< Back to previous page


Shared Leadership – The Key to Optimal Team Functioning?

Leadership, and especially effective leadership, has inspired authors from the oldest works in literature (e.g., The Odyssey) up to recent authors (e.g., John C. Maxwell, Simon Sinek, and Brené Brown). The age-old fascination for the perfect leadership model is almost akin to the search for the holy grail. Much like the tales on this legendary quest appealed to people’s fascination, the allure of great leadership has also inspired researchers to partake in a similar ‘quest’. The present PhD thesis represents a small part of the ‘quest for leadership’, and focuses its search on aspects of shared leadership. More specifically, instead of confining our investigation to effective leadership of a sole leader (e.g., the coach or captain), the present PhD thesis aims to investigate the extent to which sharing leadership amongst various members of the team can benefit team functioning and performance. More specifically, it attempts to expand upon the existing leadership literature in three ways, represented by three sections. First, the present PhD thesis describes the direct comparison between the competence support provided by either coaches or athlete leaders (i.e., Section 1, containing 2 studies). Second, we elaborate on our findings relating to the changes that leadership structures experience over time. (i.e., Section 2, containing 1 study). Finally, the present PhD presents experimental support for a recent leadership development program: the 5R Shared Leadership Program (i.e., Section 3, containing 2 studies). This summary will briefly elaborate on this PhD thesis’ most important contributions to the extant literature.

Section 1 – The Impact of Athlete Leaders Compared with Coaches

The notion proposed by Self-Determination Theory that people who are intrinsically motivated enjoy participating in their sport, absorb information faster, and thus, perform better (Cerasoli et al., 2014), has been intensively studied by researchers. However, the bulk of these studies focus on competence support by the coach, and are correlational in design (Allen & Howe, 1998; Amorose & Horn, 2000), with the exception of the experimental study of De Muynck et al. (2017). While that study revealed that competence support (i.e., providing positive feedback and encouraging team members) by the coach is beneficial for tennis players’ competence satisfaction and intrinsic motivation, the study’s design introduced an external experimenter to provide feedback, as opposed to relying on athletes' own coach, thereby limiting the transferability of these findings. This methodological limitation also applies to experimental research investigating the impact of competence support provided by athlete leaders (Fransen et al., 2015b; 2017b). While the findings from these studies highlight how athlete leaders’ positive feedback benefits their team members (e.g., feeling more competent, more intrinsically motivated, displaying heightened team confidence, and performing better), the experiments used a research confederate as athlete leader, who was unknown to the other players. Furthermore, these previous experiments used new teams consisting of players unfamiliar with each other. Section 1 first provides the first experimental comparison of the relative impact of competence support provided by either coaches’ or athletes’ leadership (Study 1), and moves on to also resolving the ecological limitations concerning the current literature on the impact of either leader (Study 2).

In Study 1, 120 male basketball players were divided into 24 groups of five players. With the aim of increasing the ecological validity compared to previous research, each group was created from 5 players of an existing team. This study used a pretest-posttest design, including four different experimental conditions, which differed in whether competence support was provided, and if so, whether it was provided by the coach, by the athlete leader, or by both. During the study, all teams had to complete a task containing basketball-specific skills twice (i.e., dribbling, passing, free-throws, and lay-ups), once for the pretest and once for the posttest. As expected, our findings indicated how both coaches and athlete leaders have the potential to positively affect their team members’ competence satisfaction, intrinsic motivation, and performance by providing positive feedback. Furthermore, Study 1 highlighted how both coaches and athlete leaders appeared to achieve a very similar influence on their team. Moreover, when both leaders provided positive feedback, a surplus effect emerged on their team’s objective performance. However, besides the creation of teams of five players (instead of a complete team), Study 1 also introduced a research confederate as the coach for the experimental task (instead of using the actual coach).

Study 2 built upon the findings from Study 1, but further improved the ecological validity of the experiment by recruiting 18 whole existing competitive basketball teams (N = 126) and by having the actual coaches and athlete leaders of those teams providing the competence support during the experiment. Besides those changes, the design implemented by Study 2 was similar to Study 1. In line with the findings from Study 1, the findings from Study 2 highlighted how positive competence support by both coaches and athlete leaders had a positive impact on team members’ performance. However, in contrast to Study 1, Study 2’s results hinted at a differential impact of coaches and athlete leaders on team members’ objective performance. Specifically, while coaches affected both players’ speed and accuracy, athlete leaders only affected teammates’ speed. Furthermore, positive feedback provided by athlete leaders enhanced players’ feelings of competence satisfaction and intrinsic motivation, whereas coaches did not impact this motivational pathway directly in Study 2.

In conclusion, Section 1’s findings suggest that coaches and athlete leaders might be best suited to fulfill different leadership roles. More specifically, athlete leaders might be most effective as ‘motivational leaders’, whereas coaches might be better suited to focus on enhancing players’ performance (e.g., by improving shot accuracy).

Section 2 – Longitudinal Evolution of Leadership Networks in Soccer Teams

While the advantages of sharing leadership amongst team members in sport teams are broadly reported, we know little to date about how these leadership structures evolve over time. More specifically, most research that has investigated the complete sport teams’ leadership structures has used cross-sectional designs, thereby limiting our knowledge of potential fluctuations over time. The only exception, that we are aware of, is the study by Duguay et al. (2019), which tracked the evolution of leadership in a youth ice hockey team. Their findings indicated how leadership structures tend to change over time, with a slight trend to towards more shared forms of leadership. However, this study consisted of only one case study, and was not able to relate these changes in leadership structures to team effectiveness or performance due to not collecting ancillary data besides those leadership structures.

To our knowledge, Study 3 was the first study that addressed the need to gain more insight into the evolution of athlete leadership structures over time. Specifically, we recruited 20 semi-professional soccer teams (N = 460) and analyzed their leadership structures at the start of the season and then again halfway through the season. More specifically, we distinguished between four different forms of leadership, as determined by Fransen et al. (2014b): task leadership, emphasizing technical and tactical guidance; motivational leadership, focusing on motivating team members; social leadership, nurturing good relations between team members and cultivating a positive atmosphere; and external leadership, representing the team towards other stakeholders such as media, fans, and sponsors. The findings of Study 3 indicated the tendency of these leadership structures to evolve towards more shared leadership structures over the course of a season. This trend towards shared leadership could be attributed to informal leaders emerging over time, as players take on leadership roles. Furthermore, teams with the most shared leadership at the end also showed the highest outcomes relating to both team functioning and team performance.

Section 3 – Testing the effectiveness of the 5R Shared Leadership Program in Basketball Teams

Even when coaches are aware of the potential benefits of a shared leadership structure, they might still face doubts when they want to implement such a structure in their own team. These uncertainties can come in different shapes and sizes. For example, coaches might worry about how to get started on the transition towards a shared leadership structure. They might fear losing some form of control over their team. Even if they are undaunted by those challenges, or they are convinced of the benefits, they could wonder how they will implement such a leadership structure in practice. How do they know which athletes to assign as leaders? Should they differentiate between different types of leaders, and if so, which ones should they implement? Finally, coaches might be willing to develop the leadership capacities of their athlete leaders, but have no real notion or guidelines on how to achieve this. This lacuna in the extant research is highlighted by Duguay et al. (2020): “To date, previous research has primarily focused on identifying the benefits of shared athlete leadership with little attention given to how it can be developed, including the potential challenges of implementing a shared athlete leadership structure.” Section 3 aims to expand the current athlete leadership literature by providing a resource (i.e., evidence-based programming) that can support coaches in learning how to implement and maintain a structure of shared leadership.

More specifically, Section 3 experimentally investigates the effectiveness of the recently developed 5R Shared Leadership Program (5RS), which contends with the previously mentioned challenges that coaches face when transitioning to a shared leadership structure (Fransen et al., 2020b). More specifically, 5RS first uses Shared Leadership Mapping to identify a team’s best athlete leaders in a bottom-up manner. This technique employs team members’ perceptions of the leadership displayed by their team members to create a map of the team’s leadership network. Team members who appear most central in the resultant network are the ‘best’ leaders as perceived by their team members. Moreover, this mapping process does not identify team members to take on generic leadership roles, but rather identifies their perceived suitability for the four previously mentioned distinct leadership roles (task, motivational, social, and external leadership; Fransen et al., 2014b). Using the information gained by the Shared Leadership Mapping process for each of these four roles separately, the team members perceived to be the best leaders by their team members are then formally assigned to a specific leadership role. Second, after assigning athlete leaders, 5RS draws on the Social Identity Approach to Leadership to help leaders nurture teammates’ identification with their team (Haslam et al., 2017). More specifically, 5RS aims to help leaders to cultivate a shared social identity (i.e., creating a shared feeling of ‘us’ in the team, instead of having individual team members perceiving themselves as unique individuals). In order to achieve this, 5RS takes the entire team, including the appointed leaders, through five different stages, with the aim to help those leaders nurture and promote a collective sense of ‘us’: Readying, Reflecting, Realizing, Representing, and Reporting. We will describe how the present PhD thesis implemented of each of these phases in three workshops. The first workshop included the first two phases of 5RS (i.e., Readying and Reflecting). The team created their own personal ‘trademark’ by completing practical exercises to emphasize their own team’s shared identity. This trademark represented the shared norms and values of the team (e.g., a “wolfpack”, as a representation of loyalty, fierceness, and togetherness). Additionally, the shared leadership structure was introduced in the team at the end of this first workshop. The second workshop then focused on the next two phases of 5RS (i.e., Representing and Realizing), during which the team created goals in line with their team identity. The team’s athlete leaders were asked to take the lead in coordinating the process of setting goals and creating strategies to achieve these goals. Finally, the third workshop encompassed the final phase of 5RS (i.e., Reporting), which aimed to evaluate both the progress towards the team’s goals, and the effectiveness of the implemented strategies. For more information, each of these phases is described in detail in “Chapter 1: General Introduction” of this PhD thesis, at page 57.

Nevertheless, while 5RS is grounded in an extensive body of theorizing, the program’s effectiveness has only been investigated by means of qualitative data from case studies. Specifically, Fransen et al. (2020b) investigated 5RS in an administrative team belonging to a Belgian university college (N = 16) and an elite female volleyball team (N = 16). Their participants generally reported positive experiences with 5RS. However, this qualitative data collection was not intended to provide any causal evidence of the efficacy of 5RS. Section 3 thus aims to move beyond those case studies by providing experimental support for the program’s effectiveness, and thus substantiating claims about the value of 5RS.

Specifically, in Study 4, we recruited eight national level Belgian male basketball teams (N = 96), of which four teams received 5RS and four teams served as a control group. After using Shared Leadership Mapping to identify the team members who are consensually seen as providing the best leadership on a specific athlete leadership role, we implemented three workshops (as described above), each lasting about 90 minutes. In line with our expectations, the findings of Study 4 highlighted how the 5R Shared Leadership program has the ability to enhance the ability of athlete leaders to nurture a shared sense of social identity (a sense of ‘us-ness’). Furthermore, 5RS aided team members to remain motivated and committed to the team goals, and improved their self-assessed feeling of general health.

Study 5 aimed to move beyond the initial support for the effectiveness of 5RS, by addressing some of Study 4’s limitations, and further improving the experimental design. Specifically, Study 4 used a self-selection procedure for including coaches in the intervention group (i.e., based on their willingness to participate), tracked only eight teams (of which only four teams participated in 5RS), and only collected data during the second half of the competitive season. To resolve these limitations, Study 5 recruited 16 basketball teams (i.e., eight female and eight male teams; N = 170) and tracked them throughout the entire competitive season. Furthermore, to address the issue of self-selection, we adopted a waitlist-control design including a homogeneous sample consisting of only teams whose coaches explicitly agreed to participate in the complete 5RS program. We then randomly assigned those teams to either the intervention group (receiving the intervention in the first half of the season) or the wait-list control group (receiving the intervention in the second half of the season). Study 5 also implemented a train-the-trainer approach to deliver 5RS, as opposed to Study 4, which used an expert with a strong theoretical background in the underpinning literature of the program to deliver 5RS.

 In line with our expectations, and the findings of Study 4, the results of Study 5 indicated that the 5R Shared leadership Program appeared to be beneficial not only for developing high-quality leadership in sport teams but also for improving team functioning, and nurturing players’ health. More importantly, Study 5 highlighted the potential of 5RS while adopting a train-the-trainer approach, opening the possibility of wider application by both researchers and practitioners. The findings of Study also suggested that 5RS achieves these benefits, regardless of the intervention’s timing during the season or the team’s gender. In conclusion, Section 3 advances the present literature on in-group leadership development, and offers practitioners direction on how and when to apply 5RS with the aim of improving team functioning and players’ health.


The present PhD thesis provides several conceptual contributes to the current literature on shared leadership in sports. First, across all studies included in this thesis, our findings highlight how athlete leaders provide a viable and valuable source of leadership for both the team’s functioning and team members’ well-being.

Second, the present PhD thesis also provides some early insight into differences between coaches and athlete leaders. While the impact of coaches and athlete leaders on their team members’ intrinsic motivation and performance were generally found to be similar in Study 1 and Study 2, these studies also highlighted how coaches and athlete leaders affect their team through different ‘pathways’. Specifically, coaches might be better at improving their team’s shot accuracy (e.g., due to coaches being formally schooled on in-depth technical advice), whereas athlete leaders might be better at enhancing their team members’ intrinsic feelings of motivation. Moreover, the findings of Study 3 indicated how task, motivational, and social leadership can more easily be extended to the player group, whereas external leadership was more restricted to coaches (e.g., due to their stronger link with the stakeholders around the team, such as the club board and media).

A third important contribution of the present PhD thesis is the insight in how leadership structures evolve over time. In line with findings on the evolution of leadership structures in self-managing teams (Small & Rentsch, 2010; Smith et al., 2018) and university student project teams (Fransen et al., 2016b), Study 3 highlighted how leadership in sport teams tends to evolve towards more shared forms of leadership over the competitive season, and that this change is driven by informal leaders emerging from the team to form an additional source of leadership.

The fourth and final important contribution of this PhD thesis is its work regarding training programs to develop athlete leadership. Early in this PhD project, Study 1 and Study 2 already hinted at the possibility of enhancing athlete leaders’ ability to lead by instructing these athletes to provide competence supportive feedback. However, while the intervention used in these studies only required minimal effort over a period of a few minutes, these experiments were typically completed during one regular practice session, and could thus only speak to short-term outcomes. In contrast, Study 4 and Study 5 examined 5RS as a viable tool for long-term athlete leadership development. The findings of both these studies suggested that 5RS has the capacity to strengthen the leaders’ ability to nurture a shared sense of social identity in their team, as well as to improve general team functioning (e.g., social support, intrinsic motivation, team confidence) and positively affect individual team members’ well-being (e.g., perceived health and burnout). Additionally, Study 5 adds to these findings that 5RS maintains its beneficial potential in a train-the-trainer approach, regardless of when the program is delivered during the season (i.e., first or second half) or the team’s gender.

In conclusion, the present PhD thesis offers a valuable contribution to the ‘quest’ for optimal leadership in sport teams in several ways. First, our findings highlight the similarities and differences in competence support provided by coaches and athlete leaders. Second, the present PhD indicates how the emergence of informal leaders is a primary driver of the changes leadership structures experience over time. Finally, we pioneered the experimental support for the 5R Shared Leadership Program as a tool to implement a shared leadership structure and develop athlete leadership. We hope that this research endeavor will inspire both researchers and aid practitioners to see the athletes as a possible key to help unlock the leadership potential of teams. Specifically, this PhD is part of a larger line of research, aiming to help teams (business and sport teams) to implement a structure of shared leadership, and develop their leadership potential. For more information, please visit: www.leadinginsights.be/en.

Date:1 Oct 2016 →  8 Dec 2020
Keywords:Leadership, Social Identity, Athlete leader, Leadership Development, Athlete leadership, Peer leadership, In-group leadership, Social network analysis
Disciplines:Human movement and sports sciences, Sports psychology
Project type:PhD project