The Shared Leadership of Teams A Meta Analysis of Proxim - 2014 - The Leadershi PDF

The Leadership Quarterly 25 (2014) 923–942
Contents lists available at ScienceDirect
The Leadership Quarterly

journal homepage: www.elsevier.com/locate/leaqua
The shared leadership of teams: A meta-analysis of proximal,

distal, and moderating relationships☆
Vias C. Nicolaides ⁎,1, Kate A. LaPort, Tiffani R. Chen 2, Alan J. Tomassetti 2, Eric J. Weis 2,
Stephen J. Zaccaro, Jose M. Cortina
George Mason University, Department of Psychology, David King Hall, Room 2086, 4400 University Drive, MSN 3F5 Fairfax, VA 222030-4422, USA
a r t i c l e i n f o a b s t r a c t
Article history: The current meta-analysis examines the relationship between shared leadership and team perfor-
Received 18 August 2012 mance. It also assesses the role of team confidence (i.e., collective efficacy and team potency) in
Received in revised form 25 May 2014 this relationship. Mediation analyses supported the hypothesis that team confidence partially medi-
Accepted 18 June 2014
ates the effects of shared leadership on team performance. We also found support for the notion that
Available online 22 July 2014
shared leadership explains unique variance in team performance, over and above that of vertical
leadership. Furthermore, a variety of substantive continuous and categorical variables were investi-
Handling Editor: Shelly Dionne
gated as moderators of the shared leadership–team performance relationship. Specifically, the rela-
tionship between shared leadership and team performance was moderated by task
Keywords: interdependence, team tenure, and whether performance was objectively versus subjectively mea-
Shared leadership
sured. Finally, results suggest that the approach used when measuring shared leadership can also
Team performance
Meta-analysis
play a role in the observed validity. Theoretical and practical implications of these findings are
Moderators discussed.
Mediators © 2014 Elsevier Inc. All rights reserved.
Introduction
Organizations have entered an era of information technology and globalization, characterized by dynamic, complex, and competitive
environments (Barkema, Baum, & Mannix, 2002; Sirmon, Hitt, & Ireland, 2007). In order to effectively navigate such environments, or-
ganizations have turned to the implementation of team-based structures (McGrath, Arrow, & Berdahl, 2000; Salas, Goodwin, & Burke,
2009; Sundstrom, McIntyre, Halfhill, & Richards, 2000). Partly underlying this propagation of teams is the evidence that they provide
faster and more flexible action, as well as increased informational processing capability than more rigid and centralized organizational
structures (Hinsz, Tindale, & Vollrath, 1997; Richardson, Vandenberg, Blum, & Roman, 2010). This tendency toward more team-based
structures has caused scholars to focus on the identification and investigation of factors that contribute to overall team effectiveness
(e.g., Hackman, 1987; Kozlowski & Ilgen, 2006; Mathieu, Maynard, Rapp, & Gilson, 2008). One factor that has been frequently identified
in the literature as being important to team success is the leadership within and of teams (Barling, Christie, & Hoption, 2011; Kozlowski,
Watola, Jensen, Kim, & Botero, 2009; Zaccaro, Heinen, & Shuffler, 2009; Zaccaro, Rittman, & Marks, 2001).
Research on leadership in teams includes two different, yet complementary, streams of research (Avolio, Walumbwa, & Weber, 2009).
The first of these had involved the application of traditional theories of leadership, which places the emphasis on a single individual that is
designated to lead the team, and on the relationships that individual leader has with his/her followers (Dansereau, Graen, & Haga, 1975;
☆ Portions of this paper were presented at the 28th Annual Conference of the Society for Industrial and Organizational Psychology, Houston, TX, 2013.
⁎ Corresponding author at: Department of Psychology, George Mason University, Fairfax, VA 22030, USA.
E-mail address: vnicolai@gmu.edu (V.C. Nicolaides).
1
The first author would like to dedicate this paper to his late grandfather, Nikos Sokratous, who passed away while the author was working on this manuscript. He
taught us how to share and he will be immensely missed.
2
The 3rd, 4th, and 5th authors contributed equally to this project, and thus the order of authorship reflects alphabetical order.
http://dx.doi.org/10.1016/j.leaqua.2014.06.006
1048-9843/© 2014 Elsevier Inc. All rights reserved.
924 V.C. Nicolaides et al. / The Leadership Quarterly 25 (2014) 923–942
Kozlowski, Gully, Salas, & Cannon-Bowers, 1996; McGrath, 1962). This approach, referred to as vertical leadership (e.g., Pearce & Sims,
2002; Pearce & Conger, 2003), has focused primarily on the behaviors and processes that such individuals use to promote team effective-
ness (Burke et al., 2006; Morgeson, DeRue, & Karam, 2010; Zaccaro et al., 2009). For example, leaders can help teams by acting as coaches
(Hackman & Wageman, 2005; Morgeson, 2005), modeling or displaying affect (Kaplan, Cortina, Ruark, LaPort, & Nicolaides, 2014; Pirola-
Merlo, Hartel, Mann, & Hirst, 2002), and by the managing team boundaries (Druskat & Wheeler, 2003).
Over the last two decades, a second approach has gained traction. In this approach, leadership is seen as emanating not only from a
designated leader, but also from team members themselves (Carson, Tesluk, & Marrone, 2007; Avolio, Jung, Murry, & Sivasbramaniam,
1996; Pearce, 2004; Pearce & Sims, 2000; Yukl, 2007). While this approach has burgeoned in the past decade, its core ideas can be traced
to earlier writings (see Follett, 1924; Gibb, 1954). In its contemporary form, this perspective has become known as shared leadership.3
According to Pearce and Conger (2003), this type of leadership entails “a dynamic, interactive influence process among individuals in
groups for which the objective is to lead one another to the achievement of group or organizational goals or both. This influence process
often involves peer, or lateral, influence, and at other times involves upward or downward hierarchical influence” (p. 1). Shared leader-
ship is a characteristic of teams that emerges when leadership behaviors are performed by multiple members of the team (Day, Gronn, &
Salas, 2006; Pearce & Conger, 2003) in a concertive and conjoint manner (Gronn, 2002). Under such a conceptualization of leadership
there is a reduced distinction between leader and follower, because team members may fill either of these roles at any given time.
In their recent review of leadership in teams, Morgeson et al. (2010) brought a functionalist perspective to the study of leadership
(e.g., Hackman & Walton, 1986; McGrath, 1962), and asserted that the role of team leadership is to satisfy needs that arise during tran-
sition and action phases of team performance episodes (see Marks, Mathieu, & Zaccaro, 2001). Moreover, they identified four distinct
sources of team leadership that can perform 15 specific team leadership functions. According to Morgeson et al. (2010), these sources
can be distinguished in a 2 × 2 matrix on the basis of leadership formality (formal versus informal) and locus (internal versus exter-
nal). They conceptualized shared leadership as leadership influence stemming informally from internal team members. These are
members of the team's core that do not have formally prescribed leadership roles. Combining Pearce and Conger (2003) and
Morgeson et al. (2010), we define shared leadership as a set of interactive influence processes in which team leadership functions
are voluntarily shared among internal team members in the pursuit of team goals.
Despite the increased attention on shared leadership, there are a number of questions that remain unanswered. First, although
some studies have reported an overall positive relationship between shared leadership and team performance (see for example stud-
ies by Carson et al., 2007; Pearce & Sims, 2002), others have not supported this prediction (see for example studies by Boies, Lvina, &
Martens, 2010; Neubert, 1999). Second, and more importantly, although there is theory that proposes boundary conditions and me-
diating mechanisms of this positive relationship (e.g., Pearce & Conger, 2003) the vast majority of empirical studies have not exam-
ined such relationships. That is, there are few primary studies on mediators and moderators of shared leadership's relationship to
team performance. In short, given that the shared leadership literature has surged in the past few years, an empirical synthesis of
the literature is appropriate. It is also important to note that meta-analysis methods can solve problems contained in primary studies
(see Schmidt, 1992) and can organize, frame, and provide a roadmap for the future (Humphrey, 2011).
Relying on prevailing theoretical frameworks (e.g., Ilgen, Hollenbeck, Johnson, & Jundt, 2005; Marks et al., 2001), the current paper
contributes to the literature in several ways. First, we meta-analytically examine the relationship between shared leadership and team
performance. In doing so, we investigate the contribution by shared leadership to the prediction of team performance over and above
vertical leadership. Second, we answer calls by examining a priori continuous and categorical moderators of this relationship (e.g.,
Barry, 1991; Conger & Pearce, 2003). Third, we examine team confidence as a mediating mechanism through which shared leadership
operates to influence team performance. Our meta-analysis follows another recent one (Wang, Waldman, & Zhang, 2014) that exam-
ined (a) the relationship between shared leadership and team effectiveness, (b) the incremental validity of shared leadership over
vertical leadership, and (c) several of the moderators we included here. However, our study goes further by also analyzing team con-
fidence as a mediator of the effects of shared leadership on team performance, as well as by exploring the roles of team tenure, team
size, and whether the team task requires higher levels of behavioral interdependence, information exchange, or both as moderators of
this relationship. The examination of a mediator such as team confidence can provide much needed information on the mechanisms
through which shared leadership fosters better team performance. In the sections that follow we elaborate on our conceptualization
of the emerging process of shared leadership and offer theoretical justifications of its relationship with various team level constructs.
Theory and hypothesis development
The added value of shared leadership to team performance
Shared leadership is a phenomenon that emerges within teams across time. The notion that shared leadership develops within a
team through a series of successful team member interactions has been echoed by many (e.g., Barry, 1991; Perry, Pearce, & Sims,
1999). For example, Carson et al. (2007) found that a high-quality internal team environment containing shared purpose, social sup-
port, and voice was a critical antecedent of shared leadership. Using the IMOI model (Input–Mediator–Output–Input; Ilgen et al.,
2005), the current study conceptualizes shared leadership as an input of other team emergent states (Marks et al., 2001) and out-
comes, namely team confidence and team performance. Indeed, Day et al. (2006) suggested that shared leadership can serve as a
3
Other commonly used terms are distributed, collective, and rotated leadership.
V.C. Nicolaides et al. / The Leadership Quarterly 25 (2014) 923–942 925
driver of future team development and performance. Similarly, a study by Hmieleski, Cole, and Baron (2012) found that shared lead-
ership influenced performance indirectly through the development of a team's positive affective tone.
The increased attention to shared leadership is rooted in the proposition that such a phenomenon contributes to team functioning
and effectiveness, partly due to greater participation, information sharing, and positive tone among team members (Mehra, Smith,
Dixon, & Robertson, 2006; Mohrman, Cohen, & Mohrman, 1995; Pearce & Conger, 2003). For example, in a recent study of consulting
teams, Carson et al. (2007) distinguished between shared leadership and other similar constructs (e.g., team empowerment) and
found that shared leadership predicted client ratings of team effectiveness. However, evidence for shared leadership's positive effect
on performance is mixed, because some studies have not found support for this prediction (e.g., Boies et al., 2010; Neubert, 1999).
Thus, a meta-analytic approach is necessary to examine the extent to which the display of shared leadership is beneficial for team per-
formance, and to derive a more accurate estimate of its magnitude of influence.
The issue of what shared leadership provides more than vertical leadership to the team is also important to consider. Researchers
have suggested that in many team contexts, vertical leaders may lack the full range of human capital or the temporal resources nec-
essary to help their teams accomplish their goals; accordingly, in such contexts, shared leadership processes can provide support to
vertical leadership efforts (Barry, 1991; Cohen & Bailey, 1997; Kerr & Jermier, 1978; Perry et al., 1999) and, thus, explain significant
variance in team performance, beyond vertical leadership. Based on these arguments we offer the following hypotheses:
Hypothesis 1a. Shared leadership explains significant variance in team performance.
Hypothesis 1b. Shared leadership contributes incremental variance in team performance, beyond vertical leadership.
Shared leadership, team confidence, and team performance
Ilgen et al. (2005) proposed the IMOI model as a useful approach for understanding a variety of group-level processes. Specifically,
these scholars argued that team inputs influence team outputs through mediating team processes and emergent states. Moreover,
they argued that outputs may in turn become inputs to later team processes. In the current paper, we examined shared leadership
as an input with effects on team performance (i.e., an output) that occur through team confidence (see Mathieu et al., 2008),
which is an umbrella term for the emergent states of collective efficacy and potency (e.g., Bandura, 1997, 1998). In doing so, we are
answering multiple calls for investigation into the psychological and team-based mechanisms that explain how shared leadership
is related to performance (Carson et al., 2007; Avolio et al., 2009; Conger & Pearce, 2003). Collective efficacy has been defined as “a
shared belief in a group's collective capability to organize and execute courses of action required to produce given levels of goal attain-
ment” (Kozlowski & Ilgen, 2006, p. 90). Potency also refers to collective beliefs regarding the team's ability, but it pertains to a broader
range of goals, tasks and activities than collective efficacy (Shea & Guzzo, 1987). Although conceptually distinct, various meta-analyses
have demonstrated the high inter-relatedness of these two constructs and their beneficial outcomes on team performance (Gully,
Incalcaterra, Joshi, & Beaubien, 2002; Stajkovic, Lee, & Nyberg, 2009).
Furthermore, past research has supported the notion that team confidence mediates the effects on team performance of leadership
behaviors displayed by formal leaders (e.g., Bass, Avolio, Jung, & Berson, 2003; Lester, Meglino, & Korsgaard, 2002). Similarly, engage-
ment in shared leadership behaviors and satisfaction of team needs (e.g., setting realistic goals for one another and helping the team
generate solutions to overcome obstacles) should increase the confidence team members have that collectively they can produce ef-
ficient team functioning that leads to team success. As members share team leadership functions, they gain more voice in team direc-
tion and in the management of team processes, which in turn should foster a shared commitment to team action. The result should be
a stronger and more positive sense of the team's ability to succeed. Indeed, in an early discussion of shared leadership and team con-
fidence, Bradford and Cohen (1984) argued that having members who act in ways that compliment each other's competencies adds to
the overall efficacy of the team. As a consequence of shared leadership, team confidence in turn propels the group to expend effort and
achieve high performance (Sivasubramaniam, Murry, Avolio, & Jung, 2002; Gully et al., 2002; Stajkovic et al., 2009). In short, the afore-
mentioned discussions suggest both a direct and an indirect effect of shared leadership on team performance. Given that there are
likely to be multiple transmitters that convey such effects, we anticipate partial mediation through team confidence.
Hypothesis 2. Team confidence partially mediates the effects of shared leadership on team performance.
In the following section we offer a priori moderators of the relationship between shared leadership and team performance, as sug-
gested by previous authors (e.g., Barry, 1991; Carson et al., 2007; Conger & Pearce, 2003; Day et al., 2004; Perry et al., 1999).
Moderators of the effects of shared leadership on team performance
In his review of moderator practices in meta-analysis, Cortina (2003) stressed the importance of stating a priori moderators. In line
with his recommendation, we examined a variety of continuous and categorical moderators of the shared leadership–team perfor-
mance relationship (Barry, 1991; Conger & Pearce, 2003; Perry et al., 1999). Indeed, meta-analysis gives us the opportunity to test
for moderators (and mediators) even if they have not been tested in primary studies. They can be grouped into two overall categories:
substantive and methodological. Regarding substantive moderators, we examined variability in the relationship between shared lead-
ership and team performance as a function of (a) task interdependence and (b) team performance index (i.e., subjective versus ob-
jective). We also tested in an exploratory fashion the possibility that (a) team size, (b) mean team tenure, (c) team type, and
(d) shared leadership measurement approach moderate the shared leadership–team performance relationship. As is common and ap-
propriate in meta-analysis, we examined the methodological moderators of (a) criterion source (same versus other), (b) study setting
(school versus field), and (c) publication type (peer reviewed versus non-peer reviewed). In the next section we offer our theoretical
rationale for all substantive moderators and briefly discuss the exploratory moderatos.
Substantive moderators of the shared leadership–team performance relationship
Task interdependence
Kozlowski and Ilgen (2006) have argued that research that fails to incorporate interdependence is of limited value for building theory
and knowledge about teams. Although a variety of interdependence types have been proposed (e.g., Saavedra, Earley, & Van Dyne, 1993;
Shea & Guzzo, 1987; Thompson, 1967; Wageman, 1995), our conceptualization of task interdependence takes into account input, pro-
cess, and outcome interdependencies within the team (e.g., Wageman, 1995). In short, we expect a stronger relationship between shared
leadership and team performance when interdependence is high. This is because interdependence necessitates higher levels of distrib-
uted expertise, coordination, and a greater need for interaction and guidance (Alper, Tjosvold, & Law, 1998; Guzzo & Shea, 1992;
Wageman, 1995), which should magnify the effects of sharing leadership on team performance. That is, task interdependence increases
demands for leadership behaviors that foster more effective member coordination (Zaccaro et al., 2009). In a recent qualitative study,
Weibler and Rohn-Endres (2010) found that coordination and embeddedness in a high-quality relational environment were critical
for the emergence and influence of shared leadership. Moreover, several studies have reported that shared leadership appears to enhance
performance in interdependent contexts, such as consulting groups (Carson et al., 2007), top management teams (Ensley, Hmieleski, &
Pearce, 2006), and medical teams (Klein, Ziegert, Knight, & Xiao, 2006). In addition, such a rationale is consistent with past researchers
who have found higher validities between leadership behaviors in general and team performance when interdependence was high
(e.g., Burke et al., 2006).
Hypothesis 3. Task interdependence moderates the relationship between shared leadership and team performance, such that shared
leadership will be more strongly related to team performance under conditions of higher versus lower levels of interdependence.
Team performance index

Campbell and colleagues (Campbell, 1990; Campbell, McCloy, Oppler, & Sager, 1993) called for researchers to more carefully distin-
guish between different types of performance measures. In the current paper, we distinguish between subjective and objective indices of
task performance. We further point out that objective indices of performance are confounded with potential impediments that are out-
side the control of individuals. Such contamination may attenuate shared leadership's correlation with team performance. For example,
group sales may be determined by economic conditions, which are not under the control of team members. Therefore, we hypothesize
the following:
Hypothesis 4. Team performance index will moderate the relationship between shared leadership and team performance, such that
stronger relationships are observed when performance is assessed in terms of subjective indices than in terms of objective indices.
Exploratory moderators of the shared leadership–team performance relationship
In addition to offering and testing specific hypotheses concerning the aforementioned moderators, we also examine relationships
between shared leadership and certain exploratory moderators. Because the manner in which these exploratory variables interact
with shared leadership is unclear we do not propose specific hypotheses, but rather offer research questions.
Team size
Team size has a long and venerable past in the organizational behavior literature and has been found to impact group social phe-
nomena (Hare, 1981; Kozlowski & Bell, 2003; Latané, 1981; Triplett, 1898). In the domain of shared leadership, team size has been
largely treated as a nuisance variable to be controlled for statistically (e.g., Pearce & Sims, 2002) despite calls for examinations of its
effects in the context of shared leadership (Conger & Pearce, 2003; Cox, Pearce, & Perry, 2003; Pearce, 2004; Perry et al., 1999;
Seers, Keller, & Wilkerson, 2003).
Interestingly, past research has indicated that team size can be both an asset and a liability for teams (e.g., Maier, 1967). On the one
hand, larger teams have greater decision making and information processing capabilities than smaller teams (Hill, 1982; Maier, 1967). On
the other hand, an increase in the number of team members also introduces proximity barriers and reduces coordination and commu-
nication effectiveness (Campion, Medsker, & Higgs, 1993; Shaw, 1981; Smith et al., 1994), which can hinder mutual influence processes.
In short, past research implies a variety of ways in which team size may moderate the relationship between shared leadership–team
performance.
Research Question 1. Does team size moderate the shared leadership–team performance relationship?
Team tenure
Team tenure has been identified as a demographic variable that has a significant influence on team dynamics (e.g., Barker &
Patterson, 1996; Finkelstein & Hambrick, 1990). However, in the context of shared leadership it has received scant theoretical and
empirical attention. Therefore, we sought out to test in an exploratory fashion the moderating role of team tenure in the shared lead-
ership–team performance relationship.
On the one hand, team tenure may conceivably facilitate shared leadership effectiveness if tenure is viewed as a proxy for team
experience (Tesluk & Jacobs, 1998). On the other hand, sharing leadership over an extended period of time may be detrimental to
team performance. A substantial body of literature has emphasized that leadership constitutes power (e.g., Bass & Bass, 2008;
Pfeffer, 1981, 1997; Yukl, 2007). Most importantly, and specific to shared leadership, the sharing of power can be hard to control
and balance over time (Morrill, 1995; Yukl, 2007). Consequently, longer tenured teams may suffer from power struggles and
power inequalities, which breed tension, conflict, and anger within the group (Coleman, 2000; Sell, Lovaglia, Mannix, Samuelson,
& Wilson, 2004), thus disrupting team processes and performance (Dewett, 2004). In short, theory presents competing arguments
on the moderating role of team tenure in the shared leadership–team performance relationship. Given that team tenure has not
been examined in shared leadership research we decided to explore if, and how, team tenure interacts with shared leadership in
the prediction of team performance.
Research Question 2. Does team tenure moderate the shared leadership–team performance relationship?
Team type
Team type has frequently been found to moderate team relationships in past meta-analyses (e.g., Chiocchio & Essiembre, 2009;
DeChurch & Mesmer-Magnus, 2010). In the current meta-analysis, we distinguish among decision-making, action, and project
teams, and examine whether different types of teams have different shared leadership and team performance relationships. Different
types of teams are characterized by varying levels of behavioral and informational interdependence. For example, decision-making
teams typically have higher informational interdependence, action teams have higher behavioral interdependence, and project
teams have higher levels of both (DeChurch & Mesmer-Magnus, 2010). In turn, shared leadership may plausibly satisfy the needs
of one type of team better than another. Given the lack of theory and research in this area we did not offer formal hypotheses, but
rather posed a research question.
Research Question 3. Does team type moderate the shared leadership–team performance relationship?
Shared leadership measurement

Proper focal construct measurement is essential. Given that researchers have employed a variety of ways to measure shared leader-
ship in teams (Gockel & Werth, 2010), we examine in an exploratory manner the possibility that the way one measures shared leadership
moderates the shared leadership–team performance relationship. Typically, one of three approaches has been followed. The first ap-
proach entails administering leadership behavior questionnaires and aggregating using a direct-consensus or referent-shift consensus
model (Chan, 1998) where the referent can be either the individual members or the team respectively (Pearce & Sims, 2002; Avolio
et al., 1996). For example, Pearce and Sims (2002) had participants respond twice to individual leadership behaviors (e.g., transforma-
tional, transactive, etc.); once referencing their team members as a whole (shared leadership) and once referring only to their team leader
(vertical leadership). Similarly, a second approach uses a direct-consensus or referent-shift consensus model that relies on a functional
leadership paradigm and measures shared behaviors, such as boundary spanning (Morgeson et al., 2010; see also Hoch & Kozlowski,
2012). The third way to measure shared leadership takes a social network approach in which team members nominate/rate other
team members according to the degree to which they exercised leader-like influence (Carson et al., 2007; Mayo, Meindl, & Pastor,
2003). Network measures can then be computed to answer different types of shared leadership questions (e.g., network density).4
Given that there are advantages and disadvantages associated with the use of these approaches (see Gockel & Werth, 2010), they may
plausibly yield differential shared leadership validities. Thus, we investigated if there are implications of using these different approaches.
Research Question 4. Does the way in which shared leadership is measured moderate the shared leadership–team performance
relationship?
Method
Literature search
We used several means to obtain both published and unpublished empirical studies. First, we located relevant journal articles by
conducting a search in the PsycInfo and ABI-Inform databases from January, 1990–April, 2013, using the following keywords: shared
leadership, distributed leadership, team leadership, rotated leadership, team empowerment, collective leadership, top management teams,
self-managed teams, and team leadership functions (see Morgeson et al., 2010). Next, we searched the conference programs from the
annual meeting of the Society of Industrial and Organizational Psychology (SIOP) (1998 to 2013) and the national meetings of the
Academy of Management (AoM) (2000 to 2011) with the same keywords. Then, we solicited additional unpublished empirical studies
4
A related index, the coefficient of variation can also be used to assess influence concentration within a team (e.g., Haleblian & Finkelstein, 1993). Furthermore, the
use of an actor–partner interdependence model (Kenny, Kashy, & Cook, 2006) has been suggested as a third potential approach to the study of shared leadership
(Gockel & Werth, 2010).
via the Interdisciplinary Network for Group Research (INGRoup), SIOP, and AoM Organizational Behavior listservs. We also reviewed
the reference sections of seminal shared leadership articles. Finally, we contacted a number of experts who have published articles on
shared leadership. We asked them to share any in press articles or working papers that they may have.5 The search yielded a total of
467 studies whose abstracts were deemed suitable for inclusion.
Criteria for inclusion
Errors in meta-analyses are often made due to the plethora of judgments that need to be made throughout the research synthesis
process (Wanous, Sullivan, & Malinak, 1989). Therefore, our initial step was to formulate a clear and thorough operational definition
of shared leadership. We included studies that had looked at the degree to which specific team leadership functions (see Morgeson
et al., 2010) and leadership behaviors (see Pearce & Sims, 2002) were performed informally by internal team members and measured
through a direct-consensus, referent shift consensus model (Chan, 1998), a social network approach, or the actor–partner interdepen-
dence model (see Avolio, Sivasubramaniam, Murry, Jung, & Garger, 2003; Gockel & Werth, 2010).
Items used to measure shared leadership were reviewed by the authors. For inclusion in the meta-analysis, a study's items needed to
describe specific behaviors that the team as a whole or each individual performed that satisfied team needs as described by Morgeson
et al. (2010). In doing so, we relied heavily on the leadership functions framework laid out by Morgeson et al. (2010). For example,
one such team leadership function is boundary spanning (Ancona & Caldwell, 1992; Morgeson et al., 2010). Thus, we included in our
meta-analysis a study by Marrone, Tesluk, and Carson (2007) because all team members had rated the degree to which each of their
teammates engaged in boundary spanning behaviors and the behaviors were aggregated to the team level. Similarly, we included a
study by Tesluk and Mathieu (1999) which asked teams to indicate the extent to which their team as a whole engaged in team prob-
lem–management actions and strategies. This is because taking actions to solve problems is identified as a team leadership function in
Morgeson et al.'s (2010) framework. Finally, in a few instances we included studies that had measured team self-managing behaviors,
provided that they were at the team level, used an aggregated or SNA approach, and entailed behavioral items taping the fulfillment
of team needs as listed in Morgeson et al. (2010). For example, we included a study by Rousseau and Aubé (2010) that used a referent
shift approach to measure the degree to which team members engaged in planning, monitoring, reinforcing, and adjusting behaviors.
Second, given that shared leadership is a team level phenomenon we did not retain studies that had used an individual level of
analysis (e.g., Avolio et al., 1996). In sum, the decision making process involved gathering the leadership measures used by articles
in the literature search results and distributing them to the first five authors. Of the 467 initial studies 52 were judged by the first
five authors to comply with all of the above criteria. Table 1 presents the effect sizes between shared leadership and team perfor-
mance. Furthermore, we present values coded for the substantive and exploratory moderators. All included studies can be found in
the reference section, marked by an asterisk.
Coding of studies
For each sample, in addition to any effect sizes of interest, the following information was recorded: (a) means, standard deviations,
and reliabilities of all variables of interest, (b) mean team size (c) type of shared leadership measure and style, (d) interdependence
(i.e., low, moderate, high), (e) team type (i.e., action, decision making, project), (f) type of team performance index (subjective versus
objective), (g) mean team tenure, (h) study setting (i.e., field versus school), (i) criterion report source (i.e., same versus other source),
and (j) publication type (peer review versus non peer review).
A few study judgments regarding coding deserve further elaboration and attention. First, when coding team interdependence we
followed approaches used by past scholars (e.g., DeChurch & Mesmer-Magnus, 2010; Gully, Devine, & Whitney, 1995; Gully et al.,
2002). The low interdependence category included studies in which task performance derived almost solely from individual efforts
(i.e., task requires that one team member generates the correct solution); team members did not require others to help reach task
goals; and/or, feedback, rewards and other outcomes were offered at the individual level (DeChurch & Mesmer-Magnus, 2010).
For example, sales teams whose members met monthly and whose rewards were based on individual performance were considered
to have low interdependence (e.g., Greer, De Hoogh, Patel, Thatcher, & De Dreu, 2011, study 1). In the moderate category we included
studies in which team members required some level of resources and support from other members, but task performance still
entailed considerable sole individual effort. Also, performance rewards and outcomes were offered to both individuals and to the
group as a whole (DeChurch & Mesmer-Magnus, 2010). For example, road teams whose team members did not work together on
all tasks, but whose tasks included a mix of sequential, reciprocal, and pooled coordination of resources were considered to have mod-
erate interdependence (e.g., Hiller, Day, & Vance, 2006). In the high category we included studies in which team performance derived
from high mutual dependency among members, members required high levels of resources and support from others, and only team
level rewards and outcomes were possible (e.g., Carson et al., 2007) (DeChurch & Mesmer-Magnus, 2010). In addition, top manage-
ment teams were considered to have high interdependence. Finally, we also took into account means and standard deviations when
researchers had directly measured interdependence. For example, a study by Stewart and Barrick (2000) had measured interdepen-
dence on a 7 point Likert-type scale. Its mean was 3.56 and its standard deviation was .41. This implies that 95% of the individuals
rated interdependence on a range of 2.74–4.38, which we considered as having low interdependence. Inter-rater agreement was
93.2%.
5
We are grateful and wish to thank the scholars who shared their unpublished studies with us.
Second, team tenure was coded in weeks at the time of performance measurement for both field and student teams. For student
teams, we coded tenure as 15 weeks whenever team members would work together for one semester unless specified by the author.
Third, we distinguished among objective (e.g., team sales, simulation scores) and subjective (e.g., supervisor ratings of team perfor-
mance) indices of team performance. Whenever performance was rated by two sources (i.e., same and other) we coded the correla-
tion that used the rating from the observer (i.e., other) so as to minimize common method variance. Fourth, due to the small number
Table 1
Overview of effect sizes and study characteristics.
Study (year) r N rxx ryy Task Performance Team Average Team SL measurement
interdependence index tenure team size type approach
Berdahl and Anderson (2005) Study 2 0.41 41 – – 2 0 10 3.95 3 4

Boies et al. (2010) −0.08 49 – 1 2 1 15 3.96 1 1
Carson and Tesluk (2007) Study 2 0.38 29 – 0.96 3 0 20 4.83 3 4
Carson et al. (2007) 0.46 56 0.78 0.93 3 0 20 5.93 3 4
Carte, Chidambaram, and Becker (2006) 0.39 22 0.94 – 2 0 15 4.77 3 4
Cashman (2009) 0.66 33 – – 3 0 – 4.33 – 1
Charlier (2012) 0.03 86 – – 2 1 1 4 1 4
Chen, Wolf, Zaccaro, and Odenheimer (2013) 0.25 199 – 1 2 1 12 4.13 3 2
Chen and Lee (2007) 0.001 31 – 0.95 2 0 15 – 3 1
Colbert, Barrick, and Bradley (2014) 0.09 94 0.52 1 3 1 369 6.40 1 1
Daspit, Tillman, Boyd, and Mckee (2013) 0.82 24 0.30 0.58 3 3 15 5.92 3 2
Delia (2011) 0.10 59 – 0.86 – 0 52 9.83 3 2
Ensley et al. (2006) Sample 1 0.27 66 – 1 3 1 – 2.55 1 1
Ensley et al. (2006) Sample 2 0.20 154 – 1 3 1 – 2.71 1 1
Foo, Sin, and Yiong (2006) 0.21 51 – 1 3 1 – 3.39 1 4
Greer et al. (2011) Study 1 0.03 184 0.3 1 1 1 – 5.59 2 3
Greer et al. (2011) Study 2 −0.20 53 0.77 0.73 3 0 440 9.24 3 3
Greer et al. (2011) Study 3 −0.26 48 0.84 0.78 2 0 350 10.25 – 3
Gupta, Huang, and Niranjan (2010) 0.16 28 – 1 2 1 15 4.18 1 1
Gupta, Huang, and Yayla (2011) 0.28 36 – 1 2 1 14 4.06 1 1
Hiller et al. (2006) 0.36 45 – – 2 0 – 5.50 2 2
Hmieleski et al. (2012) −0.12 179 – 1 3 1 – 4.80 1 3
Hoch (2010) Study 3 0.50 61 – 0.85 – 0 59 4.10 – 3
Hoch and Kozlowski (2012) 0.13 101 – 0.79 3 0 217 5.36 3 2
Hoch, Pearce, and Welzel (2010) 0.23 26 – 0.82 2 0 74 – 3 3
Ishikawa (2012) 0.42 119 0.71 0.89 3 0 97 5.50 3 4
Lin, Yang, Arya, Huang, and Li (2005) 0.31 45 – – 2 0 15 6.50 3 4
Marrone et al. (2007) 0.47 30 0.71 0.84 3 0 15 6.13 3 2
McIntyre and Foti (2013) 0.28 39 – 1 3 1 12 3 1 4
Mehra et al. (2006) 0.37 28 – 1 2 1 – 13.4 2 4
Mehta, Feild, Armenakis, and Mehta (2009) 0.34 91 0.98 1 3 1 15 5.07 1 2
Mendez (2010) 0.54 26 0.45 0.90 3 0 177 9.77 1 4
Muethel, Gehrlein, and Hoegl (2012) 0.18 96 0.82 0.87 3 0 – 8.27 3 2
Neubert (1999) 0.14 21 – 0.83 2 0 – 13 2 4
Pearce and Ensley (2004) 0.42 69 – 0.98 3 0 80 7.24 3 3
Pearce and Sims (2002) 0.32 71 – 0.98 3 0 61 7.24 3 1
Pearce, Yoo, and Alavi (2004) 0.18 28 – – 2 0 – 7.39 3 1
Potter, Balthazard, and Waldman (2010) 0.29 118 – – 2 0 15 3.86 1 1
Ramthun, McElravy, and Matkin (2013) 0.63 51 – – 3 0 – 4 2 4
Rasmussen (2011) 0.50 53 0.98 0.96 3 0 34 5.64 3 3
Rousseau and Aubé (2010) 0.26 97 – 0.88 2 0 22 4.34 2 2
Sanders (2006) 0.75 49 – 0.91 3 0 79 12.65 3 3
Sivasubramaniam et al. (2002) 0.45 41 – 0.95 2 0 15 3.75 3 1
Small and Rentsch (2010) 0.23 58 – – 3 Both 15 4.67 1 4
Smith, Houghton, Hood, and Ryman (2006) 0.31 51 0.79 1 3 1 – 5.7 1 4
Stewart and Barrick (2000) −0.12 45 0.80 0.83 1 0 180 14 – 3
Taggar and Seijts (2003) 0.20 59 – 1 2 1 – – 2 2
Tesluk and Mathieu (1999) 0.05 88 0.93 0.81 2 0 468 5.4 2 2
Watson, Johnson, and Zgourides (2002) 0.12 165 – – 2 0 15 5.5 3 2
Williams et al. (2010) 0.31 43 0.43 0.66 2 0 – 7.26 2 2
Yi (2012) 0.56 43 – 0.96 1 0 – 8 2 1
Ziegert, Mayer, and Piccolo (2009) 0.11 285 0.57 0.91 – 0 – 41.14 – 1
Zhang, Waldman, and Wang (2012) 0.02 74 0.76 0.89 2 0 28 4.88 2 3
Zhou (2012) 0.45 144 – – 3 0 - 3.5 1 4
Note. r = effect size coded, N = total number of teams; rxx = reliability of shared leadership measure; ryy = reliability of team performance; Task interdependence:
1 = low, 2 = moderate, 3 = high; Performance index: 0 = subjective, 1 = objective; Team tenure = average team tenure in weeks (round estimates); Average
team size = average number of team members in a team; Team type: 1 = decision making, 2 = action, 3 = project; SL measurement approach: 1 = aggregated shared
transformational leadership approach, 2 = aggregated shared functional leadership approach, 3 = other aggregated measures of shared leadership, 4 = social net-
work approach to measuring shared leadership.
of available studies, we were not able to perform separate analyses for student course teams and student laboratory teams. Thus, both
of these types of teams were included under one category (i.e., school setting) and contrasted with studies conducted in the field.
Fifth, following DeChurch and Mesmer-Magnus (2010) we coded teams into three types — action, decision making, and project
teams (Sundstrom, de Meuse, & Futrell, 1990; Sundstrom et al., 2000). According to DeChurch and Mesmer-Magnus, action teams
were those that required high levels of behavioral interdependence for success; decision-making teams required high levels of infor-
mation exchange, and project teams required high levels of both types of interdependence. Finally, we also took into account the clar-
ity of the task performed, which is consistent with past research which views that action teams work on more defined tasks whereas
project and decision teams work on more ill-defined tasks that necessitate the acquisition of resources through social interaction
(Devine, 2002; Sundstrom et al., 1990). In doing so, we could not code for five studies that had sampled a variety of team types.
Inter-rater agreement was 94.8%.
Sixth, there were six occasions in which shared leadership was measured at more than one time point. In these studies we used the
correlation from the latter time point. We chose the latest correlation because shared leadership was more likely to have developed
later in the genesis of the team, that is, after the team members had a chance to become acquainted.6 Seventh, when distinguishing
between aggregate and SNA approaches, we considered generated ratios as a simpler form of the SNA approach.
Eighth, we coded for the nature and form of the shared leadership questionnaire used in aggregation approaches to measurement.
Specifically, we distinguished among three types of aggregated measures: 1) shared transformational/charismatic leadership mea-
sures (e.g., Sivasubramaniam et al., 2002; coded as 1), 2) shared functional leadership measures (e.g., Hiller et al., 2006; coded as
2), and 3) other aggregated measures of shared leadership that did not fall in the previous two categories (e.g., shared authentic lead-
ership, generic shared leadership questionnaire, shared vision; coded as 3). If the study included a SNA approach to measurement it
was coded as 4 (e.g., Carson et al., 2007).
In four studies, two or more estimates of an effect were provided. For example, Pearce and Sims (2002) reported correlations be-
tween shared transformational leadership and two subjective indices of team performance. Treating each effect as being independent
(i.e. coding 2 effect sizes) would lead to serious errors in estimation (e.g., Glass, McGaw, & Smith, 1981, Chapter 6). This describes a
stochastic dependent effect situation (Gleser & Olkin, 1994; Hedges & Olkin, 1985). Therefore, we used the theory of composites
(Ghiselli, Campbell, & Zedeck, 1981) to arrive at a single correlation between overall shared leadership and team performance.
Next, to counter the ubiquity of significant effects in the literature (Rosenthal, 1979), we used a file drawer analysis (Hunter &
Schmidt, 2004) to determine how many null-effect studies would be needed to reduce true correlations to a value of ρ = .05.
Finally, when studies reported both centralization and density network measures, we retained density indices. This is consistent
with past shared leadership research that preferred density measures because they “[reflect] the extent to which leadership influence
is distributed among a relatively high or relatively low proportion of team members” (Carson et al., 2007, p. 1220). Furthermore, a
disadvantage of centralization scores is that a low value can be reflective of two radically different team states (Gockel & Werth, 2010).
Overall, the first five authors independently coded five studies. The initial average interrater agreement for the five initial studies
was 88%. A meeting was held to discuss discrepancies in coding until consensus was reached. After this initial set of articles, pairs of
authors coded half of the articles with the first author coding all studies. Inter-rater agreement was 97% and differences were resolved
via consensus.
Meta-analytic procedures
We used the product–moment correlation as the key metric in this meta-analysis. Because of its undesirable statistical properties
(Lipsey & Wilson, 2001), we transformed all correlations to Fisher Zs and used them in the meta-analytic method as discussed further
below. We also back-transformed all reported results into the raw correlation metric, with the exception of the slopes of the moder-
ator regression coefficients. This exception is due to the nonlinearity of the r to Fisher Z transformation, which can hinder interpreta-
tion of such slopes when they are back-transformed.
For shared leadership's main effects analyses we employed a random-effects model, using the DerSimonian–Laird estimation
(DerSimonian & Laird, 1986; Raudenbush, 2009), in order to account for heterogeneity among the population effect sizes. Moreover,
studies were weighted by their inverse variance weight (Lipsey & Wilson, 2001). To examine the influence of moderators on shared
leadership's effectiveness, we employed the conservative mixed-effects model. Initially, we started out by testing each moderator in-
dividually. After identifying significant moderators, we followed up our investigation by including all significant moderators within a
single simultaneous model. This practice of examining multiple moderators together minimizes type I errors and makes the overall
moderator analyses more meaningful. We employed the subgroup method advocated by Hunter and Schmidt (2004) to test truly cat-
egorical moderators (e.g., setting, publication status), and also examined whether the 95% confidence intervals overlapped at different
levels of the moderator. In addition, we followed up our categorical moderator analyses by creating a dummy coded variable for a
given moderator and entering it in the meta-regression. We report the effect size and significance of the dummy coded moderator
as well as the test of moderation which was provided by the metafor package in R (Viechtbauer, 2010).
We report both uncorrected and true score correlations, correcting the latter for unreliability in the constructs of shared leadership
(√rxx = .83) and team performance (√ryy = .95) using ICC(2) values as estimates of team level reliability, which is a practice consis-
tent with past research (e.g., Stewart, 2006; Whitman, Van Rooy, & Viswesvaran, 2010). An ICC(2) value indicates the degree to which
teams can be systematically differentiated on the construct of interest (Bliese, 2000). Thus, ICC(2) reflects the reliability of a measure
6
This reasoning is consistent with past team meta-analyses which coded validities from the latest occurring measurement period (e.g., Beal, Cohen, Burke, &
McLendon, 2003).
based on the number of team members responding to the measure. In instances that reliability indices were not reported, the mean
reliability from the studies that had reported reliability was used (Hunter & Schmidt, 2004). When subjective performance was
assessed by a single rater (e.g., team leader), we used Cronbach's alpha as an index of reliability. Objective performance variables
(e.g., sales, organizational performance) were coded as having perfect reliability (i.e., ryy = 1). Furthermore, on three occasions we
corrected for dichotomization, since artificial dichotomization leads to downward and upward distortions of mean correlations and
true variation of correlations respectively (Hunter & Schmidt, 1990). All analyses were performed by the first author using Microsoft
Excel and the metafor package of the R Statistical software (Viechtbauer, 2010).
Finally, we report 95% confidence intervals in order to assess sampling error effects on population effect size estimates (Whitener,
1990). When reporting our moderator results, we present the estimated true regression coefficient (β), the test of moderators (QM),
and the test for residual heterogeneity (QE). In addition, we report the intuitive I2 index (Huedo-Medina, Sánchez-Meca, Marín-
Martínez, & Botella, 2006) in order to quantify the degree of heterogeneity. In order to aid in the interpretation of the I2 index,
Higgins and Thompson (2002) proposed a tentative classification of I2 values, in which: I2 = 25%–50%, I2 = 50%–75%, I2 N 75%
would be interpreted respectively as low, medium, and high heterogeneity.
Incremental validity analysis
We followed a similar approach to that of Cortina, Goldstein, Payne, Davison, and Gilliland (2000) in order to examine the incre-
mental validity of shared leadership on team performance over and above vertical leadership. To do so, three correlations are needed:
two validities and the inter-correlation among the predictors. Given that we were not aware of past meta-analyses examining shared
leadership with these constructs, all of the correlations were generated from studies included in the current meta-analysis except for
the relationship between vertical leadership and team performance (ρ = .33), which we borrowed from Stewart (2006).7 Finally, as a
conservative test of our model we used the harmonic mean of total sample sizes (cf. Viswesvaran & Ones, 1995).
Mediation analysis
To examine the degree to which team confidence transmits the effects of shared leadership onto team performance we constructed a
meta-analytic correlation matrix among these three constructs and used Baron and Kenny's (1986) approach to mediation. Both of the
correlations between shared leadership and team confidence and performance respectively were generated from the current study. We
used the overall shared leadership corrected validity value (i.e., ρ = .32). This correlation came from 11 studies that contained a measure
of either team potency (k = 7) or collective efficacy (k = 4) with shared leadership. To get an estimate of the correlation between team
confidence and performance we averaged correlations from previous meta-analyses. We relied on Stajkovic et al. (2009) to get correla-
tions between potency and performance (ρ = .34) and between collective efficacy and performance (ρ = .38).8 We then averaged these
two correlations to arrive at a correlation between team confidence and overall team performance (ρ = .36). Again, as a conservative test
of our model we used the harmonic mean of total sample sizes (cf. Viswesvaran & Ones, 1995). Finally, we used a Sobel test to assess the
significance of the indirect effect (MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002; Sobel, 1982).
Results
We present the analyses for each hypothesis in ascending order. First, the results of the meta-analysis between shared leadership
and team performance and the hierarchical regression analysis regarding the incremental validity of shared leadership are presented
and discussed. Second, the result of the mediation analysis is presented and discussed. Third, the results of the analyses of substantive
moderators of the shared leadership–team performance relationship are presented and discussed. Fourth, the analyses of the explor-
atory moderators are presented and discussed followed by the results of the methodological moderators. Concluding, the results for
testing multiple significant moderators simultaneously are presented. When interpreting our results we focus on the width and inclu-
sion of zero of the 95% CI, the magnitude of the I2 statistic, and whether the tests for moderators (QM) and residual heterogeneity (QE)
are significant.
Shared leadership, vertical leadership, and team performance
Our literature search yielded 54 independent effect sizes between shared leadership and team performance with a total sample size of
3882 teams. As shown in Table 2 the relationship between the two constructs is positive and of a moderate magnitude (ρ = .35, p b .01).
This provides support for Hypothesis 1a. In addition, the high I2 and significant test of residual heterogeneity (QE = 213.55, df = 53,
p b .01) warrant the examination of boundary conditions.
Next, to examine the incremental validity of shared leadership over that of vertical leadership, we ran a hierarchical regression
based on a combination of our own meta-analytic estimates and estimates from Stewart (2006). We began by first regressing team
performance onto vertical leadership at Step 1 and finally on to shared leadership at Step 2. As can be seen in Table 3 shared leadership
explained an additional 5.7% of the variance in team performance beyond vertical leadership (p b .01). Moreover, shared leadership
7
We also calculated values where (a) all correlations came from different studies and (b) the shared leadership validity was calculated using Stewart's (2006)
criteria. In both situations, the results were almost identical.
8
We used estimates for which outliers were removed by the authors.
Table 2
Overview of analyses of main effects and categorical moderators.
Variable k N μr SE μρ z 95% CI I2 Failsafe N
Main effects
Overall team performance 54 3882 .28 .033 .35 8.31 (.21, .35) 75 324
Vertical leadership 13 690 .32 .074 .39 4.33 (.17, .47) 72 89
Team confidence 11 593 .61 .068 .73 10.35 (.52, .69) 55 150
Categorical moderators
Interdependence
Low 3 272 .17 .20 .22 0.87 (−.21, .56) 87 11
Moderate 23 1418 .19 .036 .24 5.29 (.12, .26) 39 84
High 25 1787 .37 .057 .47 6.61 (.27, .49) 82 194
Subjective index 37 2430 .33 .045 .42 7.43 (.25, .42) 78 274
Objective index 16 1394 .16 .042 .20 3.91 (.08, .24) 53 48
Team type
Decision-making 16 1270 .22 .052 .28 4.19 (.12, .32) 68 74
Action 11 733 .27 .074 .34 3.72 (.13, .42) 72 64
Project 22 1407 .35 .055 .44 6.30 (.24, .46) 75 172
Shared leadership measure
Transformational aggregated 14 1077 .26 .054 .33 4.77 (.15, .36) 62 79
Functional aggregated 13 1097 .27 .052 .34 5.06 (.16, .37) 64 76
Other aggregated 11 841 .19 .109 .24 1.71 (−.02, .39) 89 42
SNA 16 867 .37 .047 .47 7.96 (.28, .46) 43 135
Criterion report source
Same source 10 561 .80 .100 1 7.85 (.60, .99) 81 –
Other source 49 3604 .24 .030 .30 7.84 (.18, .30) 67 187
Setting
Field 37 2777 .29 .040 .37 6.64 (.20, .37) 79 237
School 17 1105 .26 .051 .33 5.11 (.16, .36) 60 96
Publication status
Published 37 2384 .26 .036 .33 7.22 (.19, .33) 66 208
Unpublished 17 1498 .31 .071 .39 4.41 (.17, .45) 85 116
Note. k = number of correlations meta-analyzed; N = total number of teams; μr = estimated uncorrected average correlation in the population distribution; SE =
standard error; μρ = estimated corrected average correlation in the population distribution; z = test statistic; 95% CI = 95% confidence interval around the uncorrected
μρ; I2 = index of the degree of heterogeneity in primary studies; Failsafe N = number of studies reporting null findings necessary to reduce μρ to .05.
had a similar standardized weight with that of vertical leadership (β = .26 and β = .23 respectively, p b .01). The two forms of lead-
ership explained 16.5% of the variance in team performance. So as to provide a richer account of the variance explained by vertical
leadership and shared leadership, Table 3 also shows the results of an analysis in which the order of variable entry was reversed
(i.e., does vertical leadership explain unique variance in performance above shared leadership?). This analysis revealed that vertical
leadership explained 4.3% of the variance in team performance over and above shared leadership.9 These results provide support
for Hypothesis 1b.
Shared leadership, team confidence, and team performance
Given the aforementioned results, we proceeded to test for mediation. According to Baron and Kenny (1986), in order for medi-
ation to be supported, the independent variable (i.e., shared leadership) must be related to the mediator variable (i.e., team confi-
dence). As expected, shared leadership and team confidence were positively related (ρ = .73, p b .01) explaining 53.3% of its
variance. The independent variable must also be related to the dependent variable (i.e., team performance), although mediation
can still occur without this condition being satisfied (MacKinnon, 2002). As was indicated, shared leadership had a significant corre-
lation with team performance (ρ = .35, p b .01) explaining 12.2% of its variance. Finally, we entered both predictor and mediator in the
model when predicting team performance. Partial mediation is present when the contribution of the independent variable remains
significant, but its beta value declines after the mediator is added to the model (Baron & Kenny, 1986).
As Table 4 shows, when entering both predictors in the equation we see that team confidence remains a robust predictor of team
performance (β = .22, p b .01). Most importantly, the weight for shared leadership drops, but remains significant (β = .19, p b .01).
Both of these constructs together explained 14.6% of the variance in team performance. Finally, the Sobel test was also significant (z =
6.03, p b .01) demonstrating that the indirect effect of shared leadership on team performance via team confidence is significantly dif-
ferent from zero. In summary, these analyses provide evidence that team confidence at least partially mediates the shared leadership
to team performance relationship. Thus, Hypothesis 2 was supported.
9
We thank an anonymous reviewer for suggesting this “usefulness” analysis.
Table 3
Hierarchical regression results.
Team performance
Predictors Step 1 Step 2
β SE Β SE
Vertical leadership 0.33⁎ .034 .23⁎ .034

Shared leadership .26⁎ .034
ΔR2 5.7⁎
Adjusted R2 10.8⁎ 16.5⁎
Shared leadership .35⁎ .033 .26⁎ .034
Vertical leadership .23⁎ .034
ΔR2 4.3⁎
Adjusted R2 12.2⁎ 16.5⁎
N = 844; SE = standard error.

⁎ p b .01.
Substantive moderators of the shared leadership–team performance relationship
Task interdependence
This moderator was examined using the 51 studies for which we were able to code interdependence. As shown in Table 2, when
task interdependence was low the 95% CI included zero, although this analysis was based on three studies. On the other hand, high
levels of interdependence produced relatively strong correlations between shared leadership and team performance (ρ =
.47, p b .01). When task interdependence was moderate the correlation was of a weaker positive magnitude (ρ = .24, p b .01),
and the 95% CI for high and moderate interdependence did not overlap. Indeed, the difference between the two correlations was sig-
nificant (z = 1.97, p b .05). We followed up the subgroup method to moderation by creating a dummy coded variable which was
entered as a predictor in a mixed-effects model (0 = moderate interdependence, 1 = high interdependence). As expected, the
regression coefficient was significant (β = .14, p b .01) and the distinction accounted for a significant portion in the amount of het-
erogeneity (QM = 16.48, df = 1, p b .01). The test for residual heterogeneity was significant (QE = 185.22, df = 49, p b .01), indicating
that other moderators not considered in the model are influencing shared leadership's effectiveness. Overall, these results provide
support for Hypothesis 3.

As shown in Table 2, differentiating subjective from objective team performance indices did moderate the relationship between
shared leadership and team performance, with subjective performance indices yielding a higher validity (ρ = .42, p b .01) than objec-
tive indices of performance (ρ = .20, p b .01). The difference in the two validities was significant (z = 2.23, p b .05). Finally, team per-
formance index as a moderator accounted for a significant amount of heterogeneity (QM = 4.99, df = 1, p b .05). The test for residual
heterogeneity was significant (QE = 193.92, df = 51, p b .01), indicating that other moderators not considered in the model are
influencing shared leadership's effectiveness. Thus, Hypothesis 4 was supported.
Exploratory moderators of the shared leadership–team performance relationship
Team size
Team size did not appear to interact with shared leadership (β = −.005, z = 0.04, p N .05). We tested for a curvilinear relationship,
but it was also non-significant (β = .0007, z = 0.35, p N .05). In turn, this exploratory hypothesis was not supported.
Team tenure
Team tenure was found to moderate the association between shared leadership and team performance (β = −.0009, z = −2.96,
p b .01), such that this relationship becomes weaker as team tenure increases.10 As a moderator, team tenure accounted for a signif-
icant amount of heterogeneity (QM = 8.78, df = 1, p b .05). The test for residual heterogeneity was significant (QE = 112.96, df = 34,
p b .01), likely indicating that other moderators not considered in the model are influencing shared leadership's effectiveness.
Overall, this finding lends support to the idea that the positive effects of shared leadership can be difficult to sustain over
time. We elaborate further upon on these findings in the discussion section.
Team type
As shown in Table 2 shared leadership was beneficial to team performance in all three types of teams. However, there were no
significant differences among team types as evidenced by a nonsignificant test of moderators (QM = 2.74, df = 2, p N .05) and over-
lapping 95% confidence intervals. Thus, team type did not interact with shared leadership in the prediction of team performance.
10
The same result occurred when entering team tenure and team size simultaneously in the regression equation.
Table 4
Results of mediation analysis.
Variables Step 1 Step 2 Step 3
SL ➔ TC SL ➔ P SL & TC ➔ P
β SE β SE β SE
Independent variable
Shared leadership 0.73⁎ 0.02 0.35⁎ 0.025 0.19⁎ 0.036
Mediator
Team confidence 0.22⁎ 0.036
Adjusted R2 53.3⁎ 12.2⁎ 14.6⁎
N = 1413; P = team performance; TC = team confidence; SE = standard error.

⁎ p b .01.
Shared leadership measurement approach

As shown in Table 2, shared leadership had a significant positive correlation with team performance for all measurement ap-
proaches except for the “other aggregated” category. Next we attempted to compare aggregated approaches to the SNA approach.
To do so we created a dummy coded variable (aggregated = 0, SNA = 1) and entered it in the meta-regression. Given the non-
significance and the unclear nature of the “other aggregated” category, we collapsed the shared transformational/charismatic leader-
ship and shared functional leadership categories only when making this comparison. Interestingly, there was a tendency for SNA ap-
proaches to yield higher validities (z = 1.85, QM = 3.43, df = 1, p = .06). We suspect that the small number of primary studies
available grant less power in detecting significant effects and that future research should shed more light in this area. In summary,
there is some evidence to suggest that the manner in which one measures shared leadership can matter.
Methodological moderators of the shared leadership–team performance relationship
Criterion source
As expected, criterion source moderated the relationship between shared leadership and performance. Specifically, as shown in
Table 2, when performance was rated by the same source a very strong positive uncorrected11 true correlation emerged (μr = .80,
p b .01) whereas when performance was rated by another source a moderate positive correlation emerged (ρ = .30, p b .01). The
two validities were significantly different from one another (z = 4.17, p b .01) and the test of moderators was significant (QM =
17.46, df = 1, p b .05). Finally, the test for residual heterogeneity was significant (QE = 166.36, df = 52, p b .01), indicating that
other moderators not considered in the model are influencing shared leadership's effectiveness.
Study setting
As shown in Table 2, the setting in which shared leadership studies were conducted did not seem to contribute to differences in
estimates (z = −0.25, p N .05). Overall, the results do not provide support for the moderating effect of setting.
Publication type
As indicated in Table 2, relationships coming from published and unpublished sources had overlapping 95% CIs and a non-
significant difference (z = 0.55, p N .05). Overall, the results do not provide support for the moderating effect of publication type.
Test of multiple moderators
The above analyses found strong support for the moderating effect of task interdependence, team performance index, team tenure,
and criterion report source. Thus, as a more rigorous test of moderation, we built a mixed-effects model containing all four of the afore-
mentioned moderators in the equation. This analysis was based on 33 studies and results are shown in Table 5. First, when considered
together in the equation, all moderators remain significant predictors of shared leadership's validity, with the exception of team perfor-
mance index which had a nonsignificant effect (z = −1.57, p = .11). Second, the model explained a significant amount of heterogeneity
(QM = 69.93, df = 4, p b .01) and, importantly, the test for residual heterogeneity was non-significant (QE = 38.90, df = 28, p N .05). In
summary, these results provide more confidence in our finding that task interdependence, team tenure, and criterion report source do
moderate the relationship between shared leadership and performance.
Discussion
Summary of findings and contributions
Our findings provide support for the view that shared leadership has important effects on performance, over and above the effects
of vertical leadership. Specifically, in comparison to vertical leadership, shared leadership carried a similar effect in the prediction of
11
The formula for correcting for unreliability in shared leadership and team performance yielded a corrected true correlation value above 1.
performance when both were simultaneously entered in the regression equation. Certainly, we need to know more about how vertical
and shared leadership operate together (e.g., Conger & Pearce, 2003), including how and under what situations vertical leadership fa-
cilitates, hinders, complements, and/or supplement shared leadership. This should serve as a fruitful avenue for future research.
Second, we assessed an underlying explanatory mechanism through which shared leadership operates. This extension is impor-
tant because examining mediators that underlie the relationship illuminates how and why shared leadership can have a positive im-
pact on team performance outcomes. One way in which shared leadership contributes to performance is through the enactment of a
motivational emergent state: team confidence. Thus, in moments of doubt, team stakeholders can reference the team's shared lead-
ership and point out to team members that they clearly have the potential to be successful. This finding is also in line with past
scholars positing team leadership as a key construct for the development of efficacy (e.g., Kozlowski et al., 1996; Zaccaro, Blair,
Peterson, & Zazanis, 1995). Furthermore, the team literature has also delineated a number of other powerful emergent states,
which future shared leadership researchers can employ for similar investigations (see Mathieu et al., 2008 for a review).
Third, our investigation reveals several situations in which shared leadership is more clearly related to team performance. We
found that shared leadership is particularly effective when interdependence is high. High interdependence requires team members
to work closely with one another, coordinate, and integrate actions (e.g., Stewart & Barrick, 2000). There is evidence to suggest that
under high levels of interdependence, team processes and emergent states will lead to higher team performance (e.g., Barrick,
Bradley, Kristof-Brown, & Colbert, 2007). Indeed, past meta-analyses have shown that emergent states lead to more effectiveness
as interdependence increases (e.g., DeChurch & Mesmer-Magnus, 2010; Gully et al., 1995; Stajkovic et al., 2009). Given that shared
leadership is likely to lead to stronger emergent states (e.g., Sivasubramaniam et al., 2002) the shared leadership–team performance
relationship should get stronger as interdependence increases, consistent with our results. Furthermore, interdependence can pro-
mote familiarity among individuals (Wageman, 1995) which can make members more open and amenable to the exercise and receipt
of influence. This finding is consistent with research that has found stronger relationships between broad leader behaviors and team
performance when interdependence was high (e.g., Burke et al., 2006).
Fourth, although we did not find a significant overall moderating effect for team size, we did find evidence that team tenure
interacted with shared leadership in the prediction of team performance. Our results indicate that as team tenure increases, shared
leadership validities decrease. We offer three possibilities to explain these findings and emphasize that our first explanation has to
do with shared leadership particularly, whereas the other two explanations apply to leadership more broadly.
One reason in particular may be that team members cannot sustain the sharing of leadership over a long period of time be-
cause of a likely emergence of power struggles and process conflict (Jehn, 1997; Jehn, Northcraft, & Neale, 1999). The exercise of
leadership over time by many suggests that the power distribution within a team can shift. This also implies that individuals
must give up power, which may prove difficult for some team members. Such a phenomenon can breed process conflict, such
as having members disagree about who is responsible to complete a specific duty (Jehn & Mannix, 2001), and render shared
leadership less effective. We should note that this area needs additional theoretical and empirical attention and call for research
in this direction.
A second reason for the observed effect is that longer tenured teams may become more rigid and committed to established
policies and procedures (Janis & Mann, 1977; Keck, 1997). Such rigidity may inhibit the adoption of novel strategies to solve
problems (Katz, 1982), which would impede the effectiveness of all forms of leadership. A final reason could be attributed to
team membership changes. It can be a challenge for new members to fill in and satisfy the pre-existing leadership functions
of departing members. Indeed, membership change has been found to hinder team dynamics within teams (Chandler, Honig,
& Wiklund, 2005). If these reasons are confirmed by future research, then such findings would have great implications for prac-
titioners, because interventions can be implemented to train teams to continue to capitalize on shared leadership. We invite fu-
ture researchers to further investigate and attempt to answer these questions. Again, we are not aware of any research looking at
team tenure in the context of shared leadership and we call for future researchers to zone in this construct. This implies the use
of longitudinal designs.
Finally, we found that various methodological aspects of the studies included in this meta-analysis played a role in accounting for
the heterogeneity in effect sizes. Specifically, how the criterion is rated and its nature played a substantial role in the observed vari-
ability in effects. First, shared leadership's validity was higher when performance ratings were rated by team members, as opposed to
individuals outside the team. We suspect that common method variance and cognitive biases are likely reasons for this effect. Second,
when performance was based on objective measures, as opposed to subjective measures there was a significantly lower validity. As
stated, we suspect that objective measures are amenable to influences that are outside the control of teams and, hence, exhibit a
weaker validity. Third, the manner in which one measures shared leadership can matter, since one of the aggregated approaches
had validity whose 95% confidence interval included zero. In addition, there was a marginal tendency for SNA approaches to yield
higher validities when compared to aggregated approaches.
We note that several of our results parallel those from a recent meta-analysis by Wang et al. (2014). We confirmed their findings re-
garding shared leadership and performance, the incremental validity of shared leadership over vertical leadership, and the moderating
effects (or lack thereof) of several methodological moderators. Wang et al. also found support for the moderating effects of a variable
called work complexity that appears to be somewhat similar to task interdependence. In addition, we examined team size and type of
team, but found no moderating influence of these variables. More importantly, our study expands theirs by finding that team confidence
acts as a mediator of the shared leadership–performance relationship, and that team tenure moderates this relationship. Moreover, this
latter finding is in contrast with Wang et al.'s suggestion that the passage of time may promote the sharedness of leadership and, by ex-
tension, facilitate team processes and performance, as the current study's exploratory findings regarding team tenure suggest the
opposite.
Table 5
Test of multiple moderators.
Moderators β SE z
Intercept .88 .14 6.17

Task interdependence .22⁎ .06 3.86
Tenure −.001⁎ .0002 −4.80
Team performance index −.09 .06 −1.57
Criterion report source −.62⁎ .14 −4.46
Note. k = 33; β = slope coefficient; SE = standard error; z = test statistic; Task interdependence was coded as 0 = moderate, 1 = high; Team performance index
was coded as 0 = subjective, 1 = objective; Criterion report source was coded as 0 = same source, 1 = other source.
⁎ p b .01.
Limitations
Our meta-analysis carries many of the methodological advantages and drawbacks of the primary studies included. For example,
none of the studies included used a true experimental design with randomized control conditions in the field. As a result, extraneous
variables operating to bias results could not be controlled for, although such variables would have to operate systematically across
studies to influence our meta-analytic conclusions. The variety of original samples and contexts, from top management teams to stu-
dent project teams to consulting teams, as well as the use of published and unpublished studies, both mitigate this issue. In addition,
we employed a degree of statistical control in our continuous moderator analyses. Nonetheless, for causality to be inferred in this area,
more experimental studies are needed. Indeed, laboratory studies lend themselves nicely to control important factors such as who
gets on the team, the medium through which members interact, and the interdependencies that connect them.
Another limitation concerns the relatively small sample of primary studies included in the analyses, which may display problems
of second-order sampling error (Hunter & Schmidt, 2004). As Balkundi and Harrison (2006) put it, “This is a problem endemic to
team-level research” (p. 63). It is worth mentioning that the failsafe N's for the reported effects are large, suggesting that an ample
amount of null relationships would have to be generated in the future to overturn most of our conclusions. A further limitation is
our inability to account for the dynamic nature of team tenure in the current meta-analysis. Especially in field settings, team tenure
may vary and be dynamic as members leave and new members join a given team.12 Nonetheless, future investigations of team tenure
in the context of shared leadership should take this observation into account. Furthermore, our results cannot be used to address tem-
poral issues among the focal constructs.
Future directions
The current meta-analysis emphasizes the importance of understanding shared leadership in teams and identifies important
boundary conditions and mediating processes. An immediate next step would be to examine key correlates of shared leadership, in-
cluding antecedents, mediators, and outcomes as well as moderating effects of these relationships. First, a variety of scholars have de-
veloped conceptual models of shared leadership and its antecedents that require empirical validation (see Burke, Fiore, & Salas, 2003;
Conger & Pearce, 2003; Seers et al., 2003). For example, Day et al. (2006) suggested team learning and teamwork as inputs of shared
leadership. Other constructs, such as characteristics of team members (e.g., individual competence, individual receptivity to influence)
and the team (e.g., team size, diversity) are worthy of future investigation.
Next, research should explore other mediators and moderators of this relationship. Although we were able to examine some me-
diators and moderators, more research is needed on others. Apart from other emergent states (e.g., team cohesion, team trust, team
cognition), future research should also examine mediating team processes (e.g., action, transition, interpersonal processes) that trans-
mit the effects of shared leadership on team outcomes. In a recent meta-analysis, LePine, Piccolo, Jackson, Mathieu, and Saul (2008)
provided empirical support for the framework advanced by Marks et al. (2001). This work can provide guidance to future shared lead-
ership researchers for specifying and testing relationships involving various team constructs. Additionally, it is also important to ex-
amine characteristics of the task (e.g., task time demands), the team (e.g., informational diversity, virtuality), and the organizational
context (e.g., organizational justice and values) as moderators of these relationships. Complimenting this line of work, Ilgen et al.
(2005) recently proposed the IMOI model in which team inputs influence team outputs through mediating team processes and emer-
gent states. In addition, they argued that outputs feed back to become inputs. Thus, future shared leadership researchers should use
these frameworks when specifying and testing relationships involving various team constructs. In doing so, researchers should em-
ploy the use of multiple levels of analysis (i.e., individual, team, organizational, national) in future shared leadership investigations.
Second, given the established credo of “performance is multidimensional,” future studies should examine other dimensions of per-
formance of which shared leadership is predictive. Because shared leadership depends critically on social interaction (e.g., Cox et al.,
2003), we suspect that it may be related to “behaviors that do not support the technical core itself so much as they support the broader
organizational, social, and psychological environment in which the technical core must function” (Borman & Motowidlo, 1993, p. 73).
Such contextual performance includes instances of team member assistance which we speculate is a suitable candidate for the effects
of shared leadership.
12
We thank an anonymous reviewer for pointing out this observation.
Third, it is worth examining the conditions under which shared leadership may not be appropriate (Barry, 1991; Locke, 2003) as
well as the team leadership functions that shared leadership can and cannot satisfy. For example, Morgeson et al. (2010) provided
suggestions regarding optimal pairings of leadership sources and leadership functions. Validating their suggestions would be a step
in the right direction in team leadership research. More importantly, it is worth examining why shared leadership does not have a
strong positive correlation with team performance. Given that shared leadership involves directing others and being directed by
others it is worth asking whether certain team member KSAOs (e.g., knowledge of what needs to get done, perspective taking skill,
intelligence, and agreeableness) are congenial and susceptible to shared leadership's effectiveness.
Fourth, another area for future research concerns the measurement of shared leadership. Our results indicate that measurement
can matter. Although SNA, aggregated shared leadership, and aggregated functional shared leadership measures yielded a significant
positive relationship with team performance this was not the case for other aggregated measures of shared leadership. The small
amount of primary studies did not allow for a closer examination of this category, but it contained measures of authentic shared lead-
ership, shared vision, and overall shared leadership. In addition, there was a marginal tendency for SNA approaches to yield a higher
shared leadership validity when compared to aggregated approaches combined. Overall, these results indicate that more research is
needed in this area.
We propose that because shared leadership is inherently a relational phenomenon it may be best captured by an approach whose
unit of analysis is the leadership link between team members. Indeed, SNA captures variance in patterns of links, as opposed to aggre-
gate approaches which focus on the variance in shared leadership perceptions (e.g., Carson et al., 2007). In addition, the SNA approach
to shared leadership is very promising since it offers the possibility of examining different shared leadership network structures. For
example, Mehra et al. (2006) found that two network structures of shared leadership (i.e., distributed-coordinated and distributed-
fragmented) had differential effects. As another example, Carson et al. (2007) presented various leadership sociograms. Most impor-
tantly, most, if not all, current measurement approaches employed by primary studies rely on the use of means, and to a lesser extent
variance. Although useful, the means tell us about the amount of leadership at the team level, but it does not reflect the degree to
which leadership is shared. We suspect that amount and sharedness are correlated, but we are unaware of any published studies ad-
dressing the degree to which the two are related.
It is worth mentioning that an alternative explanation for finding a stronger effect size for the social network approach may be due
to the lack of missing data in calculating shared leadership scores.13 In short, an aggregation approach (e.g., referent-shift) is prone to
missing data (Nesterkin & Ganster, 2012), which makes estimation more complicated and can attenuate team-level relationships
(Timmerman, 2005). On the other hand, the social network approach applies high thresholds that exceed a 90% within-team response
rate, and thus instances of missing data are usually not encountered (Maloney, Johnson, & Zellmer-Bruhn, 2010).
Furthermore, shared leadership research should also examine various forms of shared leadership (see also Gronn, 2002, 2009). In
the current meta-analysis, we encountered studies that used a variety of leadership styles questionnaires, others that used items tap-
ping functional behaviors (e.g., boundary spanning), and yet some that used a variety of other shared leadership questionnaires. It is
conceivable that some shared leadership behaviors may have opposing effects on performance. For example, shared monitoring and/
or shared boundary spanning might actually be negatively related with team performance. That is, if everyone is monitoring team pro-
cesses or managing relationships with key constituents outside the team then less actual work may be accomplished.14 A similar ra-
tionale can be used for shared vision. If everyone in the team is voicing their vision then the team may lose its focus and become
detracted from its work. Thus, we suspect that certain leadership styles and functional leadership behaviors may lend themselves bet-
ter to sharing than others. Identifying and elucidating the extent to which various types of leadership styles and behaviors can be
shared effectively are worthy of future investigations.
We also advise future researchers to consider incorporating various moments of the team leadership distribution (e.g., skewness
and kurtosis) in an attempt to better capture the construct of shared leadership (i.e., what is shared, how much, and by whom) as well
as improve its predictive utility (e.g., DeRue, Hollenbeck, Ilgen, & Feltz, 2010; Kane, 1996). For example, different types of distributions
may have the same mean and variance, but different skewness and kurtosis values, which are reflective of different team phenomena
(see DeRue et al., 2010 for a similar discussion in the context of shared collective efficacy). In the context of shared leadership, such
differences would reflect different forms of leadership dispersion. Identifying various configurations of shared leadership as well as
examining their antecedents, outcomes, and operating mechanisms would be an important step in advancing our understanding of
this construct. Moreover, the development of an index, showing the extent to which leadership within a group is shared, would be
another way the sharedness problem could be addressed. In sum, creating comprehensible, reliable, and valid measures of constructs
is an indispensible first step for any scientific field.
Fifth, an important question concerning measurement when using a compositional model aggregation approach, one that has not
been addressed previously, has to do with who is the referent? Because of a lack of primary studies, we could not examine if there is a
difference when the referent is the team as a whole (i.e., referent-shift model) versus each individual of the team (i.e., direct-
consensus model). In the team-efficacy literature, Mischel and Northcraft (1997) argued that changing the referent from “I” to
“We” entails different cognitive processes, which can influence conclusions drawn from using such measures. In a similar manner,
we suspect that in the context of shared leadership it is unlikely that these two rating approaches are cognitively isomorphic. Research
has shown that going from “I” to “We” can improve within-group homogeneity and between-group variability (Klein, Conn, Smith, &
Sorra, 2001), and to improve criterion related validity (Arthur, Bell, & Edwards, 2007). Consequently, one fruitful avenue for future
researchers would be to examine the cognitive processes that are involved when individuals make such ratings. Another approach
13
We thank an anonymous reviewer for suggesting this alternative explanation.
14
We thank the anonymous reviewer for suggesting this point and example.
to data collection would be to use a team consensus model and juxtapose it to aggregation and SNA approaches. For example in the
context of team empowerment, Kirkman, Tesluk, and Rosen (2001) reported that a team consensus method explained significantly
more variance in team effectiveness than an aggregated approach. Interestingly, researchers have reported similar findings for con-
struct measures employing a team discussion methodology (Stajkovic et al., 2009). Similar investigations in the context of shared
leadership are laudable and necessary.
Finally, as many have noted, shared leadership is a process, and time has to be taken into account in order to understand it. There-
fore, it is essential that future researchers implement longitudinal designs in its study. In our searches we only found a handful of stud-
ies that included more than one time point. It is also worth noting that for a study to be considered longitudinal three or more time
points are needed (Ployhart & Vandenberg, 2010).
Conclusion
A relatively new perspective has emerged in team leadership research: shared leadership. Having become the basic unit through
which most work is done in modern organizations, teams can enact shared leadership and have the potential to achieve higher per-
formance in various manners and under certain conditions. Our meta-analysis has contributed to the literature by bringing to light one
of these manners and some of these conditions. As the interested reader can imagine, there is ample ground to be covered in shared
leadership research and its measurement. We have only explored some of it. It is our hope that this meta-analysis serves as a stepping-
stone on the way to understanding the construct of shared leadership, its benefits, and its pitfalls.
References
Alper, S., Tjosvold, D., & Law, S. A. (1998). Interdependence and controversy in group decision making: Antecedents to effective self managing teams. Organizational
Behavior and Human Decision Processes, 74, 33–52.
Ancona, D.G., & Caldwell, D. F. (1992). Bridging the boundary: External activity and performance in organizational teams. Administrative Science Quarterly, 37, 634–665.
Arthur, W., Bell, S. T., & Edwards, B.D. (2007). A longitudinal examination of the comparative criterion-related validity of additive and referent-shift consensus
operationalizations of team efficacy. Organizational Research Methods, 10, 35–58.
Avolio, B. J., Jung, D. I., Murry, W., & Sivasbramaniam, N. (1996). Building highly developed teams: Focusing on shared leadership process, efficacy, trust, and perfor-
mance. Advances in interdisciplinary studies of work teams: Team leadership, vol. 3. (pp. 173–209). US: Elsevier Science/JAI Press.
Avolio, B. J., Sivasubramaniam, N., Murry, W. D., Jung, D., & Garger, J. W. (2003). Assessing shared leadership: Development and preliminary validation of a team multifactor
leadership questionnaire. In C. L. Pearce, & J. A. Conger (Eds.), Shared leadership: Reframing the hows and whys of leadership (pp. 143–172). Thousand Oaks, CA: Sage.
Avolio, B. J., Walumbwa, F. O., & Weber, T. J. (2009). Leadership: Current theories, research, and future directions. Annual Review of Psychology, 60, 421–449.
Balkundi, P., & Harrison, D. A. (2006). Ties, leaders, and time in teams: Strong inference about network structure's effects on team viability and performance. Academy
of Management Journal, 49(1), 49–68.
Bandura, A. (1997). Self-efficacy: The exercise of control. New York: Freeman.
Bandura, A. (1998). Personal and collective efficacy in human adaptation and change. In J. G. Adair, D. Bélanger, & K. L. Dion (Eds.), Advances in psychological science.
Social, personal, and cultural aspects, Vol. 1. (pp. 51–71). Hove, England: Psychology Press/Erlbaum (UK) Taylor & Francis.
Barkema, H. G., Baum, J. A.C., & Mannix, E. A. (2002). Management challenges in a new time. The Academy of Management Journal, 45(5), 916–930.
Barker, V. L., & Patterson, P. W. (1996). Top management team tenure and top manager causal attributions at declining firms attempting turnarounds. Group and
Organization Studies, 21(3), 304–336.
Barling, J., Christie, A., & Hoption, C. (2011). Leadership. In S. Zedeck (Ed.), APA handbook of industrial and organizational psychology. Building and developing the
organization, Vol. 1. (pp. 183–240). Washington, DC, US: American Psychological Association.
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations.
Journal of Personality and Social Psychology, 51(6), 1173–1182.
Barrick, M. R., Bradley, B. H., Kristof-Brown, A. L., & Colbert, A. E. (2007). The moderating role of top management team interdependence: Implications for real teams
and working groups. Academy of Management Journal, 50(3), 544–557.
Barry, D. (1991). Managing the bossless team: Lessons in distributed leadership. Organizational Dynamics, 20(1), 31–47.
Bass, B.M., Avolio, B. J., Jung, D. I., & Berson, Y. (2003). Predicting unit performance by assessing transformational and transactional leadership. Journal of Applied
Psychology, 88(2), 207–218.
Bass, B.M., & Bass, R. (2008). The Bass handbook of leadership: Theory, research and managerial applications. New York: Free Press.
Beal, D. J., Cohen, R. R., Burke, M. J., & McLendon, C. L. (2003). Cohesion and performance in groups: A meta-analytic clarification of construct relations. Journal of Applied
Psychology, 88(6), 989–1004.
*Berdahl, J. L., & Anderson, C. (2005). Men, women, and leadership centralization in groups over time. Group Dynamics: Theory, Research, and Practice, 9(1), 45–57.
Bliese, P. D. (2000). Within-group agreement, non-independence, and reliability: Implications for data aggregation and analysis. In K. J. Klein, & S. W. Kozlowski (Eds.),
Multilevel theory, research, and methods in organizations (pp. 349–381). San Francisco: Jossey-Bass.
*Boies, K., Lvina, E., & Martens, M. L. (2010). Shared leadership and team performance in a business strategy simulation. Journal of Personnel Psychology, 9(4), 195–202.
Borman, W. C., & Motowidlo, S. J. (1993). Expanding the criterion domain to include elements of contextual performance. In N. Schmitt, & W. C. Borman (Eds.), Per-
sonnel selection in organizations (pp. 71–98). San Francisco: Jossey-Bass.
Bradford, D. L., & Cohen, J. (1984). Managing for excellence: The guide to developing high performance organizations. New York: John Wiley.
Burke, C. S., Fiore, S. M., & Salas, E. (2003). The role of shared cognition in enabling shared leadership and team adaptability. In C. L. Pearce, & J. A. Conger (Eds.), Shared
leadership: Reframing the hows and whys of leadership (pp. 103–122). Thousand Oaks, CA: Sage.
Burke, C., Stagl, K., Klein, C., Salas, E., Halpin, S., & Goodwin, G. (2006). What type of leadership behaviors are functional in teams? A meta-analysis. Leadership Quarterly,
17, 288–307.
Campbell, J. P. (1990). Modeling the performance prediction problem in industrial and organizational psychology. In M.D. Dunnette, & L. M. Hough (Eds.), (2nd ed.).
Handbook of industrial and organizational psychology, Vol. 1. (pp. 687–732). Palo Alto, CA: Consulting Psychologists Press.
Campbell, J. P., McCloy, R. A., Oppler, S. H., & Sager, C. E. (1993). A theory of performance. In N. Schmitt, & W. C. Borman (Eds.), Personnel selection in organizations
(pp. 35–70). San Francisco: Jossey-Bass.
Campion, M.A., Medsker, G. J., & Higgs, A.C. (1993). Relations between work group characteristics and effectiveness: Implications for designing effective work groups.
Personnel Psychology, 46(4), 823–850.
⁎Carson, J. B. & Tesluk, P. E. 2007. Leadership from within: A look at leadership roles in teams. Unpublished Manuscript, University of Maryland, College Park, USA.
*Carson, J. B., Tesluk, P. E., & Marrone, J. A. (2007). Shared leadership in teams: An investigation of antecedent conditions and performance. Academy of Management
Journal, 50(5), 1217–1234.
*Carte, T. A., Chidambaram, L., & Becker, A. (2006). Emergent leadership in self-managed virtual teams: A longitudinal study of concentrated and shared leadership
behaviors. Group Decision and Negotiation, 15(4), 323–343.
*Cashman, D.M. (2009). The effects of vertical leadership, team demographics, and group potency upon shared leadership emergence within technical organizations. Disser-
tation Abstracts International Section A: Humanities and Social Sciences, AAT 3320543.
Chan, D. (1998). Functional relations among constructs in the same content domain at different levels of analysis: A typology of composition models. Journal of Applied
Psychology, 83(2), 234–246.
Chandler, G. N., Honig, B., & Wiklund, J. (2005). Antecedents, moderators, and performance consequences of membership change in new venture teams. Journal of
Business Venturing, 20(5), 705–725.
*Charlier, S. D. (2012). A multi-level investigation of emergent leadership and dispersion effects in virtual teams. Dissertation Abstracts International Section A: Humanities
and Social Sciences, AAT 3526798.
*Chen, C. V., & Lee, H. (2007). Effects of transformational team leadership on collective efficacy and team performance. International Journal of Management and
Enterprise Development, 4(2), 202–217.
*Chen, T. R., Wolf, P. P., Zaccaro, S. J., & Odenheimer, S. L. (2013). Team composition and emergent states as predictors of shared leadership. Paper presented at the 28th
Annual Conference of the Society for Industrial and Organizational Psychology (Houston, TX).
Chiocchio, F., & Essiembre, H. (2009). Cohesion and performance: A meta-analytic review of disparities between project teams, production teams, and service teams.
Small Group Research, 40(4), 382–420.
Cohen, S. G., & Bailey, D. E. (1997). What makes teams work: Group effectiveness research from the shop floor to the executive suite. Journal of Management, 23(3),
239–290.
*Colbert, A. E., Barrick, M. R., & Bradley, B. H. (2014). Personality and leadership composition in top management teams: Implications for organizational effectiveness.
Personnel Psychology, 67(2), 351–387.
Coleman, P. T. (2000). Intractable conflict. In M. Deutch, & P. T. Coleman (Eds.), The handbook of conflict resolution: Theory and practice (pp. 428–450). San Francisco:
Jossey-Bass.
Conger, J. A., & Pearce, C. L. (2003). A landscape of opportunities: Future research on shared leadership. In C. L. Pearce, & J. A. Conger (Eds.), Shared leadership: Reframing
the hows and whys of leadership (pp. 193–214). Thousand Oaks, CA: Sage.
Cortina, J. M. (2003). Apples and oranges (and pears, oh my!): The search for moderators in meta-analysis. Organizational Research Methods, 6(4), 415–439.
Cortina, J. M., Goldstein, N.B., Payne, S.C., Davison, H. K., & Gilliland, S. W. (2000). The incremental validity of interview scores over and above cognitive ability and
conscientiousness scores. Personnel Psychology, 53(2), 325–351.
Cox, J. F., Pearce, C. L., & Perry, M. L. (2003). Toward a model of shared leadership and distributed influence in the innovation process: How shared leadership can en-
hance new product development team dynamics and effectiveness. In C. L. Pearce, & J. A. Conger (Eds.), Shared leadership: Reframing the hows and whys of lead-
ership (pp. 48–76). Thousand Oaks, CA: Sage Publishing.
Dansereau, F., Graen, G., & Haga, W. J. (1975). A vertical dyad linkage approach to leadership within formal organizations: A longitudinal investigation of the role mak-
ing process. Organizational Behavior & Human Performance, 13(1), 46–78.
*Daspit, J., Tillman, C. J., Boyd, N. G., & Mckee, V. (2013). Cross-functional team effectiveness: An examination of internal team environment, shared leadership, and
cohesion influences. Team Performance Management, 19(1/2), 34–56.
Day, D. V., Gronn, P., & Salas, E. (2004). Leadership capacity in teams. The Leadership Quarterly, 15(6), 857–880.
Day, D.V., Gronn, P., & Salas, E. (2006). Leadership in team-based organizations: On the threshold of a new era. Leadership Quarterly, 17(3), 211–216.
DeChurch, L. A., & Mesmer-Magnus, J. R. (2010). The cognitive underpinnings of effective teamwork: A meta-analysis. Journal of Applied Psychology, 95(1), 32–53.
*Delia, E. (2011). Complexity leadership in industrial innovation teams: A field study of leading, learning and innovating in heterogeneous teams. Dissertation Abstracts In-
ternational Section A: Humanities and Social Sciences, AAT 3443409.
DerSimonian, R., & Laird, N. (1986). Meta-analysis in clinical trials. Controlled Clinical Trials, 7(3), 177–188.
DeRue, D. S., Hollenbeck, J., Ilgen, D., & Feltz, D. (2010). Efficacy dispersion in teams: Moving beyond agreement and aggregation. Personnel Psychology, 63(1), 1–40.
Devine, D. J. (2002). A review and integration of classification systems relevant to teams in organizations. Group Dynamics: Theory, Research, and Practice, 6(4),
291–310.
Dewett, T. (2004). Creativity and strategic management: individual and group considerations concerning decision alternatives in TMTs. Journal of Managerial
Psychology, 19, 156–169.
Druskat, V. U., & Wheeler, J. V. (2003). Managing from the boundary: The effective leadership of self-managing work teams. Academy of Management Journal, 46(4),
435–457.
*Ensley, M.D., Hmieleski, K. M., & Pearce, C. L. (2006). The importance of vertical and shared leadership within new venture top management teams: Implications for
the performance of startups. Leadership Quarterly, 17(3), 217–231.
Finkelstein, S., & Hambrick, D. C. (1990). Top-management-team tenure and organizational outcomes: The moderating role of managerial discretion. Administrative
Science Quarterly, 35(3), 484–503.
Follett, M. P. (1924). Creative experience. New York, NY, US: Longmans, Green and Co.
*Foo, M. -D., Sin, H. -P., & Yiong, L. -P. (2006). Research notes and commentaries: Effects of team inputs and intrateam processes on perceptions of team viability and
member satisfaction in nascent ventures. Strategic Management Journal, 27(4), 389–399.
Ghiselli, E. E., Campbell, J. P., & Zedeck, S. (1981). Measurement theory for the behavioral sciences. San Francisco: Freeman.
Gibb, C. A. (1954). Leadership. In G. Lindzey (Ed.), Handbook of social psychology, Vol. 2. (pp. 877–917). Reading, MA7: Addison-Wesley.
Glass, G. V., McGaw, B., & Smith, M. L. (1981). Meta-analysis in social research. Beverly Hills, CA: Sage.
Gleser, L. J., & Olkin, I. (1994). Stochastically dependent effect sizes. In H. Cooper, & L. Hedges (Eds.), The handbook of research synthesis (pp. 339–355). New York: Rus-
sell Sage Foundation.
Gockel, C., & Werth, L. (2010). Measuring and modeling shared leadership: Traditional approaches and new ideas. Journal of Personnel Psychology, 9(4), 172–180.
*Greer, L. L., De Hoogh, A., Patel, P., Thatcher, S. M. B., & De Dreu, C. K. W. (2011). The dark side of shared leadership: Power struggles in teams with multiple leaders. Work-
ing paper. University of Amsterdam.
Gronn, P. (2002). Distributed leadership as a unit of analysis. Leadership Quarterly, 13(4), 423–451.
Gronn, P. (2009). Leadership configurations. Leadership, 5(3), 381–394.
Gully, S. M., Devine, D. J., & Whitney, D. J. (1995). A meta-analysis of cohesion and performance: Effects of levels of analysis and task interdependence. Small Group
Research, 26(4), 497–520.
Gully, S. M., Incalcaterra, K. A., Joshi, A., & Beaubien, J. M. (2002). A meta-analysis of team-efficacy, potency, and performance: Interdependence and level of analysis as
moderators of observed relationships. Journal of Applied Psychology, 87(5), 819–832.
*Gupta, V. K., Huang, R., & Niranjan, S. (2010). A longitudinal examination of the relationship between team leadership and performance. Journal of Leadership &
Organizational Studies, 17(4), 335–350.
*Gupta, V. K., Huang, R., & Yayla, A. A. (2011). Social capital, collective transformational leadership, and performance: A resource-based view of self-managed teams.
Journal of Managerial Studies, 23(1), 31–45.
Guzzo, R. A., & Shea, D. (1992). Group performance and intergroup relations in organizations. In M.D. Dunnette, & L. M. Hough (Eds.), Handbook of industrial and
organizational psychology, Vol. 3. (pp. 269–313). Palo Alto, CA: Consulting Psychologists Press.
Hackman, J. R. (1987). The design of work teams. In J. Lorsch (Ed.), Handbook of organizational behavior (pp. 315–342). Englewood Cliffs, NJ: Prentice-Hall.
Hackman, J. R., & Wageman, R. (2005). A theory of team coaching. Academy of Management Review, 30(2), 269–287.
Hackman, J. R., & Walton, R. E. (1986). Leading groups in organizations. In P. S. Goodman & Associates (Ed.), Designing effective work groups. San Francisco, CA: Jossey-Bass.
Haleblian, J., & Finkelstein, S. (1993). Top management team size, CEO dominance, and firm performance: The moderating roles of environmental turbulence and dis-
cretion. Academy of Management Journal, 36(4), 844–863.
Hare, A. P. (1981). Group size. American Behavioral Scientist, 24(5), 695.
Hedges, L. V., & Olkin, I. (1985). Statistical methods for meta-analysis. San Diego, CA: Academic Press.
Higgins, J. P. T., & Thompson, S. G. (2002). Quantifying heterogeneity in a meta-analysis. Statistics in Medicine, 21, 1539–1558.
Hill, G. W. (1982). Group versus individual performance: Are N + 1 heads better than one? Psychological Bulletin, 91(3), 517–539.
*Hiller, N. J., Day, D.V., & Vance, R. J. (2006). Collective enactment of leadership roles and team effectiveness: A field study. Leadership Quarterly, 17(4), 387–397.
Hinsz, V. B., Tindale, R. S., & Vollrath, D. A. (1997). The emerging conceptualization of groups as information processors. Psychological Bulletin, 121(1), 43–64.
*Hmieleski, K. M., Cole, M. S., & Baron, R. A. (2012). Shared authentic leadership and new venture performance. Journal of Management, 38(5), 1476–1499.
*Hoch, J. E. (2010). Developing a shared and vertical leadership short scale: Validating Pearce and Sims (2002) questionnaire using in three samples of German work
teams. Paper presented at the annual meeting of the Society of Industrial and Organizational Psychology, Atlanta, Georgia.
*Hoch, J. E., & Kozlowski, S. W. J. (2012). Leading virtual teams: Hierarchical leadership, structural supports, and shared team leadership. Journal of Applied Psychology.
http://dx.doi.org/10.1037/a0030264 (Advance online publication).
*Hoch, J. E., Pearce, C. L., & Welzel, L. (2010). Is the most effective team leadership shared? The impact of shared leadership, age diversity, and coordination on team
performance. Journal of Personnel Psychology, 9(3), 105–116.
Huedo-Medina, T. B., Sánchez-Meca, J., Marín-Martínez, F., & Botella, J. (2006). Assessing heterogeneity in meta-analysis: Q statistic or I2 index? Psychological Methods,
11(2), 193–206.
Humphrey, S. E. (2011). What does a great meta-analysis look like. Organizational Psychology Review, 1(2), 99–103.
Hunter, J. E., & Schmidt, F. L. (1990). Dichotomization of continuous variables: The implications for meta-analysis. Journal of Applied Psychology, 75(3), 334–349.
Hunter, J. E., & Schmidt, F. L. (2004). Methods of meta-analysis: Correcting error and bias in research findings (2nd ed.). New York, NY: Sage.
Ilgen, D. R., Hollenbeck, J. R., Johnson, M., & Jundt, D. (2005). Teams in organizations: From input–process–output models to IMOI models. Annual Review of Psychology,
56, 517–543.
*Ishikawa, J. (2012). Transformational leadership and gatekeeping leadership: The roles of norm for maintaining consensus and shared leadership in team perfor-
mance. Asia Pacific Journal of Management, 29(2), 265–283.
Janis, I. L., & Mann, L. (1977). Decision making: A psychological analysis of conflict, choice, and commitment. New York: The Free Press.
Jehn, K. (1997). A qualitative analysis of conflict types and dimensions in organizational groups. Administrative Science Quarterly, 42, 530–557.
Jehn, K. A., & Mannix, E. A. (2001). The dynamic nature of conflict: A longitudinal study of intragroup conflict and group performance. Academy of Management Journal,
44(2), 238–251.
Jehn, K. A., Northcraft, G. B., & Neale, M.A. (1999). Why differences make a difference: A field study of diversity, conflict, and performance in workgroups. Administrative
Science Quarterly, 44, 741–763.
Kane, J. S. (1996). The conceptualization and representation of total performance effectiveness. Human Resource Management Review, 6(2), 123–145.
Kaplan, S., Cortina, J., Ruark, G., LaPort, K., & Nicolaides, V. (2014). The role of organizational leaders in employee emotion management: A theoretical model. The
Leadership Quarterly, 25(3), 563–580.
Katz, R. (1982). The effects of group longevity on project communication and performance. Administrative Science Quarterly, 27, 81–104.
Keck, S. L. (1997). Top management team structure: Differential effects by environmental context. Organizational Science, 8(2), 143–156.
Kenny, D. A., Kashy, D., & Cook, W. L. (2006). Dyadic data analysis. New York, NY, US: Guilford Press.
Kerr, S., & Jermier, J. M. (1978). Substitutes for leadership: Their meaning and measurement. Organizational behavior and human performance, 22(3), 375–403.
Kirkman, B.L., Tesluk, P. E., & Rosen, B. (2001). Assessing the incremental validity of team consensus ratings over aggregation of individual-level data in predicting team
effectiveness. Personnel Psychology, 54(3), 645–667.
Klein, K. J., Conn, A. B., Smith, D. B., & Sorra, J. S. (2001). Is everyone in agreement? An exploration of within-group agreement in employee perceptions of the work
environment. Journal of Applied Psychology, 86, 3–16.
Klein, K. J., Ziegert, J. C., Knight, A. P., & Xiao, Y. (2006). Dynamic delegation: Shared, hierarchical, and deindividualized leadership in extreme action teams.
Administrative Science Quarterly, 51, 590–621.
Kozlowski, S. W. J., & Bell, B.S. (2003). Work groups and teams in organizations. In W. C. Borman, D. R. Ilgen, & R. J. Klimoski (Eds.), Comprehensive handbook of
psychology: Industrial and organizational psychology, Vol. 12. (pp. 333–375). New York: John Wiley.
Kozlowski, S. W. J., Gully, S. M., Salas, E., & Cannon-Bowers, J. A. (1996). Team leadership and development: Theory, principles, and guidelines for training leaders and
teams. Advances in interdisciplinary studies of work teams: Team leadership, Vol. 3. (pp. 253–291). US: Elsevier Science/JAI Press.
Kozlowski, S. W. J., & Ilgen, D. R. (2006). Enhancing the effectiveness of work groups and teams. Psychological Science in the Public Interest, 7(3), 77–124.
Kozlowski, S. W. J., Watola, D. J., Jensen, J. M., Kim, B. H., & Botero, I. C. (2009). Developing adaptive teams: A theory of dynamic team leadership. Team effectiveness in
complex organizations: Cross-disciplinary perspectives and approaches (pp. 113–155). New York, NY, US: Routledge/Taylor & Francis Group.
Latané, B. (1981). The psychology of social impact. American Psychologist, 36(4), 343–356.
LePine, J. A., Piccolo, R. F., Jackson, C. L., Mathieu, J. E., & Saul, J. R. (2008). A meta-analysis of teamwork processes: Tests of a multidimensional model and relationships
with team effectiveness criteria. Personnel Psychology, 61(2), 273–307.
Lester, S. W., Meglino, B.M., & Korsgaard, M.A. (2002). The antecedents and consequences of group potency: A longitudinal investigation of newly formed work groups.
Academy of Management Journal, 45(2), 352–368.
*Lin, Z., Yang, H., Arya, B., Huang, Z., & Li, D. (2005). Structural versus individual perspectives on the dynamics of group performance: Theoretical exploration and em-
pirical investigation. Journal of Management, 31(3), 354–380.
Lipsey, M. W., & Wilson, D. B. (2001). Practical meta-analysis. Thousand Oaks, CA: Sage.
Locke, E. A. (2003). Leadership: Starting at the top. In C. J. Pearce, & C. Conger (Eds.), Shared leadership: Reframing the hows and whys of leadership (pp. 271–284). Thousand
Oaks, CA: Sage.
MacKinnon, D. P. L. (2002). A comparison of methods to test mediation and other intervening variable effects. Psychological Methods, 7, 83–104.
MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V. (2002). A comparison of methods to test mediation and other intervening variable effects.
Psychological methods, 7(1), 83.
Maier, N. R. (1967). Assets and liabilities in group problem solving: The need for an integrative function. Psychological Review, 74(4), 239–249.
Maloney, M. M., Johnson, S. G., & Zellmer-Bruhn, M. E. (2010). Assessing group-level constructs under missing data conditions: A Monte Carlo simulation. Small Group
Research, 41, 281–307.
Marks, M.A., Mathieu, J. E., & Zaccaro, S. J. (2001). A temporally based framework and taxonomy team processes. Academy of Management Review, 26(3), 356–376.
*Marrone, J. A., Tesluk, P. E., & Carson, J. B. (2007). A multilevel investigation of antecedents and consequences of team member boundary-spanning behavior. Academy
of Management Journal, 50(6), 1423–1439.
Mathieu, J., Maynard, M. T., Rapp, T., & Gilson, L. (2008). Team effectiveness 1997–2007: A review of recent advancements and a glimpse into the future. Journal of
Management, 34(3), 410–476.
Mayo, M., Meindl, J. R., & Pastor, J. C. (2003). Shared leadership in work teams. A social network approach. In C. L. Pearce, & J. A. Conger (Eds.), Shared leadership:
Reframing the hows and whys of leadership (pp. 193–214). Thousand Oaks, CA: Sage.
McGrath, J. E. (1962). Leadership behavior: Some requirements for leadership training. Washington, DC: U.S. Civil Service Commission.
McGrath, J. E., Arrow, H., & Berdahl, J. L. (2000). The study of groups: Past, present, and future. Personality and Social Psychology Review, 4(1), 95–105.
*McIntyre, H. H., & Foti, R. J. (2013). The impact of shared leadership on teamwork mental models and performance in self-directed teams. Group Processes & Intergroup
Relations, 16(1), 46–57.
*Mehra, A., Smith, B. R., Dixon, A. L., & Robertson, B. (2006). Distributed leadership in teams: The network of leadership perceptions and team performance. Leadership
Quarterly, 17(3), 232–245.
*Mehta, A., Feild, H., Armenakis, A., & Mehta, N. (2009). Team goal orientation and team performance: The mediating role of team planning. Journal of Management,
35(4), 1026–1046.
*Mendez, M. J. (2010). A closer look into collective leadership: Is leadership shared or distributed? Dissertation Abstracts International Section A: Humanities and Social Sci-
ences, AAT 3383040.
Mischel, L. J., & Northcraft, G. B. (1997). “I think we can, I think we can…”: The role of efficacy beliefs in group and team effectiveness. In B. Markoversusky, M. J.
Lovaglia, & R. Simon (Eds.), Advances in group processes, Vol. 14. (pp. 177–197). Greenwich, CT: JAI.
Mohrman, S. A., Cohen, S. G., & Mohrman, A.M., Jr. (1995). Designing team-based organizations: New forms for knowledge work. San Francisco: Jossey-Bass.
Morgeson, F. P. (2005). The external leadership of self-managing teams: Intervening in the context of novel and disruptive events. Journal of Applied Psychology, 90(3),
497–508.
Morgeson, F. P., DeRue, D. S., & Karam, E. P. (2010). Leadership in teams: A functional approach to understanding leadership structures and processes. Journal of
Management, 36(1), 5–39.
Morrill, C. (1995). The executive way. Chicago: University of Chicago Press.
*Muethel, M., Gehrlein, S., & Hoegl, M. (2012). Socio‐demographic factors and shared leadership behaviors in dispersed teams: Implications for human resource man-
agement. Human Resource Management, 51(4), 525–548.
Nesterkin, D. A., & Ganster, D.G. (2012). The effects of nonresponse rates on group-level correlations. Journal of Management. http://dx.doi.org/10.1177/
1094428106298969 (Pagination not provided).
*Neubert, M. J. (1999). Too much of a good thing or the more the merrier? Exploring the dispersion and gender composition of informal leadership in manufacturing
teams. Small Group Research, 30(5), 635–646.
Pearce, C. L. (2004). The future of leadership: Combining vertical and shared leadership to transform knowledge work. Academy of Management Executive, 18(1),
47–57.
Pearce, C. L., & Conger, J. A. (2003). Shared leadership: Reframing the hows and whys of leadership. Thousand Oaks, CA: Sage.
*Pearce, C. L., & Ensley, M.D. (2004). A reciprocal and longitudinal investigation of the innovation process: The central role of shared vision in product and process
innovation teams (PPITs). Journal of Organizational Behavior, 25(2), 259–278.
Pearce, C. L., & Sims, H. P., Jr. (2000). Shared leadership: Toward a multi-level theory of leadership. Advances in the Interdisciplinary Studies of Work Teams, 7, 115–139.
*Pearce, C. L., & Sims, H. P. (2002). Vertical versus shared leadership as predictors of the effectiveness of change management teams: An examination of aversive, di-
rective, transactional, transformational, and empowering leader behaviors. Group Dynamics: Theory, Research, and Practice, 6(2), 172–197.
*Pearce, C. L., Yoo, Y., & Alavi, M. (2004). Leadership, social work, and virtual teams: The relative influence of vertical versus shared leadership in the nonprofit sector. Im-
proving leadership in nonprofit organizations. San Francisco, CA, US: Jossey-Bass, 180–203.
Perry, M. L., Pearce, C. L., & Sims, H. P. (1999). Empowered selling teams: How shared leadership can contribute to selling team outcomes. Journal of Personal Selling &
Sales Management, 19(3), 35–51.
Pfeffer, J. (1981). Power in organizations. Marshfield, MA: Pitman.
Pfeffer, J. (1997). The ambiguity of leadership. Leadership: Understanding the dynamics of power and influence in organizations. Notre Dame: University of Notre Dame
Press, 54–68.
Pirola-Merlo, A., Hartel, C., Mann, L., & Hirst, G. (2002). How leaders influence the impact of affective events on team climate and performance in R&D teams. Leadership
Quarterly, 13(5), 561–581.
Ployhart, R. E., & Vandenberg, R. J. (2010). Longitudinal research: The theory, design, and analysis of change. Journal of Management, 36(1), 94–120.
*Potter, R., Balthazard, P., & Waldman, D. A. (2010). Shared leadership in teams: Modelling performance in traditional computer-supported environments. Paper pre-
sented at the annual meeting of the Academy of Management, Montreal, Canada.
*Ramthun, A. J., McElravy, L. J., & Matkin, G. S. (2013). Living dangerously: Shared leadership and performance for teams in dangerous environment. Manuscript sub-
mitted to the Academy of Management Annual Meeting.
⁎Rasmussen, W. O. 2011. Time is what you make of it. On increasing and effectively using volunteer's time contributions in team volunteering. Unpublished
manuscript
Raudenbush, S. W. (2009). Analyzing effect sizes: Random effects models. In H. Cooper, L. V. Hedges, & J. C. Valentine (Eds.), The handbook of research synthesis and
meta-analysis (pp. 295–315) (2nd ed.). New York: Russell Sage Foundation.
Richardson, H. A., Vandenberg, R. J., Blum, T. C., & Roman, P.M. (2010). Does decentralization make a difference for the organization? An examination of the boundary
conditions circumscribing decentralized decision-making and organizational financial performance. Readings in organizational behavior (pp. 472–497). New York,
NY, US: Routledge/Taylor & Francis Group.
Rosenthal, R. (1979). The “file drawer problem” and tolerance for null results. Psychological Bulletin, 86, 638–641.
*Rousseau, V., & Aubé, C. (2010). Team self-managing behaviors and team effectiveness: The moderating effect of task routineness. Group & Organization Management,
35(6), 751–781.
Saavedra, R., Earley, P. C., & Van Dyne, L. (1993). Complex interdependence in task-performing groups. Journal of Applied Psychology, 78(1), 61–72.
Salas, E., Goodwin, G. F., & Burke, C. S. (Eds.). (2009). Team effectiveness in complex organizations: Cross-disciplinary perspectives and approaches. The organizational
frontiers series. New York, NY, US: Routledge/Taylor & Francis Group.
*Sanders, T. O. (2006). Collectivity and influence: The nature of shared leadership and its relationship with team learning orientation, vertical leadership and team effective-
ness. Dissertation Abstracts International Section A: Humanities and Social Sciences, AAT 3237041.
Schmidt, F. L. (1992). What do data really mean? Research findings, meta-analysis, and cumulative knowledge in psychology. American Psychologist, 47(10), 1173.
Seers, A., Keller, T., & Wilkerson, J. M. (2003). Can team members share leadership: Foundations in research and theory. In C. L. Pearce, & J. A. Conger (Eds.), Shared
leadership: Reframing the hows and whys of leadership (pp. 77–102). Thousand Oaks, CA: Sage.
Sell, J., Lovaglia, M. J., Mannix, E. A., Samuelson, C. D., & Wilson, R. K. (2004). Investigating conflict, power, and status within and among groups. Small Group Research,
35(1), 44–72.
Shaw, M. E. (1981). Group dynamics. New York: McGraw-Hill.
Shea, G. P., & Guzzo, R. A. (1987). Groups as human resources. In K. M. Rowland, & G. R. Ferris (Eds.), Research in personnel and human resources management, vol. 5. (pp.
323–356). Greenwich, CT: JAI.
Sirmon, D.G., Hitt, M.A., & Ireland, R. (2007). Managing firm resources in dynamic environments to create value: Looking inside the black box. Academy of Management
Review, 32(1), 273–292.
*Sivasubramaniam, N., Murry, W. D., Avolio, B. J., & Jung, D. I. (2002). A longitudinal model of the effects of team leadership and group potency on group performance.
Group & Organization Management, 27(1), 66–96.
*Small, E. E., & Rentsch, J. R. (2010). Shared leadership in teams: A matter of distribution. Journal of Personnel Psychology, 9(4), 203–211.
*Smith, A., Houghton, S. M., Hood, J. N., & Ryman, J. A. (2006). Power relationships among top managers: Does top management team power distribution matter for
organizational performance? Journal of Business Research, 59(5), 622–629.
Smith, K. G., Smith, K. A., Olian, J.D., Sims, H. P., O'Bannon, D. P., & Scully, J. A. (1994). Top management team demography and process: The role of social integration and
communication. Administrative Science Quarterly, 39(3), 412–438.
Sobel, M. E. (1982). Asymptotic confidence intervals for indirect effects in structural equation models. Sociological methodology, 13(1982), 290–312.
*Solansky, S. T. (2008). Leadership style and team processes in self-managed teams. Journal of Leadership & Organizational Studies, 14(4), 332–341.
Stajkovic, A.D., Lee, D., & Nyberg, A. J. (2009). Collective efficacy, group potency, and group performance: Meta-analyses of their relationships, and test of a mediation
model. Journal of Applied Psychology, 94(3), 814–828.
Stewart, G. L. (2006). A meta-analytic review of relationships between team design features and team performance. Journal of Management, 32(1), 29–55.
*Stewart, G. L., & Barrick, M. R. (2000). Team structure and performance: Assessing the mediating role of intrateam process and the moderating role of task type.
Academy of Management Journal, 43(2), 135–148.
Sundstrom, E., de Meuse, K. P., & Futrell, D. (1990). Work teams: Applications and effectiveness. American Psychologist, 45(2), 120–133.
Sundstrom, E., McIntyre, M., Halfhill, T., & Richards, H. (2000). Work groups: From the Hawthorne studies to work teams of the 1990s and beyond. Group Dynamics:
Theory, Research, and Practice, 4(1), 44–67.
*Taggar, S., & Seijts, G. H. (2003). Leader and staff role-efficacy as antecedents of collective-efficacy and team performance. Human Performance, 16(2), 131–156.
Tesluk, P. E., & Jacobs, R. R. (1998). Toward an integrated model of work experience. Personnel Psychology, 51(2), 321–355.
*Tesluk, P. E., & Mathieu, J. E. (1999). Overcoming roadblocks to effectiveness: Incorporating management of performance barriers into models of work group effec-
tiveness. Journal of Applied Psychology, 84(2), 200–217.
Thompson, J.D. (1967). Organizations in action. New York: McGraw-Hill.

Timmerman, T. A. (2005). Missing persons in the study of groups. Journal of Organizational Behavior, 26, 21–36.
Triplett, N. (1898). The dynamogenic factors in pacemaking and competition. The American Journal of Psychology, 9(4), 507–533.
Viechtbauer, W. (2010). Conducting meta-analyses in R with the metafor package. Journal of Statistical Software, 36(3), 1–48.
Viswesvaran, C., & Ones, D. S. (1995). Theory testing: Combining psychometric meta-analysis and structural equations modeling. Personnel Psychology, 48(4), 865–885.
Wageman, R. (1995). Interdependence and group effectiveness. Administrative Science Quarterly, 40, 145–180.
Wang, D., Waldman, D. A., & Zhang, Z. (2014). A meta-analysis of shared leadership and team effectiveness. Journal of Applied Psychology, 99(2), 181–198.
Wanous, J. P., Sullivan, S. E., & Malinak, J. (1989). The role of judgment calls in meta-analysis. Journal of Applied Psychology, 74, 259–264.
*Watson, W. E., Johnson, L., & Zgourides, G. D. (2002). The influence of ethnic diversity on leadership, group process, and performance: An examination of learning
teams. International Journal of Intercultural Relations, 26(1), 1–16.
Weibler, J., & Rohn-Endres, S. (2010). Learning conversation and shared network leadership: Development, gestalt, and consequences. Journal of Personnel Psychology,
9, 181–194.
Whitener, E. M. (1990). Confusion of confidence intervals and credibility intervals in meta-analysis. Journal of Applied Psychology, 75, 315–321.
Whitman, D. S., Van Rooy, D. L., & Viswesvaran, C. (2010). Satisfaction, citizenship behaviors, and performance in work units: A meta-analysis of collective construct
relations. Personnel Psychology, 63, 41–81.
Williams, H. M., Parker, S. K., & Turner, N. (2010). Proactively performing teams: The role of work design, transformational leadership, and team composition. Journal of
Occupational and Organizational Psychology, 83(2), 301–324.
*Yi, C. S. (2012). Exploring the implications of vertical and shared leadership for team effectiveness in retail shops in Hong Kong. Dissertation Abstracts International Section A:
Humanities and Social Sciences.
Yukl, G. A. (2007). Leadership in organizations (7th ed.). Englewood Cliffs, NJ: Prentice Hall.
Zaccaro, S. J., Blair, V., Peterson, C., & Zazanis, M. (1995). Collective efficacy. In Self-efficacy, adaptation, and adjustment (pp. 305–328). US: Springer.
Zaccaro, S. J., Heinen, B., & Shuffler, M. (2009). Team leadership and team effectiveness. In E. Salas, G. F. Goodwin, & C. S. Burke (Eds.), Team effectiveness in complex
organizations: Cross-disciplinary perspectives and approaches (pp. 83–111). New York, NY, US: Routledge/Taylor & Francis Group.
Zaccaro, S. J., Rittman, A. L., & Marks, M.A. (2001). Team leadership. Leadership Quarterly, 12(4), 451–483.
*Zhang, Z., Waldman, D. A., & Wang, Z. (2012). A multilevel investigation of leader–member exchange, informal leader emergence, and individual and team perfor-
mance. Personnel Psychology, 65(1), 49–78.
*Zhou, W. (2012). Moderating and mediating effects of shared leadership on the relationship between entrepreneurial team diversity and performance. Dissertation Abstracts
International Section A: Humanities and Social Sciences, AAT 3541506.
*Ziegert, J. C. (2007). Does more then one cook spoil the broth? An examination of shared leadership. Philadelphia, Pennsylvania: Paper presented at the annual meeting
of the Academy of Management.
*Ziegert, J. C., Mayer, D.M., & Piccolo, R. F. (2009). Context matters: Examining contextual influences on shared leadership. Paper presented at the annual meeting of the
Society of Industrial and Organizational Psychology, New Orleans, Louisiana.

The Shared Leadership of Teams A Meta Analysis of Proxim - 2014 - The Leadershi PDF

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

The Shared Leadership of Teams A Meta Analysis of Proxim - 2014 - The Leadershi PDF

Uploaded by

Copyright:

Available Formats

The Leadership Quarterly 25 (2014) 923–942

Contents lists available at ScienceDirect

The Leadership Quarterly

The shared leadership of teams: A meta-analysis of proximal,

Theory and hypothesis development

The added value of shared leadership to team performance

Hypothesis 1a. Shared leadership explains signiﬁcant variance in team performance.

Shared leadership, team conﬁdence, and team performance

Moderators of the effects of shared leadership on team performance

Substantive moderators of the shared leadership–team performance relationship

Team performance index

Exploratory moderators of the shared leadership–team performance relationship

Shared leadership measurement

Criteria for inclusion

Berdahl and Anderson (2005) Study 2 0.41 41 – – 2 0 10 3.95 3 4

Incremental validity analysis

Shared leadership, vertical leadership, and team performance

Variable k N μr SE μρ z 95% CI I2 Failsafe N

Shared leadership, team conﬁdence, and team performance

Predictors Step 1 Step 2

Vertical leadership 0.33⁎ .034 .23⁎ .034

N = 844; SE = standard error.

Substantive moderators of the shared leadership–team performance relationship

Team performance index

Exploratory moderators of the shared leadership–team performance relationship

Variables Step 1 Step 2 Step 3

N = 1413; P = team performance; TC = team conﬁdence; SE = standard error.

Shared leadership measurement approach

Methodological moderators of the shared leadership–team performance relationship

Test of multiple moderators

Summary of ﬁndings and contributions

Intercept .88 .14 6.17

Thompson, J.D. (1967). Organizations in action. New York: McGraw-Hill.

You might also like