A Meta-Analysis of Shared Leadership: Antecedents,Consequences,and Moderators

Abstract

Many organizations are encouraging a shared leadership approach that meets the increased complexity of today’s working environment. It is therefore imperative for researchers to clearly comprehend the mechanism of shared leadership in teams. Contributing to the burgeoning research in the field of shared leadership, this study aims to advance our understanding along the many dimensions of the shared leadership phenomenon: its antecedents, moderators, and consequences. In this article, we provide a critical and comprehensive analysis of the extant literature and generate an integrated framework that presents seven hypotheses and five research questions. We then empirically test this framework via a systematic meta-analysis from 40 studies (team n = 3,019). Significantly, our findings reveal that the internal team environment and team heterogeneity are positively related to the emergence of shared leadership in teams. Moreover, we confirm the positive relationship between shared leadership and team outcomes. Our analysis also highlights how intragroup trust and task interdependence significantly moderate the shared leadership–team outcomes relations, with higher correlations observed in greater levels of intragroup trust, as well as larger levels of task interdependence. We also find the moderating effect of shared leadership measurement methods in such relations. Specifically, there is a stronger relationship when shared leadership is measured with social network analysis, rather than aggregating approaches. Overall, our study brings valuable insights into the shared leadership area and provides clear directions for future research.

Keywords

shared leadership team outcomes meta-analysis

Introduction

As a result of the proliferation of self-managed teams (Solansky, 2008) and decentralized organizational designs (Balogun & Gerry, 2004), the topic of shared leadership (also termed distributed or collective leadership) has recently emerged in the literature and received considerable attention. It offers an alternative perspective on leadership: from a traditional understanding of a leader-centric and individual-level phenomenon, to a dynamic, interactive group-level leadership phenomenon (Pearce, 2004). Shared leadership has been described as a condition where teams collectively exert leadership influence (Cox, Pearce, & Perry, 2003). Under conditions of shared leadership, leadership is regarded as a series of role functions that could be performed by any members of the group (Shane & Fields, 2007), or as an emergent team property where leadership influence and responsibilities are distributed among multiple individuals (Carson, Tesluk, & Marrone, 2007).

Research on shared leadership has gained momentum since the publication of Pearce and Conger’s seminal book in 2003 (Barnett & Weidenfeller, 2016). In addition to numerous individual studies, three meta-analytic reviews have been conducted (see D’Innocenzo, Mathieu, & Kukenberger, 2014; Nicolaides et al., 2014; D. Wang, Waldman, & Zhang, 2014). These studies examined the correlation between shared leadership and team outcomes, via conducting in-depth academic inquiry on different forms of the shared leadership concept—for example, shared traditional leadership, shared new-genre leadership, and cumulative, overall shared leadership (D. Wang et al., 2014); different shared leadership measurement techniques—for example, social network analysis and aggregation techniques (D’Innocenzo et al., 2014); as well as different moderating roles on this relationship—for example, task interdependence and team confidence (Nicolaides et al., 2014). However, the results produced are not consistent across the research stream (as summarized in Table 1). Table 1 presents a critical analysis of the past meta-analytic reviews in the area of shared leadership and specifically reveals the conflicting findings of the research and their corresponding variables. For example, in terms of shared leadership assessment approaches, both D’Innocenzo et al. (2014) and Nicolaides et al. (2014) found that there is a stronger association between shared leadership and team outcomes when shared leadership is measured with social network techniques, rather than aggregation-based methods. In contrast, D. Wang et al. (2014) argued that shared leadership measurement approaches have no such influence on this relationship. Similar conflicting findings are also visible when comparing variables such as team setting, team performance measures, and task interdependence (see Table 1). To promote a better understanding of the influence of shared leadership on team outcomes, further investigation on these contradicting results is required. It is also important to consider how recent research has shaped and contributed to these debates. Therefore, our initial research objective is to provide an updated meta-analytic assessment in order to further examine the conflicting results among previous studies.

Table 1.

A Critical Analysis on Past Meta-Analytic Reviews in the Shared Leadership Area.

Variables	Conflicting findings	Reference
Shared leadership measurement techniques • Aggregation • Social network analysis	i. The relationships between shared leadership and team performance are stronger when shared leadership is measured with social network techniques than aggregation-based methods.	D’Innocenzo et al. (2014); Nicolaides et al. (2014)
	ii. Shared leadership measurement approaches have no influence on these relationships.	D. Wang et al. (2014)
Team setting • Classroom/lab simulation • Data collected in the field	i. Sample studied in the field exhibited higher effect sizes than that in the classroom or lab.	D’Innocenzo et al. (2014)
	ii. Fail to find support that these relations would be stronger when sample is studied in the field rather than classroom or lab.	Nicolaides et al. (2014); D. Wang et al. (2014)
Team performance measures • Subjective performance • Objective performance	i. There is a lower average effect sizes when performance is measured with objective indices than subjective indices.	Nicolaides et al. (2014); D. Wang et al. (2014)
	ii. No significant influence of team performance measures on these relationships.	D’Innocenzo et al. (2014)
Task interdependence • High interdependence • Low interdependence	i. Stronger relationship is found under conditions of higher versus lower levels of interdependence.	Nicolaides et al. (2014)
	ii. The results do not provide support for the moderating effect of task interdependence.	D’Innocenzo et al. (2014)

The second purpose of our research is to analyze and compare the relationships between shared leadership and different types of team outcomes. Prior work reported the positive influence of shared leadership on teams, including team performance (Ensley, Hmieleski, & Pearce, 2006), team innovation (Hoch, 2013), team satisfaction (Serban & Roberts, 2016), team functioning (Bergman, Rentsch, Small, Davenport, & Bergman, 2012), and team proactivity (Erkutlu, 2012). In detail, for example, in the work of Hoch (2013), shared leadership was found to be positively related to innovative behavior in teams. Hoch (2013) specifically suggested that when shared leadership exists in groups, members are more likely to contribute ideas and make their unique information accessible to the other group members. This process results in high degrees of creativity and idea generation. Moreover, Bergman et al. (2012) suggested that shared leadership contributes to overall team functioning. In particular, under higher levels of shared leadership, teams experience less conflict and greater consensus, in conjunction with higher intragroup trust and cohesion. We can therefore conclude from the extant literature that there is a positive relationship between shared leadership and different types of team outcomes. However, there is a lack of existing research that clarifies and compares the specific influence of team leadership on specific facets of team outcomes. As such, we argue that there is a need to investigate the spectrum of potential variations among these effects. Whereas prior research only investigated one or two types of team outcomes, our meta-analysis study provides the opportunity to simultaneously examine multiple forms of team outcomes and their corresponding relationships with shared leadership.

The third purpose of this research is to gain a holistic understanding of how to foster the successful emergence of a mutual sharing of leadership within an organization. Despite calls for more research in this area, to date there has been no meta-analysis conducted to explicitly investigate the antecedents that support the emergence of shared leadership. For instance, Carson et al. (2007) called for an academic inquiry on the predictors of shared leadership development, such as team empowerment, team composition, and contextual factors. Additionally, Fausing, Joensson, Lewandowski, and Bligh (2015); Grille, Schulte, and Kauffeld (2015); and Zhou (2016) all advocated for further examination on identifying antecedent variables of shared leadership. Our study therefore responds to these calls and analyzes potential antecedent conditions that contribute to the emergence of shared leadership.

In summary, our study provides a comprehensive, critical meta-analytic review of shared leadership’s antecedents, consequences, and moderators to enhance an understanding of the mechanism of shared leadership within teams. We critically synthesize the extant literature and offer a theoretical framework that links shared leadership to team outcomes, in relation with group behavior, attitudinal outcomes, team cognition, and team performance. This framework also includes two groups of antecedent variables: internal team environment and team characteristics, and two groups of moderating variables: substantive moderators and methodological moderators. We empirically test this framework based on random effects, meta-analytic procedures as suggested by Schmidt and Hunter (2014). In the process, we strive to make four significant contributions. First, our research builds on and integrates the theory of shared leadership, as well as updating existing meta-analytic research on shared leadership. Second, it answers the calls from researchers to examine factors that determine the emergence of shared leadership, which is largely ignored by past meta-analyses. Third, it offers valuable insights into previously inconsistent findings. Finally, we provide an enhanced understanding of the nature of shared leadership and its influence on teams, in addition to illuminating promising directions for future research.

Literature Review

Shared Leadership Concepts

Many researchers have attempted to define what shared leadership is. For example, Pearce and Sims (2002) considered it in terms of the “serial emergence” of multiple leaders over the life of a team. Carson et al. (2007) linked it to results coming from the distribution of leadership influence among team members. Shane and Fields (2007) defined shared leadership as a state or quality of mutual influence. These definitions are more similar than different. They all emphasize the significance of the distributed, collective, and/or mutual influence among multiple group members. Under the condition of shared leadership, team members collectively exert leadership influence, participate in the decision-making process, fulfill tasks traditionally reserved for a hierarchical leader, and, when appropriate, offer guidance to other members to help achieve group goals (Shane & Fields, 2007).

Synthesis of Studies in the Area of Shared Leadership

In recent years, there have been numerous empirical studies published on shared leadership (see Table 2). As shown in Table 2, researchers have investigated shared leadership in many organizational contexts, such as decision-making teams (Bergman et al., 2012), manufacturing teams (Rolfsen et al., 2013), and entrepreneurial teams (Zhou, 2016). They have also explored the factors that benefit the emergence of shared leadership. For example, Serban and Roberts (2016) examined the ways in which both internal team environment and task ambiguity serve as predictors of shared leadership. Similarly, Fausing et al. (2015) studied how empowering leadership and task interdependence support the development of shared leadership. Other studies investigated the influence of shared leadership on team outcomes—for example, team performance (Rolfsen et al., 2013), team creativity (Gu, Chen, Huang, Liu, & Huang, 2016), and team proactivity (Erkutlu, 2012). Additional scholars explored the moderating roles of shared leadership and certain team outcomes relations. For instance, Chiu, Owens, and Tesluk (2016) studied the role of team capability in the relationship between shared leadership and team effectiveness, while Erkutlu (2012) investigated the role of organizational culture in the relationship between shared leadership and team proactivity. However, these studies are quite specific and narrow in focus, making it difficult to gain a holistic and clear understanding of the influence of shared leadership in teams. Consequently, a critical synthesis on the literature is required in order to generate a complete overview of the concept of shared leadership, and our study serves to address this gap. Based on our analyses, we develop a comprehensive, integrated framework that illustrates how shared leadership emerges, the extent of its influence on team outcomes, and the specific factors that influence relationships between shared leadership and team outcomes. This framework is the first academic inquiry of its kind, which offers a straightforward picture with novel insights into the role of shared leadership in teams.

Table 2.

A Summary of Recent Empirical Studies of Shared Leadership Contexts, Antecedents, Consequences, and Moderators.

Studies	Contexts	Antecedents	Consequences	Moderators
Bergman et al. (2012)	Decision-making teams	NA	Team functioning	NA
Chiu et al. (2016)	Work teams	Leader humility	Team effectiveness	Team capability
Choi, Kim, and Kang (2017)		NA	Team’s organizing and planning effectiveness	NA
Drescher, Korsgaard, Welpe, Picot, and Wigand (2014)	Online simulation	NA	Team performance	Group trust
Erkutlu (2012)	Commercial bank teams	NA	Team proactivity	Organizational culture
Fausing et al. (2015)	Knowledge and manufacturing teams	• Empowering leadership• Task interdependence	Team performance	NA
Grille et al. (2015)	Work teams	• Fair reward• Psychological• Empowerment• Team leader behavior	NA	NA
Gu et al. (2016)	Interorganizational teams	NA	Team creativity	• Knowledge sharing• Task interdependence
Hoch (2014)	Work teams	Demographic diversity	Team performance	Knowledge sharing
Lee, Lee, and Seo (2015)	E-learning environment	NA	Team creativity	• Task variety• Task analyzability
Muethel, Gehrlein, and Hoegl (2012)	Software project teams	Team demographics	Team performance	NA
Nielsen and Daniels (2012)	Work teams	NA	Working conditions and well-being	• Group cohesion• Meaningful work• Social support• Role conflict
Rolfsen et al. (2013)	Manufacturing teams	NA	Team performance	• Team autonomy• Team function
Serban and Roberts (2016)	Student simulation teams	• Internal team environment• Task ambiguity	• Team performance• Team satisfaction	NA
L. Wang, Jiang, Liu, and Ma (2017)	Sales teams	NA	Team effectiveness	Leader-member exchange differentiation
Zhou (2016)	Entrepreneurial teams	Team diversity and composition	Team performance	NA

Note. NA = not applicable.

Theoretical Framework

Further to the analysis of Table 2, we synthesize the common antecedents to shared leadership from cumulative research; clarify, compare, and analyze the consequences or outcomes, as well as investigating the controversial moderators that would influence the relationship between shared leadership and team outcomes. In this process, we expose the key antecedents, moderators, and consequences and then generate an integrated framework of shared leadership (see Figure 1). Our framework comprises three parts. Part 1 defines antecedent variables needed for shared leadership to successfully emerge, namely, internal team environment and team characteristics. Part 2 relates to the consequent variables that symbolize the effects of shared leadership on teams. They are divided into four types based on the work of Pearce and Conger (2003, p. 354): group behavior processes, attitudinal outcomes, team cognition, and team performance. Part 3 focuses on two forms of moderators, substantive moderators (intragroup trust and task interdependence) and methodological moderators (team setting, shared leadership measure, outcomes measure, and longitudinal survey). We discuss each of these in more detail in the following sections.

Figure 1.

The integrated framework of shared leadership in teams, its antecedents, moderators, and consequences.

Antecedents

Internal Team Environment

Carson et al. (2007) proposed that the internal team environment is an essential antecedent for effective shared leadership. In their study, internal team environment comprises three elements: shared purpose, social support, and voice. They suggested that these elements can work collectively to create an atmosphere that inspires team members’ willingness to exert leadership influence. Shared purpose occurs when group members have a mutual understanding of team goals and take the required measures to focus on these common objectives. This importance has been recognized by previous researchers such as Kirkman and Rosen (1999), who proved that shared purpose heightens the level of commitment and motivation within teams and increases individuals’ willingness to participate in leadership activities. The second dimension, social support, refers to the efforts of team members to offer emotional and psychological support to one another. Van, Haynes, Borrill, and Stride (2004) suggested that the presence of high-level social support within a team creates an environment wherein group members can more easily collaborate and develop a sense of shared responsibilities for team results. The third and last dimension is voice. Voice is described as constructive change-oriented communication, participation in decision making, and involvement in key processes (Van Dyne & LePine, 1998). When group members are willing to speak up and get involved, the likelihood that they exercise leadership can increase significantly (Carson et al., 2007). Additional research has further confirmed that the internal environment of a team is positively associated with the display and development of shared leadership (see Daspit, Justice, Boyd, & Mckee, 2013; Serban & Roberts, 2016). It can therefore be argued that when shared purpose, social support, and voice exist in groups, team members are more likely to provide leadership within teams and to respond to the leadership of others, and thus potentially increase the presence of shared leadership. We therefore propose the following:

Hypothesis 1: Internal team environment is positively related to the emergence of shared leadership.

Team Characteristics

Previous researchers have examined team characteristics that influence the exhibition of shared leadership from different dimensions. For example, Muethel et al. (2012) found that the female-to-male ratio, mean age, and national diversity can affect the emergence of shared leadership. Hoch (2014) suggested age and tenure diversity as predictors of shared leadership. Moreover, Cox et al. (2003) presented proximity, size, ability, diversity, and maturity as determinants to displays of shared leadership. Combining these findings, as well as the classical work of Campion, Papper, and Medsker (1996) who considered team characteristics in terms of heterogeneity, maturity, and size, our study examines how these three components separately affect the emergence of shared leadership.

Team heterogeneity

Team heterogeneity has been operationalized in several ways, including team members’ experience and perception (Vanaelst et al., 2006); functional, educational level and skill (Hmieleski & Ensley, 2007); as well as tenure diversity (Carpenter, 2002). Our research borrows from the work of Drach-Zahavy and Somech (2002) and considers heterogeneity to be any number of team members’ differences, including personal heterogeneity (gender and age) and task-oriented heterogeneity (functional, educational, and tenure).

There are some inconsistent findings in the literature concerning the influence of team heterogeneity on shared leadership. For example, Cox et al. (2003) reported that greater diversity hinders the successful exhibition of shared leadership, as individuals with very different specialties or professional vocabularies have difficulty in collaborating. In contrast, Hoch (2014) argued that heterogeneity benefits shared leadership, due to the higher levels of nonoverlapping and nonredundant information in heterogeneous groups. Consistent with Hoch (2014), we expect that team heterogeneity fosters shared leadership. Shared leadership relies on social interaction, integration, and willingness to provide and respond to mutual influence. We believe that more diverse teams provide opportunities for shared leadership to emerge, because it contributes to increased communication, collaboration, and cohesiveness within teams (Stahl, Makela, Zander, & Maznevski, 2010). Consequently, we propose the following:

Hypothesis 2a: Team heterogeneity is positively related to the emergence of shared leadership.

Team maturity

We understand the concept of team maturity as defined by Cox et al. (2003), in terms of the level of familiarity and interaction among team members, relational norms, as well as the extent of responsibilities and consensual effectiveness within teams. Relations among individuals are more effective, and members tend to share a strong sense of responsibility toward common goals in teams that are deemed more “mature.” Shared leadership is therefore more likely to emerge where there is a high degree of team maturity (Cox et al., 2003). Moreover, Zhang, Tremaine, Fjermestad, Milewski, and O’Sullivan (2006) suggested measuring maturity in terms of a combination of team competency, team motivation, and teamwork skills. High team maturity enhances a team’s knowledge levels, skills, and experiences, as well as members’ commitment to team objectives and their ability to cooperate. These factors have the potential to collectively stimulate the exhibition of shared leadership. Thus, we posit the following:

Hypothesis 2b: Team maturity is positively related to the emergence of shared leadership.

Team size

Team size is an antecedent variable that has been largely ignored by previous researchers. In light of this, our study explores how team size affects the emergence of shared leadership in groups. Prior work has repeatedly reported that larger teams exert a negative impact on team processes. For instance, Curral, Forrester, Dawson, and West (2001) proposed that larger teams have a more ambiguous sense of objectives, lower levels of participation, lower emphasis on quality, and lower support for innovation. In contrast, smaller leadership teams are more cohesive, which allows their members to act more effectively on strategic goals (Karriker, Madden, & Katell, 2017; Weiss & Hoegl, 2015). According to Curral et al. (2001), as group size increases, the psychological distance between individuals rise. Large teams therefore become too unwieldy to enable effective interaction, and in this vein, shared leadership may be less likely to emerge. Consequently, we hypothesize that larger teams impede the exhibition of shared leadership as they put more strain on team processes. We therefore posit the following:

Hypothesis 2c: Team size is negatively related to the emergence of shared leadership.

Consequences

A number of researchers regard shared leadership as an important predictor of team outcomes. It has been reported to be positively related to successful team performance, most notably, self-report performance (Choi et al., 2017), performance measured by managers (Pearce & Sims, 2002), and performance measured by external clients (Carson et al., 2007). Several studies also point to the impact of shared leadership on objective measures of performance, such as team sales (Mehra, Smith, Dixon, & Robertson, 2006), growth in revenue (Ensley et al., 2006), and fair reward (Grille et al., 2015). Furthermore, shared leadership has been proven to positively influence team creativity (Wu & Cormican, 2016b), team proactivity (Erkutlu, 2012), and team satisfaction (Serban & Roberts, 2016). In lieu of these results, we posit that shared leadership is positively associated with team outcomes. When members provide leadership to teams and respond to the leadership of others, they exchange and share more information (Carson et al., 2007), experience higher commitment (Cox et al., 2003), and become more responsible for the decision-making processes (Bergman et al., 2012). Collectively, these consequences would benefit team outcomes. Thus, we suggest the following:

Hypothesis 3: Shared leadership is positively related to team outcomes.

Although shared leadership has been proposed to exert positive influence on team outcomes, the magnitude of these relations across multiple forms of outcomes remains unclear. In other words, shared leadership is found to have various effects on different types of outcomes. We thus conducted a comparative analysis to unpack this further. First, we categorized team outcomes into four aspects in accordance with Pearce and Conger (2003, p. 354): (a) group behavioral process, which represents specific forms of behavioral responses such as networking behaviors and problem-solving behaviors; (b) attitudinal outcomes, which refers to a topology of attitudinal effects, such as team satisfaction, social integration, and team trust; (c) team cognition, which comprises team efficacy and potency, as well as team creative processes; and (d) team performance, which includes subjective performance rated by team leaders, members themselves, or external experts, and objective indicators, such as sales and productivity. In the research of D. Wang et al. (2014), they suggested that shared leadership tends to be more strongly associated with group behavioral process and attitudinal outcomes than team performance. However, there remains a shortage of evidence to support these claims, and we therefore examine it further in an exploratory fashion and propose the following central research question:

Research Question 1: What are the differences in the strength of relationships between shared leadership and multiple team outcomes: group behavioral process, attitudinal outcomes, team cognition, and team performance?

Moderators

There are several alternatives for classifying a meta-analytic data set to estimate specific moderator effects. Our research adopts the most common method suggested by Cortina (2003) and categorizes them into substantive moderators and methodological moderators. The difference between these two moderators lies in their substantive or methodological attributes. They represent different kinds of conditions that affect the direction and strength of the relationship between shared leadership and team outcomes. As for substantive moderators, we examine the moderating role of (a) intragroup trust and (b) task interdependence in the shared leadership–team outcomes relations. In terms of methodological moderators, we investigated (a) team setting, (b) shared leadership measure, (c) outcome measure, and (d) longitudinal survey. Besides, followed by the research of D’Innocenzo et al. (2014), Leong and Fischer (2011), Nicolaides et al. (2014), and D. Wang et al. (2014), we studied our moderators by adopting different formats, where substantive moderators are presented in hypotheses format while methodological moderators are employed in an exploratory fashion with questions format. The details are discussed below.

Substantive Moderators

Intragroup trust

The literature suggests that links between shared leadership and team outcomes can be strengthened or constrained by a series of factors, such as knowledge sharing (Hoch, 2014), social support (Nielsen & Daniels, 2012), and team autonomy (Rolfsen et al., 2013). This study focuses on intragroup trust, because it is one of the most significant factors that can moderate the effectiveness of leadership in terms of yielding team performance outcomes (Casimir, Waldman, Bartram, & Yang, 2006). The theory also suggests that establishing trust among team members is one route through which changes in shared leadership bring benefits to team performance (Drescher et al., 2014). On this basis, we posit that trust is a critical mechanism in the shared leadership–team outcomes relations. Through sharing leadership roles within teams, members are willing to accept and influence one another’s behaviors toward common objectives. This provides opportunities to enhance cooperation and build trust (Bergman et al., 2012). Increases in intragroup trust subsequently foster a higher level of team outcomes (Small, 2007). Therefore, we propose the following:

Hypothesis 4: Intragroup trust moderates the relationship between shared leadership and team outcomes such that this relationship is more positive when intragroup trust is higher rather than lower.

Task interdependence

Bishop and Scott (2000) define task interdependence as the degree to which team members depend on their interactions and support from others in order to perform tasks. It has been examined to be an important variable that influences shared leadership, like Pearce (2004) who found that a high level of task interdependence optimized the benefits of shared leadership. In recent meta-analyses such as the study by Nicolaides et al. (2014), it has been reported that the shared leadership–team performance relationship is stronger under conditions of higher versus lower levels of interdependence. However, a study by D’Innocenzo et al. (2014) disagreed with this point and argued that there is no significant influence of task interdependence on this relationship. In our study, we reexamine the moderator role of task interdependence and propose that the shared leadership–team outcomes relationship is more positive with a higher degree of interdependence. This is due to research that claims that greater task interdependence enhances team members’ ego involvement (Bishop & Scott, 2000) and also creates high coordination requirements to interact in an effective manner (Liden, Erdogan, Wayne, & Sparrowe, 2006). This provides a suitable context for magnifying the positive influence of shared leadership on team achievements. Consequently, we put forward the following:

Hypothesis 5: Task interdependence moderates the relationship between shared leadership and team outcomes such that this relationship is more positive when the level of interdependence is higher rather than lower.

Methodological Moderators

Team setting

Compared with work teams, student teams are often believed to be nonrepresentative of organizational settings (Peterson, 2001). Thus, team setting has received attention from researchers who examined whether or not it affects the relationship between shared leadership and team outcomes. However, past research findings have been conflicted on this point. While D’Innocenzo et al. (2014) suggested that stronger relationships exist in the field rather than in a classroom or a lab, both Nicolaides et al. (2014) and D. Wang et al. (2014) argued that team setting has no moderating effect on these relationships. Our research focuses on solving this conflict and further investigates the moderating role of team setting in the shared leadership–team outcomes relations. We explore the question below:

Research Question 2a: Does team setting moderate the relationship between shared leadership and team outcomes?

Shared leadership measure

Previous studies have measured shared leadership in various ways. Among them, there are two common methods: aggregating techniques used by Daspit et al. (2013), Ensley et al. (2006), and Hoch (2014), as well as social network analysis employed in the work of Carson et al. (2007), and Wu and Cormican (2016b). The former examines the leadership influence from the team as a whole; the latter provides the assessment for distributed leadership patterns among team members via two measurement variables: network density and network centralization (Mehra et al., 2006). Conflicting results on these two measurement approaches concerning the effects on shared leadership–team outcomes relations also exist in prior studies. Specifically, researchers found that shared leadership measured with social network methods has a stronger relationship with team outcomes than using aggregation-based methods (D’Innocenzo et al., 2014; Nicolaides et al., 2014). However, D. Wang et al. (2014) argued that the measurement approaches have no influence on these relationships. Thus, we further examine the moderating role of shared leadership measure and ask the following question:

Research Question 2b: Does the shared leadership measurement approach moderate the relationship between shared leadership and team outcomes?

Outcome measure

Team outcomes can be measured by subjective ratings or objective indicators. According to the extant literature, both measures have their limitations. Subjective measure may suffer from certain biases, such as leniency effect or cuing effects (Martell & Levitt, 2002), and objective assessment fails to symbolize the whole outcome domain (D’Innocenzo et al., 2014). Previous meta-analyses (Nicolaides et al., 2014; D. Wang et al., 2014) asserted that shared leadership–team outcomes relations are reduced when using more objective indices. However, D’Innocenzo et al. (2014) suggested that there is no significant influence of outcome measure on these relationships. Therefore, this research aims to clarify the moderating role of outcome measure and posits the following:

Research Question 2c: Does the team outcomes measurement method moderate the relationship between shared leadership and team outcomes?

Longitudinal survey

Shared leadership has been defined as an emergent and dynamic leadership process in teams (Carson et al., 2007; Pearce, 2004). Thus, it is important to study how shared leadership changes over time. For example, Wu and Cormican (2016a) explored the evolvement process of shared leadership over a project life cycle. They recommended that “the optimal level of shared leadership appears in the early phase of a project, and when the team advances into later phases, the leadership changes and focuses on a few individuals” (p. 299). According to Gupta, Huang, and Niranjan (2010), time is an important factor that affects leadership patterns within groups. For all these reasons, we propose that the time period of a survey influences the ability to sufficiently capture leadership distribution patterns, and therefore, we suggest the following question:

Research Question 2d: Does the longitudinal survey moderate the relationship between shared leadership and team outcomes?

Method

Literature Research and Criteria for Inclusion

We conducted an extensive search (electronic and manual) for studies related to shared leadership published from 2007 to 2017 using databases such as PsycINFO, Google Scholar, Web of Science, and ProQuest Dissertation and theses databases. In this process, we used the following keywords: shared leadership, distributed leadership, collective leadership, team leadership, peer leadership, and overall leadership. We also manually researched the reference lists of existing narrative and meta-analyses, the Society for Industrial and Organizational Psychology, and the proceedings of the Academy of Management conference. This initial search yielded a total of 243 articles.

Following the advice of a study by Aguinis, Dalton, Bosco, Pierce, and Dalton (2011), we established several inclusion criteria to frame the subsequent analysis. First, studies with no antecedent or consequence data were excluded. Second, we only included studies that report sample sizes, correlations, and/or statistical results adequate to compute a correlation coefficient or effect size between shared leadership and its antecedent or consequence. Studies that only describe conceptual models (e.g., Hoch & Dulebohn, 2017; Ramthun & Matkin, 2012) and present theoretical propositions (e.g., Wu & Cormican, 2016a) were excluded. Third, given that shared leadership is a team-level phenomenon, we also excluded articles that examine shared leadership at the individual level of analysis (e.g., Drescher & Garbers, 2016). Fourth, we only included studies in an organizational or industrial setting or student studies simulating this environment. Studies that come from the educational context (e.g., Leithwood & Mascall, 2008; Terrell, 2010) were excluded, because their measurement was not exclusively concentrated on shared leadership among team members. We found that these studies also measured vertical leadership, that is, school principals and administrators (D. Wang et al., 2014), which was beyond the scope of our analysis.

At last, these filters resulted in 40 studies, pertaining to 3,019 teams. Additionally, among these 40 studies we filtered, the research of Ensley et al. (2006) have two groups of independent samples that can be adopted in our meta-analysis. Therefore, there are 41 overall effect sizes produced in total.

Coding

Each study was reviewed and independently coded by the first author and one research partner for the types of antecedent and consequent variables, as well as methodological moderators. We also independently evaluated the degree of two substantive moderators, intragroup trust and task interdependence. In detail, we rated intragroup trust according to the research of Simons and Peterson (2000) and task interdependence by Bishop and Scott (2000), in which high levels of trust (profound group-wide expectations of truthfulness, integrity, and a strong sense of shared respect for team members’ competence) and high degrees of task interdependence (members strongly rely on interactions with others to perform team tasks) are displayed. Overall, there is a 95% average interrater agreement across the study variables. In situations where discrepancies exist between two coders, we followed the approaches of Podsakoff, Bommer, Podsakoff, and MacKenzie (2006) and double checked our work and discussions to reach a consensus.

Meta-Analytic Procedures

In our meta-analysis, we extracted the Pearson correlation (r) from each individual study and used it as the key metric. As the effect sizes (correlations) have undesirable statistical properties (Lipsey & Wilson, 2001), we used Fisher’s z transformation to normalize our effect sizes followed by the research of D’Innocenzo et al. (2014) and Nicolaides et al. (2014). We then back-transformed into r to interpret and report our results because Fisher’s z is not readily interpretable (D’Innocenzo et al., 2014). In this research, we followed by random effects, meta-analytic procedures recommended by Schmidt and Hunter (2014) to calculate effect sizes between shared leadership and its antecedents as well as consequences. To present the most accurate estimates, the sample-size weighted, mean correlations and their variances were corrected for measurement errors (but not for range restriction). Specifically, we corrected each correlation for measurement error using reliability coefficients. For example, the constructs of the internal team environment and team outcomes (group behavior process outcomes, attitudinal outcomes, and team cognition) were corrected for unreliability based on ICC₂ (intraclass correlation coefficient) reported in the original study. The value of ICC₂ reflects the estimate of group-level mean scores (Combs, Liu, & Hall, 2006), which has been consistently employed in the past meta-analysis as the basis for correcting unreliability of team-level constructs (see Nicolaides et al., 2014; Seibert, Wang, & Courtright, 2011; D. Wang et al., 2014). In instances where subjective performance was measured by a single respondent (e.g., team manager or external expert), we used the Cronbach’s alpha to correct the unreliability. As not all the artefact information was reported to correct the correlation for unreliability in the individual studies, a mean reliability was used and calculated based on the available information (Schmidt & Hunter, 2014). For instance, the average reliabilities are group behavior processes = .88, attitudinal outcomes = .93, team cognition = .85, and team performance = .91. Besides, we adopted the subgroup method suggested by Schmidt and Hunter (2014) to test categorical moderators (e.g., team setting, shared leadership measure, team outcomes measure, and longitudinal survey in our research). Additionally, when multiple indicators were available from a single sample, we used linear composites to combine them into one indicator following the best practice of Schmidt and Hunter (2014). Moreover, we calculated 95% confidence intervals to judge the statistical significance of these correlations (Whitener, 1990) and 90% credibility value to evaluate the generalizability of these effects (Schmidt & Hunter, 1977). As a supplemental analysis, we also tested potential publication biases by creating contour-enhanced funnel plots (Kepes, Banks, McDaniel, & Whetzel, 2012). We found that the distribution of samples is symmetric in the plot, which means that the publication bias is likely to be absent (Kepes et al., 2012).

Results

The research results are presented in the following sections: antecedents of shared leadership (to test Hypotheses 1, 2a, 2b, and 2c), consequences of shared leadership (to test Hypotheses 3 and answer Research Question 1), as well as test of moderating effects, containing substantive moderators (to test Hypotheses 4 and 5) and methodological moderators (to answer Research Questions 2a, 2b, 2c, and 2d).

Antecedents of Shared Leadership

Internal Team Environment

Results of the meta-analysis of the internal team environment–shared leadership relationship are illustrated in Table 3. The uncorrected sample-size weighted correlation between internal team environment and shared leadership is .41. After correcting for measurement errors in both internal team environment and shared leadership, the correlation becomes ρ = .50. For this overall estimate, both 95% confidence interval and 90% credibility value excluded zero, which indicates that the mean true correlation between internal team environment and shared leadership is distinguishable from zero and is generalizable in most situations, respectively. This result thus provides a strong support for Hypothesis 1 (internal team environment is positively related to the emergence of shared leadership). In particular, shared purpose, social support, and voice are all positively associated with shared leadership (ρ = .50, .51, and .49, separately).

Table 3.

Meta-Analysis of the Antecedents of Shared Leadership.

							95% CI		90% CV
Antecedent variables	k	N	r	ρ	SD ρ	SE	Lower	Upper	Lower	Upper
Internal team environment	27	1818	.41	.50	.10	.002	.47	.56	.37	.63
Shared purpose	10	688	.41	.50	.09	.004	.44	.57	.38	.62
Social support	8	608	.42	.51	.11	.004	.44	.61	.37	.65
Voice	7	522	.40	.49	.09	.004	.42	.59	.37	.61
Team characteristics
Team heterogeneity	10	1013	.10	.11	.08	.003	.05	.17	.01	.21
Team maturity	18	1864	−.03	−.03	.16	.005	−.11	.05	−.23	.17
Team size	20	2085	−.01	−.01	.13	.003	−.08	.03	−.18	.16

Note. k = number of correlations meta-analyzed; N = sum of team numbers; r = sample-size weighted correlations, not corrected for measurement errors; ρ = sample-size weighted correlations, corrected for measurement errors; SDρ = standard deviation of ρ; SE = standard error; 95% CI = 95% confidence interval; 90% CV = 90% credibility value.

Team Characteristics

A series of team characteristics are hypothesized to be positively related to the emergence of shared leadership in Hypothesis 2. First, Hypothesis 2a posits a positive relationship between team heterogeneity and shared leadership. Our results (see Table 3) show that the corrected sample-size weighted correlation between team heterogeneity and shared leadership is ρ = .11. For this relationship, neither 95% confidence interval nor 90% credibility value includes zero, indicating support for Hypotheses 2a. Second, the relationships proposed in Hypothesis 2b (between team maturity and shared leadership, ρ = −.03) and Hypothesis 2c (between team size and shared leadership, ρ = −.01) were not supported, because both 95% confidence interval and 90% credibility value include zero. Therefore, we failed to prove that team maturity and team size can influence the exhibition of shared leadership in teams.

Consequences of Shared Leadership

Our results related to the consequences of shared leadership are illustrated in Table 4. As shown in this table, the sample-size weighted mean correlation between shared leadership and overall team outcomes is r = .31, uncorrected for measurement error. When corrected for unreliability in shared leadership and team outcomes measures, the estimated correlation is ρ = .39. Both values indicate a positive association between shared leadership and team outcomes. Moreover, 95% confidence interval and 90% credibility value do not include zero, suggesting the generality of the positive correlations across studies. Taken together, Hypothesis 3 (shared leadership is positively related to team outcomes) is fully supported. Additionally, Table 4 shows that although shared leadership demonstrates different relationships with group behavior processes (ρ = .28), attitudinal outcomes (ρ = .25), team cognition (ρ = .44), and team performance (ρ = .37), their 95% confidence intervals overlap with one another. Therefore, Research Question 1 is answered: There are no differences in the strengths of relationships between shared leadership and multiple team outcomes: group behavior processes, attitudinal outcomes, team cognition, and team performance.

Table 4.

Meta-Analysis of the Consequences and Methodological Moderators of Shared leadership.

							95% CI		90% CV
Variables	k	N	r	ρ	SDρ	SE	Lower	Upper	Lower	Upper
Consequences
Overall outcomes	49	3885	.31	.39	.17	.007	.35	.41	.17	.61
Group behavior processes	9	567	.23	.28	.16	.006	.18	.49	.08	.48
Attitudinal outcomes	3	121	.21	.25	.08	.007	.21	.51	.15	.35
Team cognition	9	538	.36	.44	.05	.002	.37	.51	.38	.50
Team performance	28	2659	.30	.37	.16	.003	.32	.49	.17	.57
Methodological moderators
Team setting
Student teams	12	726	.25	.29	.16	.006	.23	.46	.09	.49
Work teams	29	2444	.32	.39	.15	.003	.36	.51	.20	.58
Shared leadership measure
Aggregating	28	2520	.29	.35	.14	.003	.33	.40	.18	.53
Social network analysis	13	780	.36	.46	.19	.007	.45	.58	.22	.70
Network density	12	741	.37	.47	.19	.007	.46	.57	.23	.71
Network centralization	3	127	.26	.30	.02	.007	.15	.45	.27	.33
Team outcomes measure
Subjective	35	2816	.31	.38	.17	.003	.36	.51	.16	.60
Objective	8	654	.27	.32	.07	.003	.25	.39	.23	.41
Longitudinal survey
Yes	38	2919	.30	.37	.16	.003	.35	.49	.17	.57
No	3	251	.24	.28	.09	.005	.24	.49	.16	.40

Test of Moderating Effects

Substantive Moderators

There are two substantive moderators in our research, intragroup trust and task interdependence. They were proposed to exert moderating influence on shared leadership–team outcomes relations. As these two moderators are continuous variables, we employed the weighted least squared regression to test Hypothesis 4 and 5 as suggested by Steel and Kammeyer (2002). The regression results are shown in Table 5. With 41 independent samples, intragroup trust is positively related to the effect size of shared leadership–team outcomes relationship (β = .13, p < .01); thus, Hypothesis 4 (intragroup trust moderates the relationship between shared leadership and team outcomes such that this relationship is more positive when intragroup trust is higher rather than lower) is supported. Similarly, task interdependence also shows a positive association with the effect size of this relationship (β = .13, p < .01). Therefore, Hypothesis 5 (task interdependence moderates the relationship between shared leadership and team outcomes such that this relationship is more positive when the level of interdependence is higher rather than lower) is supported.

Table 5.

Weighted Least Squared (by Sample Size) Regression Analysis on Substantive Moderators.

		Shared leadership–team outcomes
Model	Variables	B	SE	Adjusted R²
1	Constant	.13*	.06	.21
	Intragroup trust	.13**	.04
2	Constant	.15*	.06	.16
	Task interdependence	.13**	.04

Note. Sample size = 41. Unstandardized coefficients are illustrated.

p < .05. **p < .01.

Methodological Moderators

Team setting

The results of the methodological moderators are reported in Table 4. It suggests that the relationship between shared leadership and team outcomes does not differ across the different team settings. Even though the corrected corrections for student teams and work teams are different (ρ = .29 and .39, respectively), their 95% confidence intervals overlap with each other. In other words, alterative team settings (student teams or work teams) appear to not result in different relationships between shared leadership and team outcomes. Thus, Research Question 2a is answered.

Shared leadership measure

As illustrated in Table 4, there is a positive relationship between shared leadership and team outcomes for all categories of shared leadership measurement approaches. However, the corrected correlations suggest that the relationship between shared leadership and team outcomes with aggregating methods is weaker (ρ = .35) than with social network analysis (ρ = .46). For such relationships, their 95% confidence intervals do not overlap over other (CI = [.33, .40] and [.45, .58], respectively), suggesting that these are two different relationships. Therefore, we can say that shared leadership measurement techniques moderate the relationship between shared leadership and team outcomes: Such a relationship will be stronger when shared leadership is measured with social network analysis versus aggregation methods. Research Question 2b is answered. Interestingly, we also find that when shared leadership is measured with social network analysis, stronger relationships are observed using network density (ρ = .47) versus network centralization (ρ = .30), where their 95% confidence intervals do not overlap.

Team outcomes measure

In terms of team outcomes measurement, the results (see Table 4) suggest that although a higher average effect size is observed when team outcomes are assessed with subjective measurement, for example, rating by team leaders, members themselves, or external experts (ρ = .38), rather than objective measurement, for example, using sales and productivity (ρ = .32), their 95% confidence intervals overlap with each other (CI = [.36, .51] and [.25, .39], respectively). This indicates that there is no difference between subjective and objective measurement on the influence of the shared leadership–team outcomes relationship. Thus, Research Question 2c is answered.

Longitudinal survey

Results in Table 4 show that studies with longitudinal survey do not exert an effect on the overall relationship between shared leadership and team outcomes. Although studies with longitudinal survey (ρ = .37) and studies without longitudinal survey (ρ = .28) present somewhat different correlations, their 95% confidence intervals overlap each other (CI = [.35, .49] and [.24, .49], respectively). This suggests that alterative survey periods (longitudinal or not) do not necessarily result in differential relationships between shared leadership and team outcomes. Research Question 2d is therefore answered.

Discussion and Future Research

This research critically synthesizes the literature in the shared leadership area and provides a systematic meta-analytic review of its antecedents, moderators, and consequences. First, our results support the claim that the internal team environment exerts positive influence on the emergence of shared leadership. This is consistent with the work of Daspit et al. (2013) and Serban and Roberts (2016). They reported that within a positive internal team environment consisting of team members who have a similar (or collective) understanding of objectives, who provide emotional and psychological strength to each other, and who actively participate in decision making, there are increased opportunities for the development of shared leadership. Regarding team characteristics (another antecedent variable), while our study did not reveal positive correlations of shared leadership with team maturity, and shared leadership with team size, we did conclude that the supports for team heterogeneity foster the exhibition of shared leadership. One possible explanation for this finding is that compared with homogeneous teams, heterogeneous teams have a greater amount of nonoverlapping knowledge and information available and such factors can enhance engagement and communication among individual members. Positive interactions within groups enables team members to provide leadership and to respond to the leadership of others (Stahl et al., 2010), therefore benefiting the overall development of shared leadership. Significantly, our article offers an in-depth analysis of potential antecedents that influence the emergence of shared leadership. We recognize, however, that in order to foster a deeper understanding, more antecedents should be identified and examined, such as team empowering (Fausing et al., 2015), task ambiguity (Serban & Roberts, 2016), team autonomy, and task variety. This should serve as a fruitful avenue for future research.

Second, our findings support claims made in previous research that there is a positive relationship between shared leadership and team outcomes. These findings are consistent with many previous studies (e.g., Chiu et al., 2016; Fausing et al., 2015; Muethel et al., 2012) and recognize that shared leadership plays a crucial role in fostering positive team outcomes. We also offer a detailed comparison concerning the nature of the relationships between shared leadership and multiple team outcomes: group behavior processes, attitudinal outcomes, team cognition, and team performance. However, our results show that there is no significant difference among these relations. This finding is different from a meta-analytic research on shared leadership conducted by D. Wang et al. (2014), who suggest that shared leadership tends to be more strongly associated with team attitudinal outcomes and group behavior processes, compared with performance measures. The reason for this conflicting result would lie in the different criteria for categorizing team outcomes and different samples used in the meta-analysis. It thus serves as a guide for future researchers seeking to investigate, analyze, and/or compare the consequences of shared leadership from different perspectives.

Third, we prove that the moderating role of intragroup trust in the shared leadership–team outcomes relationship strengthens in conditions of higher (vs. lower) levels of trust. This finding provides credence to the theory of Drescher et al. (2014). They indicated that trust, as one important facet of social functioning, is essential for team effectiveness. Drescher et al. (2014) also argued that shared leadership can improve trust; as team members influence one another and are influenced by one another, they are exposed to positive social exchanges that provide the foundation for intragroup trust. In terms of task interdependence, another moderator, our findings support the results of Nicolaides et al. (2014) who proposed that the correlation between shared leadership and team outcomes is stronger in circumstances when the degree of task interdependence is higher (vs. lower). High interdependence facilitates cooperative relations and mutual influence, which consequentially enhances shared leadership (Fausing et al., 2015). Hoch (2014) also suggested that a team process with high-levels of coordination and integration in turn fosters overall team performance. Ultimately, this study examines two kinds of substantive moderators. However, we recognize that future research endeavors should focus on additional moderating variables, such as team function (Rolfsen et al., 2013), team conflict, and team life cycle, in order to fully comprehend the shared leadership–team outcomes relationship.

Fourth, our findings provide valuable insights into methodological moderators. Among the four moderators (team setting, shared leadership measure, team outcomes measure, and longitudinal survey), we found the moderating effects of shared leadership measurement methods solely in the shared leadership–team outcomes relationship. Our results reveal that this correlation is stronger when using social network approaches than aggregation-based methods. This finding is consistent with the study by D’Innocenzo et al. (2014) who proposed that social network analysis provides a better estimate of shared leadership than aggregating techniques. One potential explanation for this is that under a social network analysis, team members are required to provide specific ratings for each other member. A richer and more informative assessment is produced when social network analysis techniques are used to analyze the ratings collected from group members, compared with an overall aggregate score. More interesting, we also found that under the social network analysis, using network density to study shared leadership exhibits a greater effect size than network centralization. This suggests that network centralization, offering a measurement of leadership distribution patterns, does not holistically capture the nuances of dyadic interactions among members to the same extent that network density does. We thus propose this as another interesting avenue for future research, because a more fine-grained analysis on the differences of shared leadership measurement approaches would greatly contribute to research in this field.

Managerial Implications

Our research also has numerous managerial implications. Most notably, our meta-analysis findings confirm the team outcome benefits of employing shared forms of leadership. We therefore suggest that organizations should focus on ways to develop strong shared leadership patterns within their groups in order to boost team outcomes. Second, our research results illuminate the specific facets of the internal team environment (shared purpose, social support, and voice) that support the development of shared leadership. Managers or leaders should thus seek to build a positive working environment, in which team members have a clear and shared sense of purpose, support one another, recognize individual and team contributions, as well as become proactively involved in decision making and constructive discussions (Carson et al., 2007). Simultaneously, our results suggest the importance of team heterogeneity and the ways in which this supports the emergence of shared leadership among group members. Third, in order to maximize the benefits of shared leadership on positive team outcomes, we encourage managers to help develop strong task interdependence and intragroup trust. Organizations should therefore carefully analyze specific team conditions to assure interdependence among tasks within each team, while taking steps to promote a cooperative climate wherein group members have an adequate opportunity to interact positively, build trust, and work toward common objectives.

Limitations and Conclusion

While our research contributes new knowledge to the concept of shared leadership, this study is not without limitations. First, even though we conducted an exhaustive search of shared leadership literature over the past 10 years, our meta-analysis is limited by the availability of empirical research. The second limitation is the diversity of antecedents. This research examines antecedents from the “team” perspective, for example, team environment and team characteristics. We acknowledge that this set of antecedent variables is not exhaustive. For example, some scholars have focused on examining the characteristics of “task” that would benefit the emergence of shared leadership, such as task interdependence (Fausing et al., 2015) and task ambiguity (Serban & Roberts, 2016). Third, in our research, we used reliability coefficients ICC₂ to correct unreliability for team-level constructs. However, for the variable, subjective performance, which was measured by a single respondent (e.g., an external expert), we used Cronbach’s alpha to correct for unreliability. Le, Schmidt, and Putka, (2009) assert that alpha coefficients only capture random response error, not scale-specific and transient error; thus, they overestimate reliability and cause underestimates of true score relationships. This must be taken into consideration in our study. Fourth, we used Fisher’s z transformation to normalize our effect sizes in the analysis followed by the research of D’Innocenzo et al. (2014) and Nicolaides et al. (2014). However, Schmidt and Hunter (2014, p. 66) presented that Fisher’s z transformation can produce an upward bias and cause less accurate results in random-effects meta-analysis models. Although they added that Fisher’s z transformation does serve its original purpose quite well (p. 220), the limitation of this transformation cannot be ignored.

In summary, this research significantly advances existing literature and debate by critically integrating the shared leadership theory in a holistic manner and providing a comprehensive, updated meta-analytic review. Our study is consistent with past meta-analyses that investigated the correlations between shared leadership and team outcomes, and we analyzed several moderating roles that would influence these relationships. However, we go further than previous research in three main ways: (a) by investigating the antecedent variables that determine the emergence of shared leadership to respond to the calls from researchers, (b) by presenting new evidence to clarify the inconsistencies reported in prior work, and (c) by developing an integrated framework to define and shape the concept of shared leadership as understood in the literature. Overall, this article contributes new academic and practical knowledge on shared leadership to benefit both organizations and scholarship in this field, and it serves as a valuable foundation to guide future research.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by National University of Ireland, Galway; National Natural Science Foundation of China (No. 71772099); Tsinghua University Research Plan Foundation (No. 2016THZWLJ03); and National Natural Science Foundation of China (No. 71421061 and 71121001).

ORCID iD

Qiong Wu

Author Biographies

Qiong Wu is a PhD candidate at National University of Ireland, Galway. Her research focuses on shared leadership and project management.

Kathryn Cormican (PhD, National University of Ireland, Galway) is a lecturer in the College of Engineering & Informatics at NUIG. Her research interests lies in the areas of entrepreneurship and technology innovation management. Kathryn leads a number of applied research projects in this area. She has published widely at international conferences and peer reviewed journals.

Guoquan Chen (PhD, Tsinghua University) is a professor of and vice-chair of the Department of Leadership and Organisation Management in the School of Economics and Management, Tsinghua University. His research areas are leadership, team learning, organizational learning, and team management. He has published more than 60 research articles in peer reviewed journals, like Journal of Management Studies and International Journal of Human Resource Management.

References

Aguinis

Dalton

D. R.

Bosco

F. A.

Pierce

C. A.

Dalton

C. M.

(2011). Meta-analytic choices and judgment calls: Implications for theory building and testing, obtained effect sizes, and scholarly impact. Journal of Management, 37, 5-38.

Balogun

Gerry

(2004). Organizational restructuring and middle manager sensemaking. Academy of Management Journal, 47, 523-549.

Barnett

R. C.

Weidenfeller

N. K.

(2016). Shared leadership and team performance. Advances in Developing Human Resources, 18, 334-351.

Bergman

J. Z.

Rentsch

J. R.

Small

E. E.

Davenport

S. W.

Bergman

S. M.

(2012). The shared leadership process in decision-making teams. Journal of Social Psychology, 152, 17-42.

Bishop

J. W.

Scott

(2000). An examination of organizational and team commitment in a self-directed team environment. Journal of Applied Psychology, 85, 439-450.

Campion

M. A.

Papper

E. M.

Medsker

G. J.

(1996). Relations between work team characteristics and effectiveness: A replication and extension. Personnel Psychology, 49, 429-452.

Carpenter

M. A.

(2002). The implications of strategy and social context for the relationship between top management team heterogeneity and firm performance. Strategic Management Journal, 23, 275-284.

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, 1217-1234.

Casimir

Waldman

D. A.

Bartram

Yang

(2006), Trust and the relationship between leadership and follower performance: Opening the black box in Australia and China. Journal of Leadership & Organizational Studies, 12(3), 68-84.

10.

Chiu

C. Y. C.

Owens

B. P.

Tesluk

P. E.

(2016). Initiating and utilizing shared leadership in teams: The role of leader humility, team proactive personality, and team performance capability. Journal of Applied Psychology, 101, 1705-1720.

11.

Choi

S. B.

Kim

Kang

S. W.

(2017). Effects of transformational and shared leadership styles on employees’ perception of team effectiveness. Social Behavior and Personality: An International Journal, 45, 377-386.

12.

Combs

Liu

Hall

(2006). How much do high-performance work practices matter? A meta-analysis of their effects on organizational performance. Personnel Psychology, 59, 501-528.

13.

Cortina

J. M.

(2003). Apples and oranges (and pears, oh my!): The search for moderators in meta-analysis. Organizational Research Methods, 6, 415-439.

14.

Cox

J. F.

Pearce

C. L.

Perry

M. L.

(2003). Toward a model of shared leadership and distributed influence in the innovation process. In Pearce

C. A.

Conger

J. A.

(Eds.), Shared leadership: Reframing the hows and whys of leadership (pp. 69-102). Thousand Oaks, CA: Sage.

15.

Curral

L. A.

Forrester

R. H.

Dawson

J. F.

West

M. A.

(2001). It’s what you do and the way that you do it: Team task, team size, and innovation-related group processes. European Journal of Work and Organizational Psychology, 10, 187-204.

16.

D’Innocenzo

Mathieu

J. E.

Kukenberger

M. R.

(2014). A meta-analysis of different forms of shared leadership–team performance relations. Journal of Management, 42, 1964-1991.

17.

Daspit

Justice

T. C.

Boyd

N. G.

Mckee

(2013). Cross-functional team effectiveness: An examination of internal team environment, shared leadership, and cohesion influences. Team Performance Management: An International Journal, 19, 34-56.

18.

Drach-Zahavy

Somech

(2002). Team heterogeneity and its relationship with team support and team effectiveness. Journal of Educational Administration, 40, 44-66.

19.

Drescher

Garbers

(2016). Shared leadership and commonality: A policy-capturing study. Leadership Quarterly, 27, 200-217.

20.

Drescher

M. A.

Korsgaard

M. A.

Welpe

I. M.

Picot

Wigand

R. T.

(2014). The dynamics of shared leadership: Building trust and enhancing performance. Journal of Applied Psychology, 99, 771-783.

21.

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, 217-231.

22.

Erkutlu

(2012). The impact of organizational culture on the relationship between shared leadership and team proactivity. Team Performance Management: An International Journal, 18, 102-119.

23.

Fausing

M. S.

Joensson

T. S.

Lewandowski

Bligh

(2015). Antecedents of shared leadership: Empowering leadership and interdependence. Leadership & Organization Development Journal, 36, 271-291.

24.

Grille

Schulte

E. M.

Kauffeld

(2015). Promoting shared leadership: A multilevel analysis investigating the role of prototypical team leader behavior, psychological empowerment, and fair rewards. Journal of Leadership & Organizational Studies, 22, 324-339.

25.

Chen

Huang

Liu

Huang

(2016). A multilevel analysis of the relationship between shared leadership and creativity in inter-organizational teams. Journal of Creative Behavior, 52, 106-129.

26.

Gupta

V. K.

Huang

Niranjan

(2010). A longitudinal examination of the relationship between team leadership and performance. Journal of Leadership & Organizational Studies, 17, 335-350.

27.

Hmieleski

K. M.

Ensley

M. D.

(2007). A contextual examination of new venture performance: Entrepreneur leadership behavior, top management team heterogeneity, and environmental dynamism. Journal of Organizational Behavior, 28, 865-889.

28.

Hoch

J. E.

(2013). Shared leadership and innovation: The role of vertical leadership and employee integrity. Journal of Business and Psychology, 28, 159-174.

29.

Hoch

J. E.

(2014). Shared leadership, diversity, and information sharing in teams. Journal of Managerial Psychology, 29, 541-564.

30.

Hoch

J. E.

Dulebohn

J. H.

(2017). Team personality composition, emergent leadership and shared leadership in virtual teams: A theoretical framework. Human Resource Management Review, 27, 678-693.

31.

Karriker

J. H.

Madden

L. T.

Katell

L. A.

(2017). Team composition, distributed leadership and performance: It’s good to share. Journal of Leadership & Organizational Studies, 24, 507-518.

32.

Kepes

Banks

G. C.

McDaniel

Whetzel

D. L.

(2012). Publication bias in the organizational sciences. Organizational Research Methods, 15, 624-662.

33.

Kirkman

B. L.

Rosen

(1999). Beyond self-management: Antecedents and consequences of team empowerment. Academy of Management Journal, 42, 58-74.

34.

Schmidt

F. L.

Putka

D. J.

(2009). The multifaceted nature of measurement artifacts and its implications for estimating construct-level relationships. Organizational Research Methods, 12, 165-200.

35.

Lee

D. S.

Lee

K. C.

Seo

Y. W.

(2015). An analysis of shared leadership, diversity, and team creativity in an e-learning environment. Computers in Human Behavior, 42, 47-56.

36.

Leithwood

Mascall

(2008). Collective leadership effects on student achievement. Educational administration quarterly, 44, 529-561.

37.

Leong

L. Y. C.

Fischer

(2011). Is transformational leadership universal? A meta-analytical investigation of multifactor leadership questionnaire means across cultures. Journal of Leadership & Organizational Studies, 18, 164-174.

38.

Liden

R. C.

Erdogan

Wayne

S. J.

Sparrowe

R. T.

(2006). Leader-member exchange, differentiation, and task interdependence: Implications for individual and group performance. Journal of Organizational Behavior, 27, 723-746.

39.

Lipsey

M. W.

Wilson

D. B.

(2001). Practical meta-analysis. Thousand Oaks, CA: Sage.

40.

Martell

R. F.

Leavitt

K. N.

(2002). Reducing the performance-cue bias in work behavior ratings: Can groups help? Journal of Applied Psychology, 87, 1032-1041.

41.

Mehra

Smith

B. R.

Dixon

A. L.

Robertson

(2006). Distributed leadership in teams: The network of leadership perceptions and team performance. Leadership Quarterly, 17, 232-245.

42.

Muethel

Gehrlein

Hoegl

(2012). Socio-demographic factors and shared leadership behaviors in dispersed teams: Implications for human resource management. Human Resource Management, 51, 525-548.

43.

Nicolaides

V. C.

LaPort

K. A.

Chen

T. R.

Tomassetti

A. J.

Weis

E. J.

Zaccaro

S. J.

Cortina

J. M.

(2014). The shared leadership of teams: A meta-analysis of proximal, distal, and moderating relationships. Leadership Quarterly, 25, 923-942.

44.

Nielsen

Daniels

(2012). Does shared and differentiated transformational leadership predict followers’ working conditions and well-being? Leadership Quarterly, 23, 383-397.

45.

Pearce

C. L.

(2004). The future of leadership: Combining vertical and shared leadership to transform knowledge work. Academy of Management Executive, 18, 47-57.

46.

Pearce

C. L.

Conger

J. A.

(2003). Shared leadership: Reframing the hows and whys of leadership. Thousand Oaks, CA: Sage.

47.

Pearce

C. L.

Sims

H. P.

(2002). Vertical versus shared leadership as predictors of the effectiveness of change management teams: An examination of aversive, directive, transactional, transformational, and empowering leader behaviors. Group Dynamics: Theory, Research, and Practice, 6, 172-197.

48.

Peterson

R. A.

(2001). On the use of college students in social science research: Insights from a second-order meta-analysis. Journal of Consumer Research, 28, 450-461.

49.

Podsakoff

P. M.

Bommer

W. H.

Podsakoff

N. P.

MacKenzie

S. B.

(2006). Relationships between leader reward and punishment behavior and subordinate attitudes, perceptions and behaviors: A meta-analytic review of existing and new research. Organizational Behavior and Human Decision Processes, 99, 113-142.

50.

Ramthun

A. J.

Matkin

G. S.

(2012). Multicultural shared leadership, a conceptual model of shared leadership in culturally diverse teams. Journal of Leadership & Organizational Studies, 19, 303-314.

51.

Rolfsen

Fausing

M. S.

Jeppe

J. H.

Jonsson

T. S.

Lewandowski

Bligh

M. C.

(2013). Moderators of shared leadership: Work function and team autonomy. Team Performance Management: An International Journal, 19, 244-262.

52.

Schmidt

F. L.

Hunter

J. E

. (1977). Development of a general solution to the problem of validity generalization. Journal of Applied Psychology, 62, 529-540.

53.

Schmidt

F. L.

Hunter

J. E.

(2014). Methods of meta-analysis: Correcting error and bias in research findings. Thousand Oaks, CA: Sage.

54.

Seibert

S. E.

Wang

Courtright

S. H

. (2011). Antecedents and consequences of psychological and team empowerment in organizations: A meta-analytic review. Journal of Applied Psychology, 96, 981-1003.

55.

Serban

Roberts

A. J.

(2016). Exploring antecedents and outcomes of shared leadership in a creative context: A mixed-methods approach. Leadership Quarterly, 27, 181-199.

56.

Shane

W. M.

Fields

(2007). Exploring the impact of shared leadership on management team member job outcomes. Baltic Journal of Management, 2, 251-272.

57.

Simons

T. L.

Peterson

R. S.

(2000). Task conflict and relationship conflict in top management teams: The pivotal role of intragroup trust. Journal of Applied Psychology, 85, 102-111.

58.

Small

E. E.

(2007). Shared leadership: A social network analysis (Doctoral dissertation). University of Tennessee, Knoxville. Retrieved from https://trace.tennessee.edu/utk_graddiss/305/

59.

Solansky

S. T.

(2008). Leadership style and team processes in self-managed teams. Journal of Leadership & Organizational Studies, 14, 332-241.

60.

Stahl

G. K.

Makela

Zander

Maznevski

M. L.

(2010). A look at the bright side of multicultural team diversity. Scandinavian Journal of Management, 26, 439-447.

61.

Steel

P. D.

Kammeyer

M. J. D.

(2002). Comparing meta-analytic moderator estimation techniques under realistic conditions. Journal of Applied Psychology, 87, 96-111.

62.

Terrell

H. P.

(2010). The relationship of the dimensions of distributed leadership in elementary schools of urban districts and student achievement (Doctoral dissertation). George Washington University, Washington, DC.

63.

Van

Haynes

Borrill

Stride

(2004). Leadership behavior and subordinate well-being. Journal of Occupational Health Psychology, 9, 165-175.

64.

Van Dyne

LePine

J. A

. (1998). Helping and voice extra-role behaviors: Evidence of construct and predictive validity. Academy of Management Journal, 41, 108-119.

65.

Vanaelst

Clarysse

Wright

Lockett

Moray

S’Jegers

(2006). Entrepreneurial team development in academic spinouts: An Examination of Team Heterogeneity. Entrepreneurship Theory and Practice, 30, 249-271.

66.

Wang

Waldman

D. A.

Zhang

(2014). A meta-analysis of shared leadership and team effectiveness. Journal of Applied Psychology, 99, 181-198.

67.

Wang

Jiang

Liu

(2017). Shared leadership and team effectiveness: The examination of LMX differentiation and servant leadership on the emergence and consequences of shared leadership. Human Performance, 30, 155-168.

68.

Weiss

Hoegl

(2015). Effects of relative team size on teams with innovative tasks: An understaffing theory perspective. Organizational Psychology Review, 6, 324-351.

69.

Whitener

E. M.

(1990). Confusion of confidence intervals and credibility intervals in meta-analysis. Journal of Applied Psychology, 75, 315-321.

70.

Cormican

(2016a). Shared leadership: An analysis of the evolvement process across the project life cycle. International Journal of Innovation, Management and Technology, 7, 299-303.

71.

Cormican

(2016b). Shared leadership and team creativity: A social network analysis in engineering design teams. Journal of Technology Management & Innovation, 11(2), 2-12.

72.

Zhang

Tremaine

Fjermestad

Milewski

O’Sullivan

(2006). Delegation in virtual team: The moderating effects of team maturity and team distance. In Computer Society. Proceedings of the 2006 IEEE International Conference on Global Software Engineering, Florianopolis, Brazil (pp. 62-68). Piscataway, NJ: IEEE.

73.

Zhou

(2016). When does shared leadership matter in entrepreneurial teams: The role of personality composition. International Entrepreneurship and Management Journal, 12, 153-169.