. (1992). Goal and process clarity: Specification of multiple constructs of role ambiguity and a structural equation model of their antecedents and
consequences. Journal of Applied Psychology,77,130143.
Schaubroeck, J., Lam, S.S.K., & Peng, A.C. (2011). Cognition-based and affect-based trust as mediators of leader behavior influences on team performance. Journal of
Applied Psychology,96,863871.
Sears, D. (1986). College sophomores in the laboratory: Influences of a narrow data base on social psychology's view of human nature. Journal of Personality and Social
Seibert, S. E., Wang, G., & Courtright, S. H. (2011). Antecedents and consequences of psychological and team empowerment in organizations: A meta-analytic review.
Journal of Applied Psychology,96,9811003.
Smith, G.T., McCarthy, D.M., & Anderson, K.R. (2000). On the sins of short-form development. Psychological Assessment,12,102111.
Smith, A.C., Organ, D.W., & Near, J.P. (1983). Organizational citizenship behavior: Its nature and antecedents. Journal of Applied Psychology,68,653663.
Society for Industrial and Organizational Psychology (SIOP) (2003). Principles for the validation and use of personnel selection procedures. Bowling Green, OH: SIOP.
268 R.C. Liden et al. / The Leadership Quarterly 26 (2015) 254269
Spears, L.C., & Lawrence, M. (2002). Focus on leadership: Servant-leadership for the 21st century. New York: John Wiley & Sons.
Stajkovic, A.D., & Luthans, F. (1998). Self-efficacy and work-related performance: A meta-analysis. Psychological Bulletin,124,240261.
van Dierendonck, D. (2011). Servant leadership: A review and synthesis. Journal of Management,37,12281261.
van Dierendonck, D., & Nuijten, I. (2011). The servant-leadership survey: Development and validation of a multidimen sional measure. Journal of Business and
van Dierendonck, D., Stam, D., Boersma, P., de Windt, N., & Alkema, J. (2014). Same difference? Exploring the differential mechanisms linking servant leadership and
transformational leadership to follower outcomes. Leadership Quarterly,25,544562.
Walumbwa, F.O., Hartnell, C.A., & Oke, A. (2010). Servant leadership, procedural justice climate, service climate, employee attitudes, and organizational citizenship be-
havior: A cross level investigation. Journal of Applied Psychology,95,517529.
Wayne, S.J., Shore, L.M., & Liden, R.C. (1997). Perceived organizational support and leadermember exchange: A social exchange perspective. Academy of Management
Whetten, D.A. (1989). What constitutes a theoretical contribution? Academy of Management Review,14,490495.
Wiecko, F.M. (2010). Research note: Assessing the validity of college samples: Are students really that different? Journal of Criminal Jus tice,38,11861190.
Williams, L.J., & Anderson, S.E. (1991). Job satisfaction and organizational commitment as predictors of organizational citizenship and in-role behaviors. Journal of
Yaffe, T., & Kark, R. (2011). Leading by example: The case of leader OCB. Journal of Applied Psychology,96,806826.
269R.C. Liden et al. / The Leadership Quarterly 26 (2015) 254269
... Thereafter, each principal was provided with a questionnaire in which they were prompted to select one word from five, and match the selected word with its nearest meaning, also as from an option of five. Each principal was then asked to do a self-assessment by completing Liden's quantitative Servant Leadership Test (Liden et al., 2015), in order for their own views to be compared to those of their subordinates. The chosen samples of teachers from both School A and School B also completed Liden's 28-scale Servant Leadership Test questionnaire in order to assess their views of their principal as a servant leader (Liden et al., 2015). ...
... Each principal was then asked to do a self-assessment by completing Liden's quantitative Servant Leadership Test (Liden et al., 2015), in order for their own views to be compared to those of their subordinates. The chosen samples of teachers from both School A and School B also completed Liden's 28-scale Servant Leadership Test questionnaire in order to assess their views of their principal as a servant leader (Liden et al., 2015). Each group of teachers also completed Gallup's Q2 Employee Engagement Survey (, to measure the engagement level of each group of teachers from School A and School B. ...
... Each group of learners also completed Liden's quantitative 28-scale Servant Leadership Test in order to assess their views of their principal as a servant leader (Liden et al., 2015). Lastly, each group of learners completed an adapted version of Gallup's Q12 Employee Engagement Survey ( ...
Full-text available
This study aimed to investigate the impact of “purpose” as a principal leadership skill on the performance of two township schools using a quantitative research design and collecting data from the school principals, teachers and matric learners, using the 28-scale Servant Leadership Test as well as Gallup’s Q12 Employee Engagement survey. The questionnaires addressed the key objectives, namely, the extent to which the principals of the participating schools exhibited servant leadership and their understanding of “purpose” as one word in leadership and how teachers and learners perceived the impact of a “one-word” purpose-driven leader on the performance of the selected schools. Although no relationship could be demonstrated between ‘’purpose’’ and the performance of the two township schools, it became evident that a significant increase in Servant Leadership leads to a significant increase in engagement and performance, as measured by the matric pass rate. It is recommended that workshops be facilitated with principals and teachers in order to entrench ‘’purpose’’ deeper throughout the schools. In addition, Servant Leadership training has to conduced to increase the leadership ability of the school principals. Future research in the area of ‘’purpose as one word’’, as well as Servant Leadership as a principal skillset within the South Africa’s public school’s leadership is recommended.
... influences organizational outcomes, such as in-role performance (Liden et al. 2015) and affective commitment . Moreover, organizational commitment itself was identified as an essential element for facilitating positive individual and organizational outcomes (DeCotiis and Summers 1987). ...
... Moreover, the studies by Saleem et al. (2020) and Liden et al. (2014) revealed that servant leadership predicts subordinates' performance. Notably, servant leadership can have an influence on several performance levels and types, such as in-role performance (Liden et al. 2015), team performance (Hu and Liden 2011), and sales performance (Coetzer et al. 2017). Also the meta-analysis by Lee et al. (2020) showed strong effects of servant leadership on individuals' performance. ...
... We measured all our items with symmetric and equidistant Likert scales (Ghasemy et al. 2020) anchored by 1 (strongly disagree) to 5 (strongly agree). With respect to servant leadership, we employed the seven-item servant leadership scale (SL-7), which was developed and validated by Liden et al. (2015). This scale is one of the recommended scales by Eva et al. (2019), due to its inclusion of creating value for the community and conceptual skills, as well as its easiness of application and parsimony, and has been used in other studies before (e.g., Sun et al. (2019)). ...
Full-text available
In our study, we investigated the change patterns in academics’ servant leadership behavior and affective commitment during and after the first Covid-19 lockdown (April 2020–August 2020) in Malaysia. Furthermore, we evaluated the influence of academics’ servant leadership behavior on job performance through affective commitment. We applied additional analyses to determine the degree to which the two former constructs are influenced by age and academic rank. To do so, we leveraged multivariate latent growth curve (LGC) modeling in analyzing the longitudinal data collected from 220 academics at three time points over a course of four months during the Covid-19 pandemic. Our results showed that, while servant leadership and affective commitment were stable over time, servant leadership was a driver of affective commitment and job performance. We also observed that affective commitment significantly mediated the relationship between academics’ servant leadership behavior and job performance. Moreover, we found interindividual differences in servant leadership and affective commitment in terms of age and academic rank at the beginning of the Covid-19 lockdown. Our multi-disciplinary research is among the limited number of studies applying a rigorous, longitudinal approach to examine behaviors (i.e., servant leadership and job performance) and attitudes (i.e., affective commitment) during the Covid-19 pandemic in higher education research.
... Van Dierendonck and Nuijten (2011) later developed a multidimensional measurement instrument to capture these essential elements of servant-leadership. Presently, several multidimensional measurement instruments exist, affording future researchers ample opportunities with which to test the theory's complexity and underlying foundation (Liden et al., 2015;Liden, Wayne, Zhao, & Henderson, 2008;Van Dierendonck & Nuijten, 2011). ...
In this paper, the authors describe a conceptual model of servant leadership and organizational performance, taking into account the moderating influences of physical distance (i.e. leader geographical dispersion) and unified communication solutions. The authors position physical distance as a barrier to effective servant leadership implementation and introduce unified communication solutions as a means by which to mitigate those negative effects and simultaneously foster the development of community while serving the legitimate needs of subordinates.
... Servant Leadership is mostly concerned with employee participation, solidarity, valuing unity, and human resource development. The servant leadership framework is an amalgamation of seven characteristics (Liden et al., 2015) including forming relationships, behaving ethically, signifying conceptual skills, putting first, empowering, helping followers grow and succeed, and creating value for others external to the organization. Transactional Leadership is largely associated with conversing with the team members to develop detailed and plain goals and establishing that workers acquire what is guaranteed for the purpose accomplished as per the accord, trade rewards for employees' effort, and are alert to their endeavor instantly (Bryant, 2003). ...
... Therefore, the leadership construct is appropriately defined as residing on a continuum. For instance, leadership approaches can range from authoritative to lenient and supportive to directive (Liden et al., 2015). These perspectives for defining the concept of leadership are essential to understanding the various situations in which the organizations found themselves; they are regarded as important aspects through which the practice of dynamic and flexible leadership may be introduced. ...
Full-text available
This study utilized a survey to explore the relationships between servant leadership characteristics of school teachers, student success, and organizational health in selected public schools. This study hypothesized that the servant leadership of school teachers would positively affect student success and the organizational health of public schools. The Organizational Leadership Assessment was used (Laub, 1999) to collect data from fifteen public schools in Houston, Texas. Data were screened for normality, linearity, and homoscedasticity to ensure that the regression results were robust, and univariate and multivariate analyses were used to determine the nature of the relationship between organizational health and servant leadership. Descriptive analysis explained variance in dependent and independent variables. The values people subscale was statistically significant.
... Standardized scales were used to measure all the study variables. To measure SL, a sevenitem global SL measure was adopted from Liden et al. (2015). For measuring JS of the team members, the study employed a scale used by Jamal and Baba (1997), which was initially developed by Parker and DeCotiis (1983) and consisted of nine items. ...
Purpose This study investigated the impact of servant leadership on project success in nongovernment organizations (NGOs) working in a developing country like Pakistan. A moderated mediation design was employed, and the mediating role of employees' emotional intelligence (EI) and job stress (JS) was tested between servant leadership on project success. Moreover, the study also examined the moderating role of team effectiveness. Design/methodology/approach Data were collected from 441 project team members working on different developed projects. Data were analyzed using partial least square-structural equation modeling (PLS-SEM) technique. Findings Results revealed that servant leadership exerts a significant positive impact on project success. Also, it is noted that servant leadership significantly increases the employee's emotional intelligence that contributes to project success although it does not reduce JS. However, JS was found to be a significant mediator in the association between servant leadership on project success. The findings also revealed that team effect plays an imperative moderating role in ensuring project success. Originality/value The study is one of the very few studies conducted to assess the impact of servant leadership on project success in not-for-profit organizations. The study contributes to the literature and methodology by adopting a holistic approach to investigate the mediation of EI and JS along with the moderation of team effectiveness in the nexus of servant leadership and project success.
... Servant leadership explicitly incorporates stewardship as an essential element of effective leadership (Eva et al., 2019); therefore, servant leadership in social work innovation practice belongs to a certain extent to the high command type. On the other hand, servant leaders empower their employees, prioritizing their interests, provide them greater work autonomy and develop their competencies, the followers reciprocate the favorable treatments with desirable behaviors including engaging in innovative behavior (Liden et al., 2015;Schaubroeck et al., 2011;Wang et al., 2019). Servant leadership develops an exchange relationship between leaders and followers and several studies have confirmed the underlying mechanisms such as leader-member exchange (Bao et al., 2018) and affective commitment (Jang & Kandampully, 2018). ...
Purpose: Using situational leadership theory, this study examined the innovation of social workers in community-level social work institutions under different leadership styles. Methods: We conducted an online survey experiment on the professional strengths of social workers in eight cities of mainland China from December 8–20, 2020. Results: Both ethical leadership and servant leadership are conducive to social workers’ innovative behaviors. Our full model also suggests that organizational climate moderates the effects of leadership styles on innovative behaviors, with servant leadership having a greater impact than ethical leadership on social workers’ innovative behavior in low-innovation climate environments, but ethical leadership has a greater impact than servant leadership on social workers’ innovative behavior in high-innovation climate environments. Conclusion: Our study provides important strategies to fully stimulate the innovation effects of leadership based on the stage characteristics of social work innovation practices.
The main focus of the chapter is to introduce and describe the central characteristics of common good leadership and establish existing connections to servant leadership. The multiple notions of the common good have originated in different philosophical traditions, and common good leadership overlaps with a variety of positive and propositive leadership styles, particularly servant leadership; therefore, the most relevant traits and skills of both leadership styles will be analyzed, providing insights into whether leading for the common good can be considered an act of service and if acts of service are enough to drive collective actions towards the common good. Literature review and content analysis were used as the methodologies of the study.
Purpose Given the intensified need to be responsive to change, employees' discretionary and constructive efforts, such as those aimed at effecting workplace functional change (i.e. taking charge), are required to enhance organizational effectiveness. Therefore, the authors reckon that due to their serving attitude of prioritizing the needs of others above the self and their motivational qualities, the servant leadership approach can enhance the confidence of subordinates' capabilities to perform a range of meaningful activities (i.e. role breadth self-efficacy; RBSE), which in turn should facilitate their engagement in taking charge. Design/methodology/approach The authors collected data from 324 leader-subordinate dyads (i.e. academicians) from two federal universities in Nigeria. The authors assessed the measurement and structural models with partial least squares structural equation modeling (PLS-SEM). Findings This study found that servant leadership and RBSE were crucial enablers of subordinates' taking charge. Furthermore, a positive relationship between servant leadership and RBSE was found. Lastly, RBSE was a partial mediating mechanism partly underlying the positive relationship between servant leadership and taking charge. Practical implications Selecting and training leaders to practice servant leadership in Nigerian public universities may serve as a springboard for employees to take charge because it also enables them to develop their RBSE. Originality/value The current study sheds light on the psychological process through which servant leadership affects subordinates' taking charge by identifying RBSE as a crucial motivational state partly underlying the process.
Full-text available
Current conceptualizations of procedural justice focus largely on the individual level of analysis; no framework exists for examining procedural justice's social context. Empirical tests reported here offer some support for group-level and cross-level hypotheses. Work group perceptions of cohesion and supervisor visibility in demonstrating procedural justice were associated with the development of procedural justice climate. Procedural justice climate was positively associated with helping behaviors after the effects of individual procedural justice perceptions were controlled for.
The statistical tests used in the analysis of structural equation models with unobservable variables and measurement error are examined. A drawback of the commonly applied chi square test, in addition to the known problems related to sample size and power, is that it may indicate an increasing correspondence between the hypothesized model and the observed data as both the measurement properties and the relationship between constructs decline. Further, and contrary to common assertion, the risk of making a Type II error can be substantial even when the sample size is large. Moreover, the present testing methods are unable to assess a model's explanatory power. To overcome these problems, the authors develop and apply a testing system based on measures of shared variance within the structural model, measurement model, and overall model.
The article focuses on the development of a theory. A discussion is presented about steps involved in developing a theory, such as seeing which factors logically should be considered as part of the explanation of the social or individual phenomena of interest. The authors assert that authors developing theories are considering these factors, they should err in favor of including too many factors, recognizing that over time their ideas will be refined. The article presents information about the importance of sensitivity to the competing virtues of parsimony and comprehensiveness.
Psychological research involving scale construction has been hindered considerably by a widespread lack of understanding of coefficient alpha and reliability theory in general. A discussion of the assumptions and meaning of coefficient alpha is presented. This discussion is followed by a demonstration of the effects of test length and dimensionality on alpha by calculating the statistic for hypothetical tests with varying numbers of items, numbers of orthogonal dimensions, and average item intercorrelations. Recommendations for the proper use of coefficient alpha are offered.
The manner in which the concept of reciprocity is implicated in functional theory is explored, enabling a reanalysis of the concepts of "survival" and "exploitation." The need to distinguish between the concepts of complementarity and reciprocity is stressed. Distinctions are also drawn between (1) reciprocity as a pattern of mutually contingent exchange of gratifications, (2) the existential or folk belief in reciprocity, and (3) the generalized moral norm of reciprocity. Reciprocity as a moral norm is analyzed; it is hypothesized that it is one of the universal "principal components" of moral codes. As Westermarck states, "To requite a benefit, or to be grateful to him who bestows it, is probably everywhere, at least under certain circumstances, regarded as a duty. This is a subject which in the present connection calls for special consideration." Ways in which the norm of reciprocity is implicated in the maintenance of stable social systems are examined.
We propose a taxonomy of multidimensional constructs based on the relations between the construct and its dimensions. Multidimensional constructs that exist at deeper levels than their dimensions we term latent model. We call constructs formed as algebraic functions of their dimensions aggregate model, whereas constructs formed as different profiles of dimensional characteristics we term profile model. We discuss the nature of multidimensional constructs defined under these models and their operationalizations in empirical research.