Humility

r Academy of Management Journal
2016, Vol. 59, No. 3, 1088–1111.

http://dx.doi.org/10.5465/amj.2013.0660
HOW DOES LEADER HUMILITY INFLUENCE TEAM

PERFORMANCE? EXPLORING THE MECHANISMS OF
CONTAGION AND COLLECTIVE PROMOTION FOCUS
BRADLEY P. OWENS
Brigham Young University
DAVID R. HEKMAN
University of Colorado
Using data from 607 subjects organized into 161 teams (84 laboratory teams, 77
organizational field teams), we examined how leader humility influences team in-
teraction patterns, emergent states, and team performance. We develop and test
a theoretical model, positing that, when leaders behave humbly, followers emulate
their humble behaviors, creating a shared interpersonal team process (collective
humility). This, in turn, creates a team emergent state focused on progressively
striving toward achieving the team’s highest potential (collective promotion focus),
which ultimately enhances team performance. We test our model across three
studies wherein we manipulate leader humility to test the social contagion hy-
pothesis (Study 1), examine the impact of humility on team processes and perfor-
mance in a longitudinal team simulation (Study 2), and test the full model in
a multistage field study in a health services context (Study 3). Our findings collec-
tively support our theoretical model, demonstrating that leader behavior can spread
via social contagion to followers, producing an emergent state that ultimately affects
team performance. We contribute to the leadership literature by suggesting the need
for leaders to lead by example and showing how a specific set of leader behaviors
influence team performance, providing a template for future leadership research on
a wide variety of leader behaviors.
Nothing is so contagious as example; and we never do or openness to new ideas and feedback (Owens,
any great good or evil which does not produce its like. Johnson, & Mitchell, 2013). Partly in response to
extensive research showing that leaders tend to
—François de La Rochefoucauld, Maximes, 1665
see themselves in an overly positive light (Board &
“Humility” has been defined as an inter- Fritzon, 2005; Chatterjee & Hambrick, 2011; Park,
personal characteristic that emerges in social Westphal, & Stern, 2011), inductive and empirical
contexts and connotes (a) a willingness to view research on leader humility has emerged sug-
oneself accurately, (b) an appreciation of others’ gesting that leader humility fosters supportive
strengths and contributions, and (c) teachability, organizational contexts, including top manage-
ment team integration and empowering climate
The authors thank Raymond Sparrowe and three anon- (Ou, Tsui, Kinicki, Waldman, Xiao, & Song, 2014);
ymous reviewers for their insightful feedback on this legitimizes follower growth and development
project. We also thank Chia Yen Chui for assistance in (Owens & Hekman, 2012); encourages follower
gathering data, Elsa Chan for writing the experimental loyalty and commitment (Basford, Offermann, &
scripts to manipulate leader humility, and Kim Cameron, Behrend, 2014); reinforces employee learning
Jane Dutton, Terry Mitchell, and David Waldman for their
orientation, job satisfaction, work engagement,
input on early drafts of this manuscript. This research was
and retention (Owens et al., 2013); and tempers
supported by a grant from the John Templeton Found-
ation (Grant 29630) entitled “The Development, Validation, the ill effects of leader narcissism, leading to
and Dissemination of Measures of Intellectual Humility and positive follower outcomes (Owens, Wallace, &
Humility” and by a fellowship from the Wheatley Institu- Waldman, 2015). Though these initial findings are
tion at Brigham Young University. encouraging regarding the value of leader humility in
1088
Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s express
written permission. Users may print, download, or email articles for individual use only.
2016 Owens and Hekman 1089
organizational contexts, understanding whether and introducing and operationalizing humility on the
exactly how leader humility influences an entire team level, and theorizing its function on team
team’s performance remains largely unexplored. emergent states and outcomes. Lastly, we illuminate
While leadership has been argued to be the most important leadership and team process antecedents
important contextual factor that influences team of collective promotion focus (i.e., a collective team
performance (Williams, Parker, & Turner, 2010), focus on progressively striving toward achieving the
past reviews have asserted that much more re- team’s highest potential), fostering theoretical and
search is needed to foster understanding of the practical insight regarding the development and
mechanisms linking leader behaviors to team application of this potent team property. We con-
performance (Burke, Stagl, Klein, Goodwin, Salas, ducted three studies to test our model: the first es-
& Halpin, 2006). While some traditional leader- tablishes the direction of our model’s causal arrow,
ship approaches have been “criticized for failing the second establishes that our model predicts ob-
to fully appreciate and model the dynamism and jective performance, and the third shows that our
complexities of team leadership” (Burke et al., model operates in an organizational field context.
2006: 302), our specific research question centers
on understanding the impact of the nontraditional
leadership approach of leader humility on team TOWARD A THEORY OF HOW LEADER
processes leading to team performance. We pro- HUMILITY INFLUENCES TEAM PERFORMANCE
pose that the modeling of humble behaviors by
Leader Humility
team leaders may be one uniquely impactful way
(among other leadership approaches) to foster ef- Reflecting the definition given above, leader
fective team functioning because humble leader humility is manifest by a set of power equalizing
behaviors are equally imitable by team members behaviors that co-occur and foster each other and
and are relevant to the core team processes of that are unified by the theme of growth. Leaders’
constructive interrelating, task allocation effec- knowledge of their own limits and recognition of
tiveness, information exchange, constant updat- others’ strengths fosters awareness of where they
ing and monitoring, and self-correction (Burke need to grow and of the people around them from
et al., 2006; Johnson, Hollenbeck, DeRue, Barnes, whom they can learn to grow. Acknowledging
& Jundt, 2013; Zaccaro, Rittman, & Marks, 2002). weaknesses leaves leaders open to learning from
We theorize that, through social contagion, the and appreciating those who are skilled in areas where
modeling of leader humility will foster the team the leader may be lacking. As an organizing theoreti-
interpersonal process of collective humility (i.e., cal basis, scholars have effectively used Baumeister’s
a group tendency toward owning limitations and (1998) self-experience framework to justify the di-
mistakes, appreciating group members’ strengths, mensions of humility as comprehensively capturing
and being teachable), which will in turn shape the core ways that individuals understand and expe-
a team orientation that reflects the essence of the rience themselves (see Ou et al., 2014). This frame-
leader’s modeled values in actions. As humility is work entails understanding the self in relation to the
centered in the idea of growth (Owens & Hekman, world (reflexive consciousness), in relation to others
2012), we propose that the influence of leader hu- (interpersonal being), and by what one does (execu-
mility and collective team humility will shape a team tive function). Reflecting each component of the self-
orientation of focusing on continual improvement, experience framework, our view of humility captures
advancement, and accomplishment (i.e., a collective how one views themselves in relation to the world
promotion focus; Beersma, Homan, Van Kleef, & de (more objectively), how they view others (more ap-
Dreu, 2013; Higgins, 1997). preciatively), and how they receive new information
By examining how leader humility influences or perspectives (more openly).
team performance, we extend leadership research by Humility has also been connected to the concept
detailing the process of how this relatively new of self-transcendence, or acknowledging some-
character-based leadership construct operates thing greater than the self and connecting with
(Quick & Wright, 2011). We describe how humble things outside the self (Dennett, 1995; Tangney,
behaviors, enacted by leaders and then emulated by 2000; Templeton, 1997). Accordingly, the pro-
team members, enhance the vital team processes posed dimensions of humility—drawing attention
mentioned above. We also contribute to the general to others’ strengths, being open to others’ ideas and
humility and the team process literatures by perspectives, and being willing to acknowledge
1090 Academy of Management Journal June
personal limits—are all manifestations of transcending behaviors, which then determine group perfor-
the self (see Gordon, 2010; Morris, Brotheridge, & mance (Deal & Kennedy, 2000; see also Berson,
Urbanski, 2005; Seligman, 2002).1 Leader 360-degree Da’as, & Waldman, 2015; Sun, Xu, & Shang, 2014).
evaluation correlational analyses (Owens & Hekman, For example, transformational leaders, through
2012: 801–802) and confirmatory factor analyses em- their powerful analogies, inspiring visions, and
pirically support the co-occurrence of these three hu- uplifting stated values (Schein, 1990), produce
mility dimensions (Owens et al., 2013). a “transformational culture,” which then in turn
Like Costa and McCrae’s (1992) conceptualization influences team performance behavior (i.e., Bass &
of extraversion and agreeableness as individual Avolio, 1993: 119; Parry & Proctor-Thomson, 2003).
traits that are manifest on the “interpersonal plane,” It may be that leaders are thought to first shape
we view humility as an individual characteristic that group culture because many existing leadership
“emerges in social interactions, is behavior based, theories describe leader behaviors that are not di-
and is recognizable to others” (Owens et al., 2013: rectly imitable by team members and do not repre-
1518). Though humility has been examined on the sent the modeling of effective team functioning
intrapersonal, cognitive level (Roberts & Wood, (e.g., charismatic behavior, setting up reward sys-
2003), our observable, social view of humility is ap- tems). However, when behaviors are more imitable,
propriate here since our purpose is to understand such as those comprised in the dimensions of hu-
humility as it pertains to leadership influence pro- mility, we theorize a more complex dynamic:
cesses and team member interaction patterns. A core humble leaders first unify follower behavior, which
premise of our theoretical model is that leaders’ then produces a team culture or shared goal that
modeling of humility is an influential team input reflects that behavior (in our specific case, a collec-
that will lead to positive team interpersonal pro- tive promotion focus that reflects collective humil-
cesses, task processes, and emergent states. ity), and ultimately influences team performance.
Many leadership writings assume that leaders We suggest that collective promotion focus is a lat-
shape group culture and values, and these drive ter mediator that comes about through group
member interactions (Mathieu, Maynard, Rapp, &
1
While philosophical- and theological-based writings Gilson, 2008: 414).2 Our model summary in Table 1
hold self-transcendence as the core underlying basis of contains each team construct type, aggregation
humility (Dennett, 1995; Tangney, 2000; Templeton, logic, definition, and function.
1997), inductive examination of leaders and followers Leader humility social contagion. Leaders are
suggests that the core underlying, unifying logic of hu- theorized to be vital in providing the “enabling
mility in organizational contexts is centered on the con- structure” (Burke et al., 2006: 289) for team function-
cept of growth (Owens & Hekman, 2012). Growth shares ing and performance largely by modeling positive
some similarity to transcendence in that it implies con- ways of interrelating: “leader behavior ‘models the
ceiving of a higher state or condition greater than one’s way’ organizational/group goals should be pursued”
current self. However, growth extends beyond transcen-
(Yaffe & Kark, 2011: 809), and leaders “model team-
dence in that it also implies reaching and stretching to-
work, or how team members should work together”
ward that higher state or ideal. An underlying sense of
transcendence alone may still reflect the weakness-based (Zaccaro et al., 2002: 468). Thus, leaders can have
connotations of humility (i.e., that humility is merely a vital influence in shaping how team members in-
a “monkish virtue” that serves no real or practical pur- teract through the leaders’ own social modeling (see
pose; see Hume, 1994: 219). In line with this, Grenberg Dragoni, 2005; Naumann & Ehrhart, 2005).
(2005: 181) states that, in the context of acknowledging This social modeling idea fits with evidence that
God or a higher ideal (i.e., connoting transcendence), followers emulate leaders’ emotions (Johnson, 2009;
“The humble person takes her awareness of limit as an Sy, Cote, & Saavedra, 2005) and behaviors (Fast &
impetus to action instead of as a warrant for despairing Tiedens, 2010; Visser, van Knippenberg, van Kleef,
inaction” (i.e., connoting growth). Thus, while classical & Wisse, 2013). For example, one study showed that
ideas about humility entail a transcendent perspective,
followers emulate leaders’ citizenship behavior
we propose that it is one’s growth-oriented response to
that transcendent perspective that determines virtuous
2
versus non-virtuous humility. It seems fitting that hu- “Emergent states are products of team experiences”
mility is centered in the concept of growth, since its lex- (Marks, Mathieu, & Zaccaro, 2001: 358), and “shared
ical roots (humus, meaning “earth,” and humi, meaning mental models refer to common understanding established
“from the ground”) reflect the origin from which most through experience among team members” (Zaccaro et al.,
things grow—the ground. 2002: 459).
TABLE 1
Model Summary
Collective Team
Construct Name Leader Humility → Collective Humility → Promotion Focus → Performance
Team Construct Type Team Input Teamwork: Interpersonal Emergent State: Shared Team Output
Process Motivation
Aggregation Logic/ Composition: Direct Composition: Referent-Shift Composition: Referent-Shift Compilation
Operationalization Consensus Consensus Consensus
Definition Leaders modeling to followers An interpersonal team process A collective team focus on
the behaviors of admitting that captures the three progressively striving
mistakes and limitations, underlying dimensions of toward achieving the team’s
spotlighting follower humility as behavioral highest potential
strengths and deflecting interaction patterns of
praise to others, and being admitting mistakes and
teachable—open to new limitations, spotlighting
ideas, advice, and feedback team member strengths and
deflecting praise to others,
and being open to new ideas,
advice, and feedback
Function Exemplification of behaviors Patterns of interrelating that A shared motivational logic
that foster effective team foster voice, disclosure of that unifies regulatory focus
interrelating, task limits, and identification of toward attaining the team’s
coordination, and unique member strengths or highest possible
self-management expertise performance
(Yaffe & Kark, 2011). Especially when faced with constructively. Most recent attention given to hu-
ambiguous situations, such as judging appropriate mility in organizational contexts focuses on hu-
workplace behavior, workers look to their leaders for mility as an individual-level trait in leadership
modeling context-appropriate behaviors (Festinger, (Morris et al., 2005; Nielsen, Marrone, & Slay,
1954; Hardin & Higgins, 1996; Salancik & Pfeffer, 2010; Ou et al., 2014; Owens et al., 2015; Reave,
1978). Followers may be especially likely to emulate 2005), team membership (Owens et al., 2013), and
their leaders because leaders have positional power in relation to task performance (Johnson, Rowatt,
(Cialdini & Trost, 1998). As teams watch their leaders & Petrini, 2011). However, some writers have
give away some of their power by admitting limita- suggested that humility can also be a group-level
tions and mistakes, allowing themselves to be taught phenomenon, both as a property of teams (Owens
rather than doing all the teaching, and drawing at- & McCornack, 2010) and an attribute of entire or-
tention to others’ contributions and strengths, they ganizations (Vera & Rodriguez-Lopez, 2004). We
reinforce a cooperative, others-oriented interactive propose that humility on the team level is an in-
logic; they send a message about the value of col- terpersonal team process that captures the three
lective striving over personal status seeking. In con- underlying dimensions of humility as behavioral
trast, some more traditional leadership approaches interaction patterns, similar to the behavior-based
may reinforce status-seeking or attention-getting be- conception of collective personality (Hofmann &
havior as team members try to mimic a leader’s at- Jones, 2005). We view individual and collective
tempts to be impressive, decisive, charismatic, or humility as functionally isomorphic (Morgeson &
inspirational. Hofmann, 1999), meaning both constructs pro-
We propose that, when leaders model humble be- duce “regularized, consistent patterns of behavior
havior, followers will emulate the behavior, which that can be observed and described by others”
creates the shared group behavior of collective hu- (Hofmann & Jones, 2005: 510). Given the equiva-
mility. Thus, collective humility describes team in- lence of observable behavior, we operationalize
teraction patterns that reflect the dimensions of a referent-shift consensus composition model
humility—that is, team members acknowledge and (Chan, 1998) that captures collective humility as
appreciate one another’s strengths, listen to one a group-level phenomenon.
another’s feedback and new ideas with open- However, collective humility differs from indi-
ness, and acknowledge mistakes and handle them vidual humility because it captures the idea that
expressions of humility may be socially bounded or and accomplishment” (Higgins, 1997: 1282). Cer-
mainly expressed toward members of one’s own tainly, individual-level promotion focus can be
work group where such behaviors have become induced—either in the laboratory or by a host of
normal and valued. Thus, collective humility more contextual factors (Cesario, Grant, & Higgins,
fully captures the dynamic, complementary, and 2004; Förster, Grant, Idson, & Higgins, 2001;
co-occurring nature of the humility dimensions Förster, Higgins, & Bianco, 2003). Likewise, we
playing out in a social context and captures the propose that collective promotion focus is a mal-
dynamics through which teams develop collective leable team property that emerges or is shaped by
qualities (Lewin, 1951). For example, acknowl- team inputs, such as leadership and team behav-
edging personal limitations may be easier, or at ioral patterns. We theorize that the influence of
least less socially risky, in a context where one’s leader humility on collective promotion focus will
strengths are recognized; transparency about lim- work through the teamwork process of collective
itations and acknowledging others’ strengths humility by providing the opportunity, motiva-
makes being open to learning from others more tion, and ability for groups to focus on achieving
natural. The norm of reciprocity also suggests that their highest potential.
a person receiving positive feedback about their First, the behavior of admitting mistakes and limi-
strengths or who is being listened to would be more tations provides the opportunity for a collective pro-
likely to respond in kind (Cialdini, Vincent, Lewis, motion focus to emerge. Because a team’s collective
Catalan, Wheeler, & Darby, 1975). In contrast, in effort comprises the aggregation of individual inputs,
a team in which a member’s true strengths are not acknowledging limitations and weaknesses allows
acknowledged and valued, and the member’s sug- teams to identify potential areas for future improve-
gestions are not listened to, admitting mistakes or ment and minimizes the sting of failure by enabling
weaknesses would be more difficult as the member team members to see mistakes and failures as a result
may already feel underestimated and under- of a worthwhile struggle toward growth. This effort
valued. In sum, we theorize that leader displays of fits with the idea of a collective promotion focus
humble behavior will be emulated by members of reflecting team members’ having a risk-seeking bias,
their team, reflecting a behavioral social contagion based on the beliefs that failures are necessary for
process. continued success (Faddegon, Scheepers, & Ellemers,
2008: 880) and that missing an opportunity is far
Hypothesis 1. Team leader humility will posi-
worse than making a mistake (Higgins, 1997).
tively predict collective humility.
Second, the collective humble behavior of acknowl-
Collective humility and collective promotion edging team members’ strengths and contributions
focus. Leader humble behavior is perceived by fol- motivates group members to focus on maximal group
lowers as modeling how to grow and leads followers performance. Clarity of capability leads to surer striv-
to feel that their own growth and improvement ing. Praising others for their strengths and contribu-
processes are legitimate and necessary (Owens & tions fosters an awareness of exemplars on the team,
Hekman, 2012). Thus, in an organizational context, social models for learning and leadership. Team
we expect that collective humility will orient teams members receiving credit for their contributions gain
toward focusing on maximizing team achievement, immediate psychic rewards for improving and per-
which is akin to emerging laboratory research on forming. From a social exchange perspective (Homans,
collective promotion focus (Beersma et al., 2013; 1961), these motivational boosts assure group mem-
Dimotakis, Davison, & Hollenbeck, 2012). A col- bers that applying extra effort to enable greater team
lective promotion focus is a shared team focus on achievement will be noticed and appreciated. Thus,
progressively striving toward achieving the team’s a collective promotion focus characterized by setting
highest potential, and leads group members to ap- and striving toward lofty performance aspirations is
proach opportunities (Rietzschel, 2011) and moti- motivated by the belief that group members who con-
vates them to be “more focused on what they want to tribute to the team will be praised, appreciated, and
achieve than on what could go wrong” (Beersma acknowledged.
et al., 2013: 196). Indeed, “promotion-focused in- Likewise, as teams become aware of each mem-
dividuals are motivated mainly by internal motives ber’s strengths and weaknesses through collective
like growth, development, and self-actualization” humility, they are able to make more informed and
(Kark & Van Dijk, 2007: 506; see also Van Dijk & effective task allocations across members. In collec-
Kluger, 2004), as well as “advancement, growth, tively humble teams, the underlying task selection
logic would likely be one of maximizing team individual performance because admitting weak-
achievements (i.e., a collective promotion focus) as nesses highlights growth opportunities, appreciat-
opposed to other potential logics (e.g., assigning the ing others’ strengths highlights growth exemplars,
most desirable tasks to group members with the and being teachable enables personal growth to
highest status; Klein, Ziegert, Knight, & Xiao, 2006). occur (Owens et al., 2013). On the team level, we
Ultimately, understanding the unique skills and propose that collective humility behavior enhances
abilities of team members could optimize team per- team performance through the mechanism of col-
formance, because, when team members understand lective promotion focus.
how they fit with the team (Kozlowski, Gully, Nason, Collective promotion focus involves team mem-
& Smith, 1999; Mohammed, Klimoski, & Rentsch, bers focusing on “collective” goals rather than
2000), the team performs better (Lorinkova, Pearsall, “personal” goals, and maximal “promotion” goals
& Sims, 2013). rather than minimal “prevention” goals (Rietzschel,
Third, beyond producing the opportunity and 2011). Goal-setting theory explains that individuals
motivation for a collective promotion focus to and teams are more likely to attain what they in-
emerge, collective humility fosters the ability for tently and specifically focus on because they de-
a collective promotion focus to permeate a group velop strategies for realizing their target (Locke &
through the interaction pattern of “teachability,” Latham, 2002). While individual goals tend to
or openness to new ideas and feedback. Compared benefit the individual (i.e., self-enhancing), collec-
with less open teams, teams that are open to tive goals sometimes require individuals to sub-
feedback and new information are more likely to jugate their personal interests in order to benefit the
sense collectively that seeking new attainments, team (i.e., team-enhancing). Teams typically per-
future possibilities, and improvements is normal form better when team members put the team’s
and legitimate. Speaking openly, listening, and interests ahead of their own (Ashforth & Mael,
seeking new ideas are behaviors clearly associated 1989). Similarly, teams that are focused on max-
with team learning and improvement (e.g., see imal team goals, such as achieving team “gains”
Edmondson, 1999; Hirst, van Knippenberg, & (i.e., promotion-focused teams), likely out-
Zhou, 2009). By seeking out and being receptive perform teams that are focused on minimal goals,
to new information, teams recognize new strate- such as “non-losses” (i.e., prevention-focused
gies for enhanced effectiveness. In contrast, being teams), because such promotion-focused teams tend
closed to new ideas would signal a focus on to have greater positive affect and task satisfaction
maintaining the status quo. This notion of a col- (Dimotakis et al., 2012: 426). When deciding whether
lective promotion focus orienting teams to strive to engage in a behavior, promotion-focused team
toward maximizing collective achievement is akin members determine whether the behavior helps the
to a type of team-level self-regulation, which is team and whether it enables the team to attain max-
thought to be critical for team performance be- imal performance.
cause it enables team members to self-manage In summary, we theorize that teams with humble
their actions in ways that benefit the group (Cohen, leaders will be characterized by collective humble
Ledford, & Spreitzer, 1996). behaviors generating a strong team-promotion
In sum, we theorize that collective humble be- focus. In line with Marks and colleagues’ (2001)
havior naturally produces the opportunity, motiva- distinctions between team processes and team
tion, and ability for a collective promotion focus to emergent states, we propose that humble leaders’
emerge. This emergent state forms a shared motiva- behaviors contagiously lead to collective humility
tional logic that acts as a reference point toward (a teamwork process marked by behavioral simi-
which teams regulate themselves (Schein, 1990) to larity), which then leads to a strong collective
achieve their highest potential. promotion focus (a motivational emergent state).
This emergent state operates like a self-regulatory
Hypothesis 2. Collective humility will positively
reference point toward which the team regulates
predict team collective promotion focus.
their behavior. Due to this collective promotion
How leader humility influences team performance. focus, we propose that teams will continue to
Humility on the individual level has been empiri- self-correct, self-reinforce, and self-monitor their
cally shown to foster higher performance in both actions toward realizing their achievement-
work and academic contexts (Johnson et al., 2011; maximizing goal and attaining tangible gains in
Owens et al., 2013). Humility is thought to enhance team performance. Thus:
Hypothesis 3. Collective humility and collective let’s do it my way”), praised/put down the follower
promotion focus will mediate the relationship (e.g., “You were awesome on that project” vs.
between leader humility and team performance “Don’t be such a slacker”), and vocalized limits/
such that leader humility will foster collective bragged about strengths (e.g., “I’m not sure I’m an
humility, which will in turn foster collective HR expert” vs. “I’m so glad I’m the leader. This role
promotion focus, which will in turn enhance really fits me”). A summarized script is given in
team performance. Appendix A; a full script is available from the first
author upon request. A manipulation check from
To test our hypotheses, we conducted three stud-
two pretests (e.g., humility ratings by four ob-
ies. We first sought to determine whether manipu-
servers) showed that each of the four confederates
lated leader humility led to higher levels of collective
were effective in acting the part of the humble or
humility and, ultimately, to a team collective pro-
non-humble leader role (mean humility ratings in
motion focus. Second, we conducted a team simu-
these pretests were 4.52 for the humble condition
lation study to examine whether our proposed team
and 3.00 for the non-humble condition; additional
processes predicted team performance and to rule
manipulation checks are reported below).
out alternative mechanisms. Finally, we tested our
Each experimental condition included two re-
entire model in a field study.
search confederates and two to three participants.
When they arrived at the laboratory room, the par-
STUDY 1 (HAMMERCORP): LEADER HUMILITY, ticipants and the confederates were told that they
COLLECTIVE HUMILITY, AND TEAM had been assigned to play the role of leader or fol-
COLLECTIVE PROMOTION FOCUS lower; the roles of team leader and one of the fol-
lowers were always assigned to the confederates.
Participants
Participants were then asked to participate in
The experiment was conducted with 89 un- a strategic HR program rank-ordering task for a fic-
dergraduate business students taking a senior- tional chain of hardware stores (i.e., HammerCorp).
level organizational behavior course in a public We carefully selected a task with which partici-
northeastern U.S. university. Of the participants, pants would have a chance to interact in a way
57% were Asian, 38% Caucasian, and 2% African that humility could emerge (we encouraged in-
American, with an average age of 22 years, and teraction), where participants have varied back-
39% being female. Students received course credit grounds (we sampled a wide cross-section of HR,
for their participation. finance, accounting, marketing, and IT majors), and
where the task had no clear right answer. We did
this to foster an environment in which leader hu-
Experiment Design and Procedure
mility would be relevant and observable. It is im-
The participants were randomly divided into 31 portant to note that team members were together for
work teams and assigned to one of two experi- over an hour, which is a meaningful amount of time
mental conditions (16 teams in the humble leader within which to make social judgments (Kenny,
condition and 15 teams in the non-humble leader 2004). At the beginning of the tasks and at several
condition). In order to create the manipulated points throughout, the confederate leader and fol-
condition for leader humility, we recruited four lower engaged in scripted exchanges according to
research confederates to play the roles of the team the script given for the respective experimental
leader and one of the team members. The confed- condition.
erates were all Caucasian female students, with the After accomplishing the tasks, the participants
average age of 22. Confederates had no knowledge and the confederates were asked to fill out a survey
of the study hypotheses. Before the experiment that assessed team leader humility (as a manipula-
was initiated, we conducted several training ses- tion check), collective humility, and team collective
sions over a two-week period to help these con- promotion focus. Confederates’ survey responses
federates act one of three different roles: the were not included in the data analysis. After com-
humble leader, the non-humble leader, and the pleting the experiment, the experimenter debriefed
talkative follower. The scripts for the humble/ the participants and informed them about the pur-
non-humble leaders included statements to the pose of the study. The confederates’ identities were
confederate follower that validated/invalidated revealed and participants were asked not to discuss
follower ideas (e.g., “Good suggestion” vs. “No, the experiment with other potential subjects.
Measures promotion focus was operationalized as a referent-

shift consensus model, and, thus, scores were av-
Leader humility. We measured leader humility
eraged within each team. The full set of items is
using a nine-item other-report humility scale that
given in Appendix B. The alpha reliability for this
reflects the three proposed dimensions of humility
scale was .82.
and has shown strong predictive validity for humil-
Though this measure captures our view of the core
ity in a leadership role (Owens et al., 2013). Sample
indicators of collective promotion focus, we wanted
items include “This leader admits it when he/she
to ensure that the new scale reflected our intended
does not know how to do something,” “This leader
construct from the perspective of respondents. To
shows a willingness to learn from others,” and “This
test the face validity of our scale items, we conducted
leader often compliments others on their strengths.”
a sorting task. Following previous guidelines for
The scale reflects a direct-consensus model (Chan, studies involving item sorting (Anderson & Gerbing,
1998), since the construct is captured via consensus 1991), we recruited 13 researchers to sort the 4 items
of member perceptions of the leader. The full set of along with foil items from two other team emergent
items used in this study is included in Appendix B. state scales (a 7-item team psychological safety scale
The alpha reliability for this scale was .95. from Edmondson, 1999, and a 5-item team-learning
Collective humility. Collective humility was orientation scale from Bunderson & Sutcliffe, 2003).
measured using an adaptation of the nine-item peer- After providing a definition for each construct, par-
report scale (Owens et al., 2013) aimed at assessing ticipants were asked to assign each item to one of the
individual humility. In a preliminary study, Owens construct categories according to their respective
and McCornack (2010) reported that team-level hu- construct definitions. The collective promotion fo-
mility is psychometrically distinct from the con- cus items were correctly categorized in 50 of 52 in-
structs of psychological safety (Edmondson, 1999), stances (96%) of item categorizations. We viewed
team cohesion (Podsakoff & MacKenzie, 1994), team this result as strong evidence of the face validity of
demands–abilities fit (Abdel-Halim, 1981), and team the collective promotion focus items.
potency (Campion, Medsker, & Higgs, 1993). Items Controls. Our initial analysis approach, path
were adapted to reflect a team referent: “Members of analysis, and our smaller sample size limited our
this team admit it when they don’t know how to do ability to control for potential covariates of our
something,” “Members of this team show apprecia- study variables. However, our complementary,
tion for the unique contributions of other team bootstrapped, regression-based path analysis en-
members,” and “Members of this team are willing to abled us to control for average team size, which has
learn from one another.” The full set of items from been shown to influence team process and function
this study appears in Appendix B. As a referent-shift (Cummings, Huber, & Arendt, 1974; Hackman &
consensus model (Chan, 1998), collective humility Vidmar, 1970; Menon & Phillips, 2011), and average
was operationalized to reflect shared perceptions; team gender, because female team members may be
thus, collective humility scores were averaged within more responsive to humble behaviors, which are
each team. The alpha reliability for this measure more communal and more congruent to female social
was .92. preferences (Eagly, 2009). In this analysis, we also
Team collective promotion focus. We assessed controlled for average team age because humility has
collective promotion focus using van Kleef, van been theorized to be valued more by older individuals
Trijp, and Luning’s (2005) shortened version of the (Tangney, 2000).
Lockwood, Jordan, and Kunda (2002) measure of
individual-level promotion focus; again, we adapted
it to the team level. This adaptation was necessary Analyses
because team collective promotion focus has been To ensure construct distinctiveness, we conducted
exclusively manipulated in laboratory-based groups a series of confirmatory factor analyses (CFAs) using
and has never been measured (Beersma et al., 2013; Amos 19 (Arbuckle, 2010). Since our sample size
Dimotakis et al., 2012; Levine, Higgins, & Choi, 2000; was too low for CFA testing (n 5 89), to conduct CFA
Rietzschel, 2011). Items asked participants to in- tests, we supplemented this sample by administer-
dicate their team’s collective focus on “attaining ing our survey to an additional 153 students from
our ambitions,” “attaining the success we hope to the same university who were working in three-
achieve in the future,” and “achieving our hopes and month project teams but were not involved in the lab
aspirations.” Like collective humility, team collective study, for a total CFA sample size of 242. We only
used the supplemental sample when conducting the 2009; or MacKinnon, Lockwood, Hoffman, West, &
CFA test, not when conducting inter-rater agreement Sheets, 2002).
or any other analyses. We averaged the ratings of the
89 lab study participants after computing within-
Results
group inter-rater agreement using the null distribution
(rWG; James, Demaree, & Wolf, 1984) and intraclass The results of the CFA suggested that leader
correlations (ICC; Bliese, 2000) to justify aggrega- humility, collective humility, and team collective
tion. We conducted a one-way analysis-of-variance promotion focus were distinct constructs, as the
mean comparison test to determine whether mea- three-factor solution (x2: 295.96; df 5 205; root mean
sured leader humility, collective humility, and square error of approximation [RMSEA] 5 .06;
collective promotion focus were significantly dif- comparative fit index [CFI] 5 .97; Tucker–Lewis in-
ferent across the two conditions. dex [TLI] 5 .96) fit the data better than all other rival
In testing our hypotheses, we use structural models as the change in the x2 value was significant
equation modeling (SEM) path analyses (using Amos at the p , .001 level when comparing all other
19) to derive overall model fit and path coefficients models. Average rWG scores for our study variables
for each hypothesized study relationship. Given the ranged from .72 to .84. The ICC(1) scores for our team
team-level nature of our model, we use manifest construct variables ranged from .24 to .47, and the
variables in our estimations. We report the Akaike’s ICC(2) scores ranged from .41 to .72. Though our ICC
information criterion (AIC) to show fitness for the scores were low, they are similar to the ICC values of
hypothesized mediation model relative to a direct other studies with similarly low average team sizes
effects rival model (a smaller AIC value signifies (see Hofmann & Jones, 2005: 513). Lower average
a better model fit). Information criteria indices such group sizes are said to result in a less reliable mean
as AIC are used to compare nonhierarchical models (leading to lower ICC(2) values), likely attenuating
for fitness and can be computed with models that have relationships at the group level (Bliese, 1998). This
very low degrees of freedom (Hooper, Coughlan, & possible attenuation means that our results should be
Mullen, 2008). As a more rigorous test of our media- seen as conservative. Collinearity diagnostic tests of
tion hypotheses, we use bootstrapped regression- measured study variables revealed that all variance
based path analyses (i.e., PROCESS software; Hayes, inflation factor (VIF) values ranged from 1.00 to 1.86,
2013; see also Preacher & Hayes, 2008) in which we well below standard cutoffs.
included our control variables as covariates. This Table 2 contains the results of our bivariate cor-
approach entails randomly sampling 5,000 boot- relation analysis, which revealed that the manip-
strapped cases from the original data to derive ulated leader condition (“1” 5 humble leader
a bias-corrected and accelerated 95% confidence condition; “0” 5 non-humble leader condition)
interval (CI) that reflects the mediation effect. This was positively correlated with measured leader
method helps to offset the weaknesses of the humility, collective humility, and collective pro-
causal steps approach (for a review, see Hayes, motion focus. As a manipulation check, results of
TABLE 2
Study 1: HammerCorp Means, Standard Deviations, and Correlationsa
M SD 1 2 3 4 5 6 7
1. Collective promotion focus 3.56 .88 ;

2. Team size 2.83 .79 2.14 ;
3. Percent female .39 .49 .15 .03 ;
4. Average age 21.63 3.16 2.04 2.14 .03 ;
5. Undergraduate yearb 3.51 .53 .23 2.17 2.03 .34 ;
6. Leader humility (manipulated)c .52 .50 .33 2.00 2.10 2.11 2.05 ;
7. Leader humility (measured) 3.56 1.06 .49 .04 .23 2.25 .02 .58 ;
8. Collective humility 3.57 .90 .59 2.03 .23 2.16 .13 .40 .63
a
N 5 31 teams. All correlations greater than .23 are significant at p , .05.
b
“1” 5 freshman, “4” 5 senior.
c
“1” 5 humble leader condition, “0” 5 non-humble leader condition. The between-team variances for leader humility, collective humility,
and collective promotion focus are 2.03, 1.12, and 1.11, respectively.
a one-way analysis of variance test revealed that psychological safety) and a greater number of controls
the means in measured leader humility were sig- (e.g., prior team performance, number of competitors,
nificantly higher in the humble leader condition team size, percent female, and average age) to en-
(M 5 4.14) than the non-humble leader condition hance confidence in our proposed theoretical path.
(M 5 2.91, p , .001), suggesting our manipulation Our aim with Study 2 was to gain further support of
was successful. The mean levels of collective hu- our model by testing how our proposed team pro-
mility and collective promotion focus were also cesses predicted objective team performance in-
significantly higher in the humble leader condition dependent of other team-level mediators.
than in the non-humble leader condition (p ,
.001). From our SEM path analysis, manipulated
leader humility was related to collective humility STUDY 2 (CARCORP): TEAM COLLECTIVE
(b 5 .51, p , .001), and collective humility was PROMOTION FOCUS MEDIATES THE
related to collective promotion focus (b 5 .69, p , INFLUENCE OF TEAM COLLECTIVE HUMILITY
.001). The AIC of the leader humility manipulated ON TEAM PERFORMANCE
mediation model (AIC 5 10.07) was smaller than Participants and Procedures
the rival model with direct paths modeled from
leader and collective humility to promotion focus Participants were 192 undergraduate business
(AIC 5 12.00). The same finding was observed for students enrolled in three upper-level business
the leader humility measured mediational model strategy classes in a university in the midwestern
(AIC 5 10.00) compared with the rival direct ef- United States. The students were randomly assigned
fects model (AIC 5 30.33), suggesting superior fit to 53 teams averaging 3.62 individuals; 61% were
for the hypothesized model. Based on the results of male, 39% were female; average age was 23.21 years.
our indirect effects analyses (using 5,000 boot- Teams participated in a multistage computer simu-
strapped samples), we observed that manipulated lation (CarCorp) created by industry experts to re-
leader humility predicted collective promotion flect real auto-manufacturing market trends. The
focus through the mediator of collective humility simulation, nearly identical to business simulations
(b 5 .40; R2 5 .26, 95% CI [0.14, 0.85]). The 95% used in previously published management articles
confidence interval for this path excluded zero, (e.g., Bunderson, Van der Vegt, & Sparrowe, 2013;
and thus Hypotheses 1 and 2 are supported. Any Lorinkova et al., 2013), required multiple strategic
other ordered configuration of the path yielded an decisions over the course of 10 weeks (for an over-
indirect effect confidence interval that included view, see Interpretive Simulations, 2014). Teams
zero (e.g., collective promotion focus predicting competed for market share and stock value. Each
collective humility), which lent additional support week the stock values were posted based on the ef-
for our specific model. fectiveness of decisions made the previous week. At
Week 6, students independently rated their team on
the team variables (collective humility, promotion
Discussion focus, cohesion, and psychological safety). Partici-
The purpose of this study was to test the proposed pants received course credit for completing the
relationship between leader humility, collective hu- online assessments (100% response rate).
mility, and team collective promotion focus, and to
lend support for the causal flow of our model by ma-
Measures
nipulating leader humility. As hypothesized, leader
humility positively predicted team collective pro- All study survey measures were scaled to a five-
motion focus through collective humility. One limi- point agreement scale (1 5 “strongly disagree”; 5 5
tation of this model was that the teams were together “strongly agree”).
during only a short period of time, yielding limited Collective humility. Collective humility was
opportunity for team processes and emergent states to measured using the same nine-item scale that was
truly unfold. Thus, to build off these findings, we used used in Study 1. The alpha reliability for this mea-
longitudinal teams engaged in a simulation to exam- sure was .92.
ine the effect of collective humility and collective Team collective promotion focus. Team collec-
promotion focus on team performance employing tive promotion focus was measured using the same
a controlled team simulation. We also included po- four-item scale used in Study 1. The alpha reliability
tential rival mediators (e.g., team cohesion, team for this scale was .92.
Team performance. Team performance was Results

captured by the ending stock price reflecting the
The results of the CFA suggested that the hy-
value of each team’s company at the end of the sim-
pothesized five-factor model was the best fit, when
ulation. The simulation was designed to capture ef-
compared to the rival models (x2: 696.91, df 5 392;
fective strategic decision-making responsiveness to
RMSEA 5 .06; TLI 5 .93; CFI 5 .93). The change in
changing situational demands throughout the sim-
the x2 value was significant at the p , .001 level
ulation. Every team started at the same stock price
when comparing all rival models to the proposed
($50 per share), and the ending stock price ranged
model. Average rWG scores for all team variables
from $1.79 to $172.25 per share.
ranged from .79 to .94. The ICC(1) scores for our
Controls. Like Study 1, we measured and con-
team variables ranged from .09 to .45, and the ICC(2)
trolled for average team size, average team gender,
and average team age in our bootstrapped path an- scores ranged from .46 to .71. As mentioned, our low
alyses for indirect effects. In order to demonstrate average team size may explain the lower ICC(2)
the generative effects of our proposed model on values (see Hofmann & Jones, 2005) and render our
team performance, we also controlled for the team results more conservative. Given this, and the fact
stock price after the third round of the simulation. In that our rWG scores were above standard cutoffs, we
addition, to account for the level of industry com- proceeded to aggregate member ratings of team
petitiveness, we controlled for the number of other level variables.
teams that each team was competing against in the Table 3 displays the means, standard deviations,
simulation. We also controlled for the rival media- and descriptive statistics of our study variables,
tors of team psychological safety and cohesion by and Table 4 contains the results of our indirect ef-
using Edmondson’s (1999) seven-item psychologi- fects analyses (based on 5,000 bootstrapped sam-
cal safety measure (e.g., “It is safe to take a risk on ples). From our SEM path analysis results, the path
this team” and “Members of this team are able to coefficients suggest that collective humility was
bring up problems and tough issues”; a 5 .75) and positively related to collective promotion focus
Podsakoff and MacKenzie’s (1994) six-item team (b 5 .55, p , .001), and collective promotion focus
cohesion scale (e.g., “There is a great deal of trust was related to team performance (b 5 .45, p , .001).
among members of my project team” and “The The AIC of the hypothesized mediation model was
members of my project team are cooperative with smaller than the rival model with direct paths from
one another”; a 5 .92). collective humility and collective promotion focus
to team performance (17.58 vs. 34.89), suggesting
superior fit for the hypothesized model. As Model
Analyses 1 shows (in Table 4), collective humility positively
We conducted a series of CFAs on all measured predicted team performance through the mediator
study variables to ensure that they were operation- of collective promotion focus (b 5 .42, 95% CI
ally distinct. Given that individual team members [0.07, 1.08]) and when controlling for psychologi-
assessed team-level variables, we averaged the rat- cal safety and team cohesion. The overall model
ings after computing within-group inter-rater agree- explained 46% of the variance in team perfor-
ment (rWG; James et al., 1984) and ICCs (Bliese, 2000) mance (R2 5 .46), supporting Hypothesis 2. To test
to justify aggregation. As in Study 1, we conducted alternative paths using other mediators, we exam-
an initial SEM path analysis (Amos 19; Arbuckle, ined six other potential mediated models. Specif-
2010) and reported the AIC to show fitness for the ically, we tested whether team cohesion and team
hypothesized mediation model relative to a direct psychological safety culture mediated between
effects rival model, and then reported the results of collective humility and team performance instead
our bootstrapped regression-based path analyses of collective promotion focus. We also tested
(i.e., PROCESS software; Hayes, 2013; see also models that included cohesion and psychological
Preacher & Hayes, 2008) as a more rigorous test of our safety as additional mediators along with collec-
mediation hypotheses, as these analyses included all tive promotion focus. In all alternate models, the
of our control variables. Multicollinearity statistics 95% confidence intervals straddled zero, in-
were also calculated for all survey-derived variables. dicating these paths were not statistically signifi-
All VIF values were below the standard cutoffs that cant (see Table 4). Thus, our hypothesized team
signal potential multicollinearity problems (i.e., all process mediation model was supported while
VIF values were below 2.0). rival models were not.
TABLE 3
Study 2: CarCorp Means, Standard Deviations, and Correlationsa
M SD 1 2 3 4 5 6 7 8 9
1. T2 stock priceb 54.00 44.68 ;

2. Team size 3.62 .60 2.34 ;
3. T1 stock priceb 45.56 1.94 .25 2.09 ;
4. Number of competitors 1.58 .66 2.01 .03 .39 ;
5. Percent female .61 .26 .09 .03 2.01 .04 ;
6. Average age (years) 23.21 1.42 .17 .02 2.05 2.28 2.07 ;
7. Team psychological safetyc 5.60 .55 .41 2.26 .08 2.11 .08 .07 ;
8. Team cohesionc 5.46 .83 .24 2.20 .01 2.02 .13 2.10 .83 ;
9. Collective humilityc 4.03 .26 .41 2.02 .06 2.08 2.07 .13 .58 .59 ;
10. Collective promotion focusc 5.59 .73 .45 2.13 .09 2.06 .07 2.08 .55 .57 .55
a
N 5 53 teams. All correlations greater than .26 are significant at p , .05.
b
T1 and T2 refer to Time 1 and Time 2 survey administrations.
c
The between-team variances for collective humility, collective promotion focus, team cohesion, and psychological safety are .33, 1.38,
2.79, and 1.18, respectively.
Discussion humility and team performance. We included team

cohesion and team psychological safety as controls
The purpose of this study was to test the proposed
in our model to examine whether our model pre-
relationship between collective humility and team
performance, with team collective promotion focus dicted performance unique from these other vari-
as a mediator. As hypothesized, in this controlled lab ables. We were unable to identify any path where
simulation, collective humility positively predicted these rival variables mediated the relationship be-
team collective promotion focus and team perfor- tween collective humility and team performance.
mance. Furthermore, team collective promotion fo- Thus, our proposed model explains team perfor-
cus mediated the relationship between collective mance unique from these established constructs. As
TABLE 4
Study 2: CarCorp Bootstrapped Indirect Effect Model Comparison Analyses on the Mediating Role of Team Collective
Promotion Focus in the Collective Humility–Team Performance Relationshipa
Indirect Bootstrapped Bias-corrected and accelerated 95%
Indirect Paths Effect b SE t confidence interval
1. Collective humility → Collective promotion .42 .21 1.99* 95% CI [0.07, 1.08]
focus → Team performance
2. Collective humility → Team cohesion → 2.35 .12 21.12 95% CI [21.25, 0.03]
Team performance
3. Collective humility → Team psychological .10 .20 .51 95% CI [20.10, 0.75]
safety → Team performance
4. Collective humility → Collective promotion 2.06 .18 .77 95% CI [20.35, 0.02]
focus → Team cohesion → Team
performance
5. Collective humility → Collective promotion .04 .09 .44 95% CI [20.03, 0.37]
focus → Team psychological safety → Team
performance
6. Collective humility → Team cohesion → .05 .07 .71 95% CI [20.03, 0.29]
Collective promotion focus → Team
performance
7. Collective humility → Team psychological .02 .05 .42 95% CI [20.01, 0.21]
safety → Collective promotion focus → Team
performance
a
Bootstrapped results are based on 5,000 samples (see Preacher & Hayes, 2008). All models contain the following predictor/control variables:
team size, T1 stock price, number of competitors, percent female, average age, team psychological safety, team cohesion, collective humility,
and collective promotion focus.
*p , .05
collective humility positively predicted objective This health services company, hereafter called
performance in this competitive auto industry sim- HealthCorp, has offices and clientele throughout
ulation, this may provide support for the proposed the United States and in Puerto Rico and Great
idea that humility can be a source of competitive Britain. As part of an annual organizational as-
advantage (Vera & Rodriguez-Lopez, 2004). sessment of culture, we added measures of leader
Studies 1 and 2 suggest isolated support for key humility, transformational leadership, collective
parts of our theoretical model: manipulated leader humility, team collective promotion focus, and
humility predicts collective humility and team col- leader-rated team performance. Ideally, we would
lective promotion focus (Study 1); collective humil- have temporally separated all subjective measures,
ity predicts team performance, and this relationship but HealthCorp allowed us to administer only two
is mediated by team collective promotion focus separate surveys. Therefore, Time 1 included an
(Study 2). To build upon these studies, we sought to assessment of employee-rated leader humility and
replicate and extend these findings by testing the transformational leadership (response rate 67%).
entire model with a field sample. We also included Time 2, approximately one month later, assessed
a measure of transformational leadership to distin- employee-rated collective humility and team col-
guish the impact of leader humility from this estab- lective promotion focus (response rate 54%). At
lished leadership construct. As Schaubroeck, Lam, Time 2, leaders rated their team’s performance
and Peng (2011: 869) summarized, “To date, trans- (response rate 74%).
formational leadership is arguably the most reliable
and potent mainstream leadership behavior variable Measures
for predicting team performance.” One theorized
difference between these leadership approaches is All study measures were scaled to a five-point
the imitability of behaviors. With the exception of agreement scale (1 5 “strongly disagree”; 5 5
consideration, the dimensions of idealized influ- “strongly agree”).
ence, inspirational motivation, and intellectual Leader humility. Leader humility was measured
stimulation are arguably not as imitable by team using the same nine-item scale used in Studies 1 and
members compared with the behaviors comprising 2. The alpha reliability for this scale was .97.
humility. These transformational leadership di- Collective humility. Collective humility was
mensions “collectively inspire followers to achieve measured using the same nine-item scale used in
more than was thought possible [by] encourag[ing] Studies 1 and 2. The alpha reliability for this measure
followers to question assumptions and think about was .96.
new ways of doing tasks” (Williams et al., 2010: 306). Team collective promotion focus. Team collec-
Though leader encouragement is important, the in- tive promotion focus was measured using the same
fluence of humility comes more through modeling four-item scale used in Studies 1 and 2. The alpha
rather than encouragement or exhortation, which in reliability for this scale was .93.
turn leads to social emulation of the leader’s behav- Team performance. Team performance was
iors. Thus, while Schaubroeck et al. (2011) found measured using an adapted four-item scale
that transformational leadership influenced team (Walumbwa, Avolio, & Zhu, 2008). Sample items
include “In your estimation, how effectively does
performance through cognitive-based trust and
this team get their work done?” and “How would
team potency, we wanted to see whether it would
you judge the overall quality of the work per-
also predict team performance through collective
formed by this team?” The full set of items appears
humility and collective promotion focus, or if this
in Appendix B, and the alpha reliability score for
pathway for enhancing team performance was
this scale was .96.
unique to leader humility.
Controls. As in the previous studies, we measured
and controlled average team size, average gender,
STUDY 3 (HEALTHCORP): TESTING THE FULL and average age in our bootstrapped path analyses.
MODEL IN THE FIELD We also controlled for employee ratings of leader
transformational leadership to observe the impact of
Participants and Procedures
leader humility beyond this commonly studied
Participants were 326 health services employees construct. We measured transformational leadership
organized into 77 work teams (average team size 5 using the 16-item assessment that captured the four
4.23, 66% female, average age 38 years, 77% Caucasian). dimensions of transformational leadership from the
MLQ-53 (Bass & Avolio, 1990). The alpha reliability As mentioned, our low average team size may ex-
for this scale was .96. plain the lower ICC(2) values (see Hofmann & Jones,
2005). Multicollinearity diagnostic tests revealed
that all VIF values were below the standard cutoffs
Analyses
that signal potential multicollinearity problems
We conducted a CFA to determine the distinc- (i.e., all VIF values were below 3.0).
tiveness of all study variables and calculated rWG and As Table 5 shows, bivariate correlation ana-
ICC values to justify aggregating individual assess- lyses provided initial support for our proposed
ments to the team level. As with Studies 1 and 2, we relationships.
conducted an initial SEM path analysis (Amos 19; From our path analysis, the path coefficients
Arbuckle, 2010), and report the AIC to show fitness suggest that leader humility was related to collec-
for the hypothesized mediation model relative to tive humility (b 5 .30, p , .05), collective humility
a direct effects rival model. We then report the results was related to collective promotion focus (b 5 .67,
of our bootstrapped regression-based path analyses p , .001), and collective promotion focus was re-
(i.e., PROCESS software; Hayes, 2013; see also lated to team performance (b 5 .47, p , .001). The
Preacher & Hayes, 2008) as a more rigorous test of our AIC of the hypothesized mediation model was
mediation hypotheses. smaller than the rival model with direct paths
modeled from leader humility, collective humil-
ity, and collective promotion focus to team per-
Results
formance (17.37 vs. 55.15), suggesting superior fit
The results of CFA analyses suggested that five- for the hypothesized mediation model. Table 6
factor model (x2: 1454.75, df 5 838; RMSEA 5 .06, contains the results of our indirect effects analyses
CFI 5 .93, TLI 5 .93) fit the data better than all other (based on 5,000 bootstrapped samples). In this ta-
rival models (e.g., those combining our leadership ble, Model 1 shows that, when all control variables
constructs, our team constructs, or both) as these ri- and transformational leadership were included,
val models had CFI and TLI values that were below leader humility had an indirect effect on team
.90, RMSEA values above .06, and significant ∆x2 performance through collective humility and team
values at the p , .001 level. For our aggregation tests, collective promotion focus (b 5 .04; 95% CI [0.01,
the average rWG scores for our team-level variables 0.13]). The overall model explained 26% of the
ranged from .79 to .96. The ICC(1) scores for each variance in team performance (R2 5 .26). We also
team-level construct ranged from .07 to .19. The ICC examined whether leader humility influences
(2) scores for each team-level construct ranged from team performance through only one of our pro-
.25 to .50. For each team-level variable, between- posed mediators (i.e., either collective humility or
group-level effects were significant at the p , .05 collective promotion focus) and found no support
level, suggesting a team-level effect. Thus, we felt for these single-mediator models. (See Models 2
justified in aggregating the scores to the team level. and 3.) To more fully establish the unique effect of
TABLE 5
Study 3: HealthCorp Means, Standard Deviations, and Correlationsa
M SD 1 2 3 4 5 6 7
1. Team performance 3.64 .55 ;

2. Percent femaleb .66 .32 .09 ;
3. Team size 4.23 2.73 2.13 .23 ;
4. Average age 38.18 5.29 2.03 2.02 2.18 ;
5. Leader humilityc 3.87 .59 .21 .17 .02 2.07 ;
6. Collective humilityc 3.72 .55 .42 .20 2.15 .17 .33 ;
7. Collective promotion focusc 5.37 .80 .44 .25 2.25 .03 .36 .68 ;
8. Transformational leadershipc 3.51 .70 .25 .21 .17 2.09 .53 .16 .22
a
N 5 77 teams. All correlations larger than .23 are significant at p , .05.
b
Gender was coded “1” 5 female, “0” 5 male.
c
The between-team variances for leader humility, collective humility, collective promotion focus, and transformational leadership were
1.08, 1.39, 1.41, and 1.53, respectively.
TABLE 6
Study 3: HealthCorp Bootstrapped Indirect Effect Model Comparison Analyses on the Influence of Leader Humility on Team
Performance through Collective Humility and Collective Promotion Focusa
Indirect Bootstrapped Indirect Effects (Bias-corrected and
Indirect Paths Effect b SE t accelerated 95% confidence interval)
1. Leader humility → Collective humility → .04 .02 2.02* 95% CI [0.01, 0.13]
Collective Promotion focus → Team
performance
2. Leader humility → Collective promotion .02 .03 .73 95% CI [20.01, 0.12]
focus → Team performance
3. Leader humility → Collective humility → .03 .03 .99 95% CI [20.01, 0.13]
Team performance
4. Transformational leadership → Collective .00 .00 .00 95% CI [20.04, 0.02]
humility → Collective promotion focus →
Team performance
5. Transformational leadership → Collective .01 .02 .47 95% CI [20.02, 0.07]
promotion focus → Team performance
6. Transformational leadership → Collective 2.01 .02 .46 95% CI [20.08, 0.02]
humility → Team performance
7. Leader humility → Transformational 2.00 .01 2.16 95% CI [20.03, 0.01]
leadership → Collective humility →
Collective promotion focus → Team
performance
a
Bootstrapped results are based on 5,000 samples (see Preacher & Hayes, 2008). All models contain the following predictor/control
variables: percent female, team size, average age, leader humility, collective humility, collective promotion focus, and transformational
leadership.
*p , .05
leader humility, we reran all bootstrapped ana- Post hoc test. In our rationale for Hypothesis 2, we
lyses with transformational leadership as the main theorized that collective humility would foster
predictor (replacing leader humility in the model; better awareness of team member unique strengths
see Models 4–6), and the confidence intervals for leading to task allocation effectiveness, which
all potential mediational paths straddled zero would in turn foster a focus on optimizing team
(whether or not we controlled for leader humility). achievement, or collective promotion focus. In our
Finally, we examined whether transformational Study 3 data, we also had a measure of “team
leadership might mediate between leader humility demands–abilities fit” (Abdel-Halim’s [1981] five-
and team performance in addition to our proposed item individual-level measure aggregated to the
mediators of collective humility and collective team level;3 sample items include “I feel that my
promotion focus (Model 7), and, again, the confi- team tasks and I are well matched,” and “My team
dence interval straddled zero. tasks give me a chance to do the things I feel I do
Furthermore, to strengthen the causal direction of best”). As a post hoc test, we added this potential
our model, we also reran the model in the reverse mediator to our bootstrapped indirect effect ana-
direction (team performance → team collective pro- lyses. We found that team demands–abilities fit
motion focus → collective humility → leader hu- mediated between collective humility and collec-
mility), and the results were not significant. Thus, tive promotion focus when including all controls. In
the results of the AIC and bootstrapping analyses other words, the 95% confidence interval for the
support Hypotheses 1–3 and the causal direction of bootstrapped indirect effect analysis did not strad-
our proposed mediation model. Overall, these re- dle zero (i.e., [.004, .08]) for the following model:
sults suggest that leader humility influences team leader humility → collective humility → team
performance through a two-stage process of collec- demands–abilities fit → collective promotion focus →
tive humility and then collective promotion focus. team performance. Thus, teams expressing high
Moreover, this path from leadership to team perfor- levels of humble behavior tended to have a better fit
mance is distinct from the process linking trans-
formational leadership to team performance. 3
Average rWG 5 .86, ICC(1) 5 .10; ICC(2) 5 .24; a 5 .85.
between team member skills and task demands, whether imitable CEO behaviors shape performance-
which led to collective promotion focus and to enhancing interaction patterns and, in turn, strategic
higher team performance. foci, throughout the organization.
Our findings also inform the leadership literature
by showing that leaders who express humility may
Discussion
help their teams to transcend the comparative–
The purpose of Study 3 was to test our full con- competitive social lens that often leads to over-
ceptual model with an organizational field sample estimating oneself and underestimating others,
(as illustrated in Figure 1). The triangulation of our which is arguably a poor foundation for effective
three studies lends strong support to leader humility teamwork (see Anderson, Srivastava, Beer, Spataro,
being an important antecedent to team performance & Chatman, 2006; Lauber, Baetge, & Acomb, 1986).
and influencing team performance through the two- Our theoretical rationale and empirical findings also
stage process of collective humility and collective contribute to the leadership literature by under-
promotion focus. We also differentiated the impact scoring the importance of leading by example, and
of leader humility from transformational leadership may help explain why followers view leader hy-
by controlling for this construct and rerunning the pocrisy so harshly (Dufresne & Clair, 2013; Quinn,
model with transformational leadership as the in- 2004). Indeed, Dasborough, Ashkanasy, Tee, and Tse
dependent variable. We view the fact that trans- (2009) found that leader insincerity produced nega-
formational leadership did not predict collective tive follower emotions, such as cynicism toward the
humility or collective promotion focus as evidence leader. Hypocrisy, or incongruence between leader
for the differential impact of the specific, imitable behaviors and espoused values (Quinn, 2004), is
behaviors comprised in leader humility. thought to be a major weakness resulting from the
more visionary styles of leadership (Cha &
Edmondson, 2006). In fact, this finding of hypoc-
GENERAL DISCUSSION
risy flowing from charismatic leadership may have
The most important contribution our research partly motivated the growth of values-based leader-
makes is to help illuminate how one nontraditional ship, or leadership that is aligned with the “authen-
leadership approach—leader humility—influences tic, true self” (Gardner, Avolio, Luthans, May, &
team performance. Though leader humility is one of Walumbwa, 2005: 344; Sparrowe, 2005). Our theo-
many leadership approaches that have recently been retical rationale and empirical findings suggest that
found to foster positive effects on teams (e.g., Gelfand, followers are closely attuned to leaders’ behaviors,
Leslie, Keller, & de Dreu, 2012; Liden, Wayne, Liao, & and the specific behaviors that leaders model can
Meuser, 2014; Schaubroeck et al., 2011; Yaffe & Kark, profoundly influence team interaction processes and
2011), this effort sought to illuminate how it is performance.
uniquely situated to foster some of the core team ef- Like Schaubroeck and colleagues (2011), we ex-
fectiveness processes that teams theorists have iden- plored two types of leadership in our last study, and
tified. Specifically, we found that humble leader found that, while transformational leadership was
behaviors influenced group performance by fostering positively associated with team performance (see
the constructive interpersonal processes inherent in Table 6), its effect did not manifest through the same
collective humility and by catalyzing a specific col- path as leader humility. Though Schaubroeck and
lective regulatory focus. Post hoc analyses in Study 3 colleagues (2011) revealed that transformational
also confirmed that the theorized mechanism of task leadership influences team performance by fostering
allocation effectiveness also mediated the effect of cognition-based trust and team potency, leader
leader humility and team performance. Certainly, the humility’s influence was through contagion of the
model we proposed and tested could apply to many behaviors themselves, shaping specific teamwork
types of leader behaviors: leaders model a behavior, and regulatory-focus aspects of team functioning.
followers emulate it, a strategic orientation is created, Perhaps one reason why transformational leadership
and this strategic focus influences performance. Our did not predict performance as strongly in this con-
process model for how leaders influence group per- text is that the performance circumstance did not
formance may help provide insight into how leaders warrant a new compelling vision or was not one of
so greatly affect firm performance (e.g., CEOs account “extreme challenge, stress, and uncertainty,” which
for 38.5% of the variance in firm performance; are situations when transformational leadership is
Hambrick & Quigley, 2014), leading us to inquire theorized to be most important (Bass, 1985: 815). In
FIGURE 1
Study 3: Structural Equation Modeling Path Analysisa
e1 .27 e2 .35 e3 .20

1 1 1
.30* .67*** .47***

Collective Promotion
Leader Humility Collective Humility Team Performance
Focus
a
Path weights represent standardized path coefficients. Fit of the hypothesized mediation model (AIC 5 17.37) was superior to the rival
model in which direct paths were modeled between all study variables and team performance (AIC 5 55.15).
* p , .05, *** p , .001
contrast, qualitative evidence suggests that leader 2015) or compassion (Rynes, Bartunek, Dutton, &
humility is less effective in times marked by extreme Margolis, 2012), and whether these virtuous behav-
threat or time pressure (Owens & Hekman, 2012). iors enhance or undermine group performance.
Thus, leader humility may be more beneficial to team Lastly, our model also contributes to the vast reg-
effectiveness relative to transformational leadership ulatory focus literature (Beersma et al., 2013;
during everyday challenges (times of low to moderate Dimotakis et al., 2012; Higgins, 1997; Rietzschel,
amounts of challenge, stress, pressure, or threat). It 2011). Specifically, our findings advance regulatory
would be meaningful for future research to explore focus research by (a) identifying an important team
the relative benefits of both forms of leadership across outcome that results from collective promotion focus
different types of situations—specifically, whether and (b) identifying specific leader and team behav-
leader humility would be less effective when more iors that foster a collective promotion focus. Our
directive, power-centralized styles of leadership are field-based examination of collective promotion fo-
required. cus is novel, as, to our knowledge, this construct has
Our results also contribute to the burgeoning lit- been exclusively manipulated among laboratory-
erature examining humility, and contribute to the based groups rather than measured among naturally
debate about its worthiness as a virtue (Exline & occurring groups found in field settings (Beersma
Geyer, 2004; Grenberg, 2005; Owens, Rowatt, & et al., 2013; Dimotakis et al., 2012; Levine et al., 2000;
Wilkins, 2011). This effort represents additional ev- Rietzschel, 2011). Our findings also explain how this
idence for the strengths-based view of humility, that strategic orientation might emerge from team be-
it is a virtue foundational to progress and develop- haviors, such as collective humility, which is a novel
ment and has tangible value in organizational antecedent of a collective promotion focus. These
contexts. Far from being a sign of weak-willed, results build on findings that a collective promotion
stooped-shouldered meekness (see Tangney, 2000), focus can spring from group mottos (Faddegon et al.,
humility keeps individuals in a state of continual 2008) as well as team reward structures (Levine et al.,
adaptation. Our findings suggest that humility ap- 2000). As this research focused only on promotion
pears to embolden individuals to aspire to their regulatory focus, it would be meaningful for future
highest potential and enables them to make the in- research to explore the association of leader and
cremental improvements necessary to progress to- team humility and prevention regulatory focus. It
ward that potential. These studies are also the first may be that leader humility is negatively related to
empirical effort to confirm the contagious nature of team performance in situations that may call for
humility as leader-modeled humility fostered col- a prevention regulatory focus, such as high reliability
lective humility in teams. The implications are clear: contexts (Roberts, 1990). Future research should also
individuals must act virtuously if they want virtue to explore specific team behaviors spurred by a team
spread. Organizational members can develop virtues promotion focus. Although post hoc analysis of our
by practicing them, and virtue abilities, like physical Study 2 data suggested that resource investment
abilities, are subject to development or deterioration behavior (millions of dollars invested into increasing
(Vera & Rodriguez-Lopez, 2004; see also Dunning, factory capacity, new product launches, and re-
1995). Future research should examine whether our search and development) mediated the influence of
model holds for other virtues, such as courage promotion focus climate on team performance (i.e.,
(Koerner, 2014; Schilpzand, Hekman, & Mitchell, stock price), we recommend that future research
more directly theorize and test specific team behav- impact of leader humility on teams with different
iors fostered by collective promotion focus that en- types of interdependence (e.g., pooled, sequential, or
hance performance. reciprocal). Lastly, future research should explore
other potential antecedents of collective humility
besides immediate leader modeling, such as higher-
Strengths, Limitations, and Future Research
level or executive leader modeling, shared team
This series of studies has several strengths, in- leadership structures, and demographic or func-
cluding three samples used in experimental, simu- tional diversity.
lation, and longitudinal field designs. We also used
objective and subjective team performance mea-
CONCLUSION
sures, which further enhanced confidence in our
theoretical predictions (Campbell, Stanley, & Gage, Leadership is the most important contextual factor
1963). Thus, we took advantage of the strengths of the in shaping team performance (Hambrick & Quigley,
rigor and internal validity of laboratory contexts as 2014). Our theoretical rationale and empirical findings
well as the power and generality of field contexts. help extend and enrich theory showing that imitable
Although our studies in aggregate strongly support leader behaviors, like those encompassed in humility,
our model, they had individual limitations that are an important mechanism for influencing team
should be noted. In our studies, collective humility performance. Specifically, followers emulate leader
and team collective promotion focus were measured behaviors, which generate a shared team regulatory
from the same sources at the same time. Thus, the orientation that ultimately influences team perfor-
relationship between the variables may be inflated mance. This research also provides evidence toward
because of common method variance (CMV). How- the idea that historic virtues do still have relevance in
ever, as reported in the results section, the two con- predicting important, bottom-line outcomes. We hope
structs loaded onto separate factors, and regression that the theory and findings reported here spur further
analyses showed that multicollinearity was not theoretical and empirical attention toward a better
a concern in any of our studies. In addition, CMV understanding of the relevance and process of imitable
could not explain the relationship between our pre- leader behaviors within team contexts.
dictors and team performance, because team perfor-
mance in Studies 2 and 3 were assessed by a different
source. To further allay concern over CMV, we rec- REFERENCES
ommend future research replicate our findings by Abdel-Halim, A. A. 1981. Effects of role stress–job
examining different operationalizations of collective design–technology interaction on employee work
humility and team collective promotion focus, such satisfaction. Academy of Management Journal, 24:
as behavioral or extra-team observer assessments. Our 260–273.
studies are also limited in that all were conducted Anderson, C., Srivastava, S., Beer, J. S., Spataro, S. E., &
with Western samples and thus their generalizability Chatman, J. A. 2006. Knowing your place: Self-
to Eastern contexts is unclear. We recommend future perceptions of status in face-to-face groups. Jour-
research to try to replicate our findings using samples nal of Personality and Social Psychology, 91:
from Eastern contexts. 1094–1110.
Though it appears that leader and collective hu- Anderson, J. C., & Gerbing, D. W. 1991. Predicting the
mility fosters team performance effectiveness in performance of measures in a confirmatory factor
a simulated auto industry and in a health services analysis with a pretest assessment of their substantive
context, future research should also explore various validities. The Journal of Applied Psychology, 76:
team contexts with varying levels of task speciali- 732–740.
zation, hierarchical adherence, collective humility Arbuckle, J. 2010. IBM SPSS Amos 19 user’s guide. Chi-
dispersion, and time pressure. For example, leader cago, IL: SPSS Inc.
and collective humility may have different effects Ashforth, B. E., & Mael, F. 1989. Social identity theory and
within a Navy Seals squad, Marine infantry platoon, the organization. Academy of Management Review,
basketball team, or heart surgery unit. We can also 14: 20–39.
see the benefits of exploring the level of dis- Basford, T. E., Offermann, L. R., & Behrend, T. S. 2014.
junctivneness (vs. conjunctiveness) of team tasks Please accept my sincerest apologies: Examining fol-
(Steiner, 1972) to understand boundary conditions lower reactions to leader apology. Journal of Business
for the effectiveness of collective humility, and the Ethics, 119: 99–117.

Humility

Uploaded by

Copyright:

Available Formats

Humility

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Humility

Uploaded by

Copyright:

Available Formats

r Academy of Management Journal

2016, Vol. 59, No. 3, 1088–1111.

HOW DOES LEADER HUMILITY INFLUENCE TEAM

Measures promotion focus was operationalized as a referent-

1. Collective promotion focus 3.56 .88 ;

Team performance. Team performance was Results

1. T2 stock priceb 54.00 44.68 ;

Discussion humility and team performance. We included team

1. Team performance 3.64 .55 ;

e1 .27 e2 .35 e3 .20

.30* .67* .47*

You might also like