Effective leaders open up through dialogue : putting the power of leader vulnerability to practice with Dialogical leadership

(1)

UNIVERSITY OF AMSTERDAM

MSc Business Administration: Leadership & Management track

Master Thesis

Effective leaders open up through dialogue:

putting the power of leader vulnerability

to practice with Dialogical leadership.

Author: Sten Verpalen - 11378336 Supervisor: Dr. L. van Bunderen Date of submission: 19-06-2018, final version

(2)

Statement of originality

This document is written by Sten Verpalen who declares to take full

responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and

that no sources other than those mentioned in the text and its references have

been used in creating it.

The Faculty of Economics and Business is responsible solely for the

supervision of completion of the work, not for the contents.

(3)

Abstract

This study connects two concepts that are relatively new in leadership and management research: leader vulnerability and Dialogical leadership. Leader vulnerability appears to be very powerful for leaders because it builds follower trust, but research to date lacks practical focus on how leaders can safely reveal their weaknesses in a social context. Dialogical leadership, an emerging relational leadership technique grounded in psychology, is proposed as a precursor to displayed leader vulnerability, positively relating to perceived leader effectiveness. In addition, Leader-Member Exchange (LMX) is examined as a moderator of this relationship.

Regression analysis on multi-sourced survey data among 135 leader-employee dyads shows that while Dialogical leadership indeed positively relates to leader vulnerability, no statistical evidence is found for the conceptual model. Potential explanations for the lack of significant results are explored in the discussion section, together with theoretical and methodological limitations.

Dialogical leadership is measured for the first time using quantitative research, with a newly developed scale. The positive relation with leader vulnerability strengthens the value of the concept and gives rise to further exploration by future research. Implications and directions are discussed.

Keywords: leadership; leadership behavior; leader vulnerability; Dialogical leadership; Leader-Member-Exchange; Dialogical self

(4)

Acknowledgements

First, I would like to thank my supervisor Lisanne van Bunderen for her support and the constructive feedback throughout the research process, and Prof. Rens van Loon for his suggestions for a suitable research design and measurement scale. Also I would like to thank the students in my thesis group for their efforts in collecting the data and their helpful advice.

Writing a master thesis in combination with a full-time job would not have been possible without good support. I thank my employer, Air Traffic Control The Netherlands (LVNL), for the time that was made available. And finally, my family and friends – in particular my girlfriend Silvie – for their encouragement and patience when I was engaged in this research.

(5)

Index of figures and tables

Figure 1. Conceptual model

Figure 2. The conditions for dialogue (in contrast with debate) Table 1. Descriptive statistics and correlations

Table 2. Results regression analyses

Table 3. Results moderation analysis of LMX

Index of appendices

Appendix A Questionnaire Instructions Appendix B Questionnaire Items

(7)

1. Introduction

The hardest thing about being a leader is demonstrating or showing vulnerability. When the leader demonstrates vulnerability and sensibility and brings people together, the team wins.

Howard Schultz, founder of Starbucks, HBR (2010)

The times that leaders were successful by simply acting as a boss towards their employees have long been over. Present-day managers should go beyond the typical management tasks of goal-setting and problem-solving, as they are faced with a critical audience that expects charisma, authenticity, affect and empathy in return for commitment and motivation (Tims, Bakker & Xanthopoulou, 2001; Warrick, 2011). Indeed, research shows that leaders who allow followers to identify with them on an emotional level, by opening up and showing concern for their intrinsic needs, greatly increase employee

motivation and effectiveness (Judge, Piccolo & Ilies, 2004; Wang, Oh, Courtright & Colbert, 2011). On the one hand, opening up is a difficult step to take for most leaders. When being the person in charge, letting others see your emotions can be a terrifying perspective: “...you become vulnerable, even as the possibility for deeper connection with others is born” (Ladkin & Taylor, 2010, p. 69). On the other hand, it is exactly this vulnerability that can be very effective. The positive effects of self-exposure have been identified, with vulnerability

creating stronger, trustful relationships between leader and follower (e.g. Brower, Schoorman & Tan, 2000; Colquitt, Scott & LePine, 2007; Ladkin & Taylor, 2010).

In the last decade, leader vulnerability (here defined as: self-exposure of leaders, for instance by revealing their weaknesses) has become very popular in the work environment. A TED talk where researcher Brené Brown speaks about the power of vulnerability, has already

(8)

been viewed more than 32 million times1_{and the positive aspects of the concept have been} preached by influential business leaders, such as John Gerzema2_{(Harris Insights &}

Analytics) and Howard Schultz3_{(Starbucks). The main message is that to make real}

connection, we have to allow ourselves to be really seen. Vulnerability entails the courage of admitting imperfection, the compassion to be kind to yourself first and then to others, and connection as a result of authentically propagating this4_.

Nevertheless the negative associations with being vulnerable, such as ‘weakness’ and ‘possibility to be harmed’, seem to keep practitioners from adopting the concept in their daily management techniques.5_{It is not easy to render oneself vulnerable, because the outcomes} cannot be controlled or predicted6_{and leaders need to value “reality over rank, purpose over} power” (Bell, 2005, p. 19). Furthermore, while the higher-level construct and possible benefits seem clear, research to date lacks practical focus on how to be vulnerable as a leader in a way that it can be safely exposed by leaders and genuinely felt by followers78_{. Leaders} generally feel the need to focus on results, not on emotions (McColl-Kennedy & Anderson, 2002) and mitigate risk on the work floor to prevent harm be done to their image

(Edmondson, 2002).

Leader authenticity is often mentioned as a precursor to vulnerability (Daniel, 1998; Luthans & Avolio, 2003; Gardner, Avolio, Luthans, May & Walumbwa, 2005; Shamir &

1_{Brown, C.B. (2010, June). Brene Brown: The Power of Vulnerability [video file]. Retrieved December 28, 2017 from} https://www.ted.com/talks/brene_brown_on_vulnerability

2_{Gerzema, J., & D'Antonio, M. (2013). The Athena doctrine: How women (and the men who think like them) will rule the future. John} Wiley & Sons.

and John Gerzema: The Connection Between Vulnerability And Innovation - PSFK | Empathy Magazine | Scoop. and https://medium.com/the-girls-lounge/the-strength-of-vulnerability-fbeb881ec77f

3_{https://www.cnbc.com/2017/10/10/howard-schultz-on-the-most-undervalued-characteristic-of-great-leaders.html}

and https://www.youtube.com/watch?v=uH8E2pWOgSM

and https://www.theatlantic.com/business/archive/2012/03/is-howard-schultzs-spin-bad-for-starbucks/255003/

5_{Williams, K. (2013, July 18). The Best Leaders Are Vulnerable. Forbes. Retrieved January 21, 2018, from} https://www.forbes.com/sites/davidkwilliams/2013/07/18/the-best- leaders-are-vulnerable/

7_{Sepalla, E. (2014, December 11). What Bosses Gain by Being Vulnerable. Harvard Business Review. Retrieved December 28, 2017, from} https://hbr.org/2014/12/what-bosses-gain-by-being-vulnerable

8_{Kalman, F. (2017, May 18). Vulnerability Is a Leadership Skill, Not a Weakness. Retrieved December 28, 2017 from} http://www.talenteconomy.io/2017/05/18/vulnerability- skill-not-weakness/

(9)

Eilam, 2005; Ladkin & Taylor, 2010). Authentic behaviour means acting in accordance with one’s ‘true self’, expressing what you really think and believe and behave accordingly (Harter, 2002). By showing authentic behavior, followers identify more easily with their leaders, resulting in better leader-follower relationships (Ladkin & Taylor, 2010). This sounds logical in theory, but might be harder in practice. It is unclear how authentic behavior should be adopted by leaders, and if it is always desirable. Through authenticity, leaders might render themselves more vulnerable, but this may not be perceived as such by their followers (Claeys, 2017). ‘Being and expressing your true self’ implies behavioral integrity, conforming between words and deeds (Davis & Rothstein, 2006; Leroy, Palanski & Simons, 2012). As this is proven to be a challenge for managers (Simons, 1999), leaders who try to behave authentically, rather display desired behavior that is inconsistent with their real beliefs or feelings, provoking hypocrisy (Gardner et al., 2005, Shamir & Eilam, 2005; Cha &

Edmonson, 2006). Goffee and Jones (2006), in their paper ‘Why should anyone be led by you?’, strikingly summarize this problem: “As we counsel the executives we coach: ‘Be yourselves –more– with skill.’ There can be no advice more difficult to follow than that” (p. 58). This calls for a more pragmatic approach for leaders to actively and safely display vulnerability, and to transfer this to their followers.

In this thesis, the emerging concept of Dialogical leadership is proposed as a precursor of displayed leader vulnerability. Dialogical leadership is a relational leadership technique grounded in psychology, based on the Dialogical Self-Theory of Hubert Hermans, developed in the 1990s.According to this theory, all people fluctuate between different and even opposed perspectives in their mind (in the theory called “I-positions”). By being aware of them, one can make use of these different viewpoints, for instance by assessing a situation from different perspectives. This is called ‘inner dialogue’ (Hermans, 2004). The practice of Dialogical leadership actively takes this concept to the work floor, to ‘outer dialogue’. When

(10)

aware of the contradictory perspectives of themselves and their followers in a conversation or team meeting, Dialogical leaders will be able to conduct open and constructive dialogue, in which they open up their initial position and reconcile opposite opinions (Van Loon, 2017).

In practice, this means that Dialogical leaders actively seek constructive dialogue in the interaction with followers. They will try to create an open atmosphere where all opinions will be shared and questioned, including their own. The strength of Dialogical leadership is that it is relational: it connects the ‘self’ with the context. Being used to ambiguity by accepting different perspectives in their inner dialogue, Dialogical leaders also accept

uncertainty in the outer dialogue. They propagate that there may not be one single solution to the issue at hand.

Engaging in Dialogical leadership thus forces the leader to open up his or her initial position and beliefs. Dialogical leaders act as on a collective quest with their employees, actively asking others’ opinions instead of pushing through their own ideas. They question themselves, show authentic involvement and articulate to their employees that there is not a single right answer (Van Loon, 2017). It can be argued that by actively choosing such open levels of dialogue, leaders show their weaknesses and present themselves in a more

vulnerable way. Indeed, as Brecher (2017, p. 23) points out: “Vulnerability is not sharing every emotion, concern and doubt with everyone you meet; it’s not being emotionally fragile. It is being open to ideas other than your own, accepting uncertain states and recognizing your own limitations.”

Displayed vulnerability, in its turn, is proposed to have a positive effect on perceived effectiveness of the leader. It creates stronger, trustful relationships between leader and follower (e.g. Brower et al., 2000; Colquitt et al., 2007; Ladkin & Taylor, 2010), because the openness and the display of courage foster trust, strengthening the emotional connection with followers. As a result, an atmosphere of psychological safety is constituted on the work floor

(11)

(Edmondson, 1999), facilitating employee learning and performance (Hirak, Peng, Carmeli & Schaubroeck, 2012).

Furthermore, I expect that Leader-Member Exchange (LMX) moderates the relationship between Dialogical leadership and displayed leader vulnerability, such that Dialogical leadership leads to higher levels of displayed leader vulnerability when LMX is stronger. LMX theory describes the quality of the two-way relationship leaders have with followers (Graen & Uhl-Bien, 1991). High-level LMX leads to a more positive relationship between leader and follower; it proposes this relationship in a continuum between high-quality LMX and low-high-quality LMX, with the former leading to higher task performance and higher levels of recognition and trust (e.g. Graen & Uhl-Bien, 1995). High-quality LMX is believed to improve the impact of Dialogical leadership on leader vulnerability, since a better, more trustful relationship with their followers makes it easier for leaders to show authenticity and open themselves to their opinions and ideas.

To summarize, in this research I propose Dialogical leadership as an antecedent of leader vulnerability, because leaders who open themselves up through choosing dialogue actively render themselves vulnerable to their followers. Consequently, leader vulnerability is expected to have a positive effect on perceived leader effectiveness, through the atmosphere of psychological safety that it builds as a result of follower trust. Finally, I examine the moderating effects of LMX on the relationship between Dialogical leadership and leader vulnerability.

My research has several contributions. First, the literature on leader vulnerability is extended by testing if- and how it leads to more effective leadership, and by investigating Dialogical leadership as a practical technique to apply the more abstract concept of

vulnerability. Second, the emerging concept of Dialogical leadership is measured for the first time using quantitative research, developing and testing a measurement scale specifically for

(12)

this purpose. Third, I contribute to the scarce literature on Dialogical leadership by

introducing it in a leadership and management context. Where the psychological part of the concept has mostly been described, this research focuses on the more practical application of Dialogical leadership in daily business.

(13)

2. Theoretical Background

The concept of vulnerability in leadership is relatively new, but is quickly becoming more apparent. The term in and of itself evokes few positive associations: Oxford dictionary (n.d.) defines vulnerability as ‘the quality or state of being exposed to the possibility of being attacked or harmed, either physically or emotionally’. Psychology describes the concept as ‘the potential to be harmed’, ‘defenseless’ and ‘exposure to risk’ (Chambers, 1989; Mayer, Davis & Schoorman, 1995) where invulnerability is related to more positive terms as ‘resilience’ and ‘coping’ (Anthony, 1987). These associations have mainly to do with vulnerability that is felt by someone as a result of their actions, such as a trustor that is vulnerable to the actions of the trustee (Mayer et al., 1995; Colquitt et al., 2007).

Interestingly however, expressed vulnerability appears to be very powerful for leaders when it comes to the relationship with followers. Research shows that leader vulnerability builds better relationships with followers through its positive effects on follower trust (Mayer et al., 1995; Brown et al., 2000; Colquitt et al., 2007; Ladkin & Taylor, 2010). Trough

showing vulnerability, leaders open up to their followers, increase employee motivation and effectiveness by allowing them to identify with them on an emotional level (Judge et al., 2004; Ladkin & Taylor, 2010; Wang et al., 2011) and creating an atmosphere of

psychological safety (Edmondson, 1999).

I will first give a theoretical overview of research on vulnerability and show how the concept can lead to greater leader effectiveness as an antecedent of trust-building. It is then argued how authentic leader behaviour, which may be seen as a precursor of leader

vulnerability (Daniel, 1998; Luthans & Avolio, 2003; Gardner et al., 2005; Shamir & Eilam, 2005; Ladkin & Taylor, 2010), may not be the most suitable practice to create organizational commitment, because of its introspective nature and lack of means for practical application. Therefore, I propose dialogical leadership as a possible means for leaders to display and

(14)

apply vulnerability in a more social and interactive context, while having less concern ‘to be harmed’. Finally, Leader-Member-Exchange (LMX) is brought up as a moderator of this effect, since a better relationship between leader and follower is expected to increase leaders’ willingness to be vulnerable to them.

2.1 Vulnerability and trust

Vulnerability has to date been most researched in the larger context of ‘trust’, where it is often mentioned as the main antecedent of trust-building (e.g. Brower & Schoorman, 2000; Dietz & Den Hartog, 2006; Lapidot, Kark & Shamir, 2007; Scandura & Pellegrini, 2008). High levels of trust in a leader-follower relationship directly influence follower performance and perceived effectiveness of the leader (Brower et al., 2000; Colquitt et al., 2007; Ladkin & Taylor, 2010).

According to Lapidot and colleagues (2007), trust is not relevant to begin with if the trustor is not vulnerable to the acts of the trustee. The connectedness of the two concepts is reflected in the most accepted definition of trust: “the willingness of a party to be vulnerable to the actions of another party based on the expectation that the other will perform a

particular action important to the trustor, irrespective of the ability to monitor or control that other party” (Mayer et al., 1995, p. 712). The research of Mayer and colleagues (1995) was the first to embody the critical element of vulnerability into the definition. They argue that in the behavioural manifestation of trust in a relationship, risk-taking is inescapable. Where trust is the willingness to take risk, vulnerability is actually taking this risk. The belief of a party that they will not be harmed or put at risk by another party augments opportunistic behaviour, leaving one vulnerable when the actual risk is underestimated. Further research indeed shows a strong relationship between trust and risk-taking (e.g. Susan & Holmes, 1991, Colquitt et

(15)

al., 2007). As such, trust research sees expressed vulnerability as an active choice, taking risk to possibly gain trust of the other party.

Dietz and Den Hartog (2006) expand existing research by proposing that a measure of trust should exceed the belief of the trustworthiness of another party. Three forms of trust can be distinguished: trust as belief, trust as decision and trust as intention to act. Vulnerability plays a key role in the decision stage, defined as “a willingness to render oneself vulnerable” (p. 559). Leaders who present themselves in a more vulnerable way, actively open up to followers by accepting uncertain states and emotional fragility (Brecher, 2017). This is the point in time where the belief in the others’ devotedness is embodied and converted into actual trust, leading to trust-informed actions, such as risk-taking behaviour. When a shared belief exists that it is safe to take interpersonal risks, an atmosphere of psychological safety is created (Edmondson, 1999). Defined as “the employee's sense of being able to show and employ one's self without fear of negative consequences to self-image, status, or career” (Kahn, 1990, p. 708), employees’ perceptions of psychological safety have been found to raise voice behavior (Walumbwa & Schaubroeck, 2009) and effective learning processes of teams (Edmondson, 1999).

Concluding, actively making oneself vulnerable means taking risk. Leaders hope that the risk is payed back in trust by their followers, such that the willingness to be vulnerable to another party builds a trustful relationship. Research incorporates the display of vulnerability as the most important precedent for trust, resulting in an atmosphere of psychological safety, fostering effective working relationships and increased follower performance.

(16)

2.2 Vulnerability as a cause of leadership effectiveness

In the preceding paragraph I have identified trust as an outcome of leader vulnerability. Scholars agree that showing vulnerability makes leaders more open towards followers, allowing them to identify with them on an emotional level. The positive influencers of leader vulnerability that were identified earlier create stronger relationships between leader and follower. This way, leaders greatly increase employee motivation and effectiveness (Judge et al., 2004; Ladkin & Taylor, 2010; Wang et al., 2011). “Effective leadership—the type of leadership in which leaders are able to exert considerable incremental influence with their people and they with their leader—occurs when leaders develop mature leadership

relationships with their followers” (Graen & Uhl-Bien, 1991, p. 225). Therefore, I propose:

H1: Leader vulnerability is positively related to perceived leader effectiveness.

2.3 The challenging side of vulnerability

As a leader, being open about your mistakes and showing emotions to followers creates the possibility for a deeper relational connection, but being ‘open to attack’ is often believed to undermine authority: there is a tension between emotional reactions and behaving ‘leaderly’ to followers who are looking for guidance (Ladkin & Taylor, 2010). In a recent dissertation on the subject, this tension is labeled ‘the vulnerability paradox’: “Who is crazy enough to show vulnerability where success and results are concerned?”9_{. In his dissertation, Claeys} (2017) conducted qualitative research with numerous business leaders on the subject of vulnerability. This resulted in a process model, with four stages of manifestation of leader vulnerability by leaders: unrecognized, when the leader is vulnerable but unaware of this; felt, when the leader is aware; displayed, when the leader consciously displays vulnerability

9_{Lemaire, 2017 on: PhD dissertation J. Claeys, 2017. (Lemaire, A. (2017, September 14). Showing vulnerability as a leader... a delicate} balance. Antwerp Management School. Retrieved February 15, 2018, from https://blog.antwerpmanagementschool.be/en/showing-vulnerability-as-a-leader-a-delicate-balance)

(17)

by engaging in potentially harmful behaviour; and perceived vulnerability, when this

behaviour is actually identified by followers as leader vulnerability. Going through the stages, the potential to be harmed is increased. In research of Claeys, actively expressing

vulnerability once again is identified with risk-taking, and found a difficult step to take for managers. Vulnerability implies “that there is something of importance to be lost” (Mayer et al., 1995; p. 712). This indicates why leaders might find it difficult to apply leader

vulnerability actively and genuinely, as they believe it entails potential to be harmed.

2.4 Why authenticity is not the key to displaying vulnerability

In this paragraph, I will explain how authentic leadership research incorporates vulnerability, and why I believe that this is not the best way for leaders to render themselves vulnerable.

Vulnerability plays a central role in authentic leadership research. Authentic

behaviour means acting in accordance with one’s ‘true self’: expressing what you really think and believe and behave accordingly (Harter, 2002). Scholars agree that acting vulnerable is an essential subcomponent, or even an outcome of revealing one’s true self (Daniel, 1998; Luthans & Avolio, 2003; Gardner et al., 2005; Shamir & Eilam, 2005; Ladkin & Taylor, 2010). Leaders handling consistent to their core beliefs and displaying their true self, “admit their weaknesses and expose their vulnerability” (Shamir & Eilam, 2005, p. 401),

encouraging followers to do the same. They raise self-awareness and self-regulatory processes both within themselves and their employees, creating stronger relations between leader and follower (Avolio & Gardner 2005; Ladkin & Taylor, 2010). That is why Gardner and colleagues (2005) place authenticity in the broader spectrum of positive leadership styles as a “root construct at the basis of transformational leadership” (p. 350).

Despite the mentioned positive consequences, authenticity as a concept appears to be difficult to grasp, and even more difficult to practice. There are two reasons for this. First, a

(18)

recent critical observation in research is that leaders must actually practice what they preach. Leaders articulating a strong vision and presenting themselves confident, inconsistent with their beliefs or feelings, will have followers see right through their act (Gardner et al., 2005). Displaying desired behaviour that does not resemble inner believes provokes hypocrisy (Shamir & Eilam, 2005; Cha & Edmonson, 2006). As a result, leader behavioural integrity has been proven as a mediator of the relationship between authentic leadership and follower organizational commitment and follower performance (Davis & Rothstein, 2006; Leroy, Palanski & Simons, 2012). Maintaining behavioural integrity is very difficult, as managers find it a challenge to conform between words and deeds (Simons, 1999). This proves that real authentic behaviour is very challenging.

Second, when leaders might be able to overcome this difficulty, the next question arises: how does the concept of authenticity work in practice? Being a completely internal process, the transfer of authenticity between leader and follower may be very problematic (Gardiner, 2011). When leading only from their own point of view, leaders may have no eye for the context, such as the relationships with others (Sparrowe, 2005). Ford and Harding (2011) even argue that authentic leadership may lead to disastrous dynamics within

organizations when followers are expected to identify with a ‘self’ that is not their own, but collectively desired in the organization, resulting in losing their sense of autonomy and distinction. Hence, research is divided if being ‘your true self’ is not just another leadership style that is in fashion. Added to that are the practical challenges associated with being a role model, the difficulties of behavioural integrity and the lack of empirical evidence for the success of authentic leadership10_.

10_{Besides the mentioned criticism, the retraction of eight articles on authentic leadership from influential magazines as The Leadership}

Quarterly and Journal of Organisational Behaviour (Senior Editor, 2014) has also brought the concept of authentic leadership theory into

disrepute. Confronted with mistakes in data collection, the authors could not provide the raw data, or reproduce the results. Despite this controversiality, I believe authentic leadership embodies interesting and relevant elements that address present-day leadership challenges. The theory either incorporated or relates to constructs as vulnerability, relational transparency, trust and dialogical leadership. While the concept as a whole may be unsettled, the fundamentals provide good insights for this thesis and will be discussed as such.

(19)

Concluding, authentic behavior is believed to augment vulnerability, leading to stronger and more trustful relationships between leader and follower. However, being authentic is only effective when this is genuine, not fabricated. Behavioural integrity is a mediator of the effect, and research shows this is a difficult concept for leaders. Authentic leadership research does not explain how authenticity can be transferred to followers, raising the belief that it is easier in theory, than in practice. This fear for undermining authority when opening up to followers, combined with the lack of attention to context, open the way to a more pragmatic and relational approach to vulnerability. In this research gap I introduce Dialogical leadership as a practical application of leader vulnerability, that is still grounded in authentic exchanges and leader-follower trust, while integrating the self and the context.

2.5 Dialogical leadership as a practical approach to leader vulnerability

Previously, I hypothesized that displaying vulnerability has a positive relationship with perceived effectiveness of leaders. I also argued that leaders lack a practical approach to display this vulnerability, and authentic behaviour might not be the most practical and effective way. The emerging concept of Dialogical leadership can provide insights to managers in how to create a deeper connection in a dialogue with the inner self, but also in the communication with followers, and the organization (Van Loon, 2017), making the concept of potentially key value to successful application of leader vulnerability in leader-follower relationships. In this paragraph, I will explain what Dialogical leadership entails and how it influences displayed leader vulnerability. Also, I will give a practical example of the concept.

Dialogical leadership is a leadership style rooted in psychology, based on the Dialogical Self Theory developed by Dutch psychologists Hubert Hermans and Harry Kempen in the 1990s (Hermans & Kempen, 1993). The core contribution of the Dialogical

(20)

Self Theory is that it brings together the conception of the ‘self’ and the context, by which it gives a more social interpretation to the largely inner-directed perception of psychology. According to this theory, people adopt different roles in their mind (in the theory called ‘I-positions’), for instance ‘I in my role as student’ and ‘I in my role as amateur football player’ (Hermans, Kempen & Van Loon, 1992; Hermans & Kempen, 1993). Each role has its own voice and story, meaning that a person interprets the world from different perspectives, and these perspectives lead to different reactions. By acknowledging these different perspectives, one can take different viewpoints within the own mind. This is called ‘inner dialogue’; the perspectives are put into dialogue with each other, resulting in opening up to other opinions and possibilities, instead of just one.

From a social constructionist perspective, the self is not separated from society (Hermans, 1996, 2002; Hosking, 2011), thus being aware of these inner perspectives one can use them not only within the mind, but also between people. Hence, Dialogical leadership actively takes this challenge of dealing with different perspectives and the tensions that come with contradicting viewpoints on the work floor, to ‘outer dialogue’ (van Loon & van Dijk, 2015). When aware of the contradictory roles that participants have in a conversation or team meeting, dialogical leaders will be able to open up their initial position and reconcile opposite opinions (Van Loon, 2017). According to the theory of Dialogical leadership, a leader is “in dialogue with his environment and with himself”11_{. It is a two-way dynamic process within} one's own mind and with others.

The Dialogical approach sees leadership as a relational concept, where dialogue plays a central role. According to Van Loon, dialogue is not about exchanging existing opinions, or getting across your points of view. Instead, the starting point is that both parties in the

11_{Van Loon, R. (2010). The Dialogical leader. Developing leaders for the future. Deloitte University. Retrieved December 8, 2017, from} https://www.taosinstitute.net/Websites/taos/images/ResourcesManuscripts/van_Loon_Dialogical_Leader_version_2010.pdf

(21)

conversation don't know the answer to the problem at hand. All involved should open their initial position to mutual criticism and doubt, via other’s opinions and views: “Leaders of the future listen before they judge”12_{. Leaders become enablers, instead of directive bosses. Van} Loon emphasizes that the process of dialogue is more important than the exchange. Through creating the conditions for dialogue in their organization or team, leaders achieve a more advanced contact in the relationship with followers (Van Loon, 2003; Van Loon, 2017). To be able to conduct dialogue, leaders must suspend their own judgement and open up to the opinions of others; something that most managers find very difficult because it is their nature to come up with a solution (Van Loon, 2017).

2.6 How Dialogical leadership influences leader vulnerability

To apply outer dialogue, leaders need to open their initial positions to tensions, critique and debate, while staying true to themselves in the sense of the multiple roles they fulfil. Through creating the conditions for dialogue in their organization or team, leaders achieve a more advanced contact in the relationship with followers (Van Loon, 2003; Van Loon, 2017). From the conditions for dialogue (see: Figure 2) may be understood that Dialogical leaders have to leave their comfort zone, show real involvement, create an open-minded attitude and question themselves. They take time for self-reflection and are not afraid to show their ‘true self’ and make themselves vulnerable. They question themselves, show authentic involvement and articulate to their employees that there is not a single right answer (Wijsbek, 2009; Van Loon, 2017). This way Dialogical leaders promote an atmosphere of trust, engagement and

enthusiasm, where all opinions matter. Hence, engaging in Dialogical leadership forces

12_{Tilburg University on the book of Van Loon (2017) (Tilburg University (2016, December 09). Press release: Leiders van toekomst} luisteren voordat ze oordelen. Tilburg University. Retrieved December 09, 2017, from

(22)

leaders to open up their initial positions and beliefs and present themselves in a more vulnerable way.

The concept is relatively new: Dialogical leadership as a technique was first

mentioned five years ago. To date, the effects on perceived leader effectiveness have not been empirically researched. In this research, I propose Dialogical leadership as a possible means for leaders to display and apply vulnerability in a more social and interactive context, while having less concern to be harmed.

(23)

2.7 A practical example of Dialogical leadership13

Peter (45) is a successful CFO of a big Dutch conglomerate. He describes himself as ‘impatient’ and ‘impulsive’. Due to his entrepreneurial skills and his drive for success, he works very hard and gets good results, but he is constantly on the road and stressed. In the eyes of his followers he comes across very self-assured and arrogant, and seems to be a know-it-all who does not tolerate contradiction. In team meetings, Peter talks a lot, pushes his view through and puts pressure on his team. At home, in the countryside, Peter comes to rest. Especially when he works on the farmland with his crops and animals. There he feels

connected with the members of his family and with nature.

These two worlds were completely separated until Peter entered a coaching trajectory based on Dialogical leadership. In session, the coach motivated Peter to transfer the calming aspect of his role as a hobby farmer, to his role as CFO. This way a so-called ‘dialogical space’ was created between the two positions in which the different perspectives were exchanged. This ‘inner dialogue’ made the creation of a third, combined role possible: ‘Peter as CFO and hobby farmer’. While this may seem logical to another person, the two roles in Peter’s mind were completely separated until that point.

At the work floor, Peter started recognizing that the different perspectives and tensions in his own mind, also existed in his relationships with other people. In the coaching lessons, he was motivated to suspend his own judgement and open up to the opinions of others. From the combined position, Peter naturally went into a more relaxed way with his employees and opened himself more to their ideas and opinions. Instead of pushing his view through, he recognized conflicting perspectives and motivated his employees to bring them forward in constructive dialogue. Slowly, the team meetings changed from one-man shows were Peter simply was ‘telling’ his employees what to do, into open dialogue sessions where

(24)

opinions were shared in a safe and trustful atmosphere. When his employees brought two conflicting perspectives to the table, Peter was able to take on a third, new position, even while this angle was far away from his own point of view. This way, Peter combined ‘inner dialogue’ and ‘outer dialogue’. Peter recognized that his judgement was not always correct and, most importantly, he had the courage to voice this to his employees.

This is the basic idea of Dialogical leadership: being able to take different

perspectives on a situation, by inner dialogue between the different roles a leader fulfils, as well as by motivating followers to do the same in outer dialogue. In this thesis, I investigate the effect of choosing dialogue by leaders when they are communicating with others, for instance in a team meeting. Instead of pushing through their own ideas and opinions and distributing tasks, Dialogue is key: the leader is able to switch between different perspectives or roles, and is able to take on a third position when two conflicting perspectives are brought to the table. A Dialogical leader makes optimal use of the criteria of a constructive dialogue and creates an atmosphere where the level of conversation and problem solving is brought to a higher level (Hermans & Lyddon, 2006; Wijsbek, 2009; Van Loon, 2017). Dialogical leaders create free space within the organization to jointly think and talk about problems and challenges. They are open to self-reflection and are not afraid to show their ‘true selves’ and weaknesses (Van Loon, 2017). Concluding, leaders applying this management technique will be able to display vulnerability in an interactive context, being protected from harm in a safe environment of constructive dialogue. Therefore, I propose:

H2: Dialogical leadership positively relates to leader vulnerability.

H3: The relationship between Dialogical leadership and perceived leader effectiveness is mediated by leader vulnerability.

(25)

2.8 Leader-Member Exchange (LMX) as moderator of the relationship

In this paragraph I introduce leader-member Exchange (LMX) as a moderator of the relationship between Dialogical leadership and leader vulnerability, because I presume that the application of Dialogical leadership will lead to greater positive effects on displayed vulnerability, when there is a more intensive working relationship between leader and follower.

LMX is described as the quality of the two-way relationship leaders have with

followers (Graen & Uhl-Bien, 1991). It focuses on the dyadic relationship between leader and follower (member). Both have role expectations of the other party, for instance in terms of performance and rewards, resulting in a dynamic process of social exchange, meaning that a feeling of interdependence is created through receiving positive actions in the interaction (Bernerth, Armenakis, Feild, Giles & Walker, 2007). Leaders develop, through this process, relationships with their members that vary in quality (e.g. Wang, Law, Hackett, Wang & Chen, 2005). When a leader-follower relationship is established with high-quality exchanges, this brings benefits to both parties.

All relationships between leaders and members go through three stages: role-taking, when the member first joins the group, role-making, when members begin to work as part of the team, and routinization, when routines between members and their leader are established. In the role-taking stage, leaders categorize their members in one of two groups, often

subconsciously; the in-group or the out-group. Members belonging to the in-group are believed to be hard workers, who are loyal and trusted by their manager. This group gets more attention, support and time from their manager. As a result, they will work hard and are more likely to develop themselves. If members prove that they are incompetent or betray the trust of their manager, they are put into the out-group. They receive little attention and

(26)

difficult to escape from the out-group once a member has been categorized there (Wayne & Green, 1993; Graen & Uhl-Bien, 1991, 1995; Liden and Maslyn, 1998). LMX theory alerts leaders about their perceptions of followers and motivates them to make reasonable effort to all of them, also the ones in the out-group (e.g. Yukl, O'Donnell & Taber, 2009). According to the leader-member exchange (LMX) theory, high-quality relationships between leader and follower lead to increased follower performance through higher levels of perceived

recognition, trust (Liden and Maslyn, 1998) and followers’ organizational citizenship behaviour (OCB), meaning an employees’ voluntary commitment within an organization (Graen & Uhl-Bien, 1995; Hackett & Lapierre, 2004).

The first reason why LMX strengthens the relationship between Dialogical leadership and leader vulnerability, is because Dialogical leadership is a relational concept, of which the success depends on the quality of the relationship: it is created and sustained through

discourse (Van Loon & Van Dijk, 2015). It is a two-way dynamic process within the mind and with others (Van Loon, 2017). Therefore, when high-quality exchanges are already common sense at individual or group level, there will be a more solid foundation for open dialogue. Employees belonging to the in-group receive high levels of trust from their manager and are more valued, making it easier for them to give voice to their opinions, and for the leader to value them.

A second argument is that displaying vulnerability will feel more risk-free for leaders when relationships between leader and follower are more mature. Earlier, I showed how vulnerability is connected with taking risk (e.g. Mayer et al., 1995). The belief of a party that they will not be harmed or put at risk by another party augments opportunistic behaviour. Dialogical leaders try to foster an atmosphere of trust, engagement and enthusiasm, where all opinions matter, and they can safely show their weaknesses and question themselves. This is much easier to do in dialogue with followers belonging to the in-group, than to the out-group;

(27)

the effect of Dialogical leadership on the willingness to be vulnerable is bigger when followers are more committed and there are high levels of trust. Without this, leaders may still engage in Dialogical leadership, but it will be harder to show their weaknesses and open themselves up, resulting in less displayed vulnerability.

Therefore, I expect LMX to be a moderator of the relationship between Dialogical leadership and leader vulnerability, in a way that high-quality LMX strengthens the positive relationship between the two concepts rather than low-quality LMX:

H4: The relationship betweenDialogical leadership and leader vulnerability is moderated by leader-member exchange (LMX), such that Dialogical leadership is positively related to leader vulnerability when there is high-quality LMX, but not when there is low-quality LMX.

Finally, a moderated mediation effect is proposed for the whole conceptual model:

H5: There is a moderated mediation effect of Dialogical leadership and LMX on perceived leader effectiveness through leader vulnerability.

(28)

3. Method

3.1 Procedure

The research is of explanatory nature: to test the hypotheses, a deductive approach was chosen. Data was gathered through quantitative research among leader-follower dyads. Two surveys were composed together with four other students who research the concept of leader vulnerability; one for the leaders and one for the followers. An invitation was first sent out to either a leader or a follower, followed by detailed instructions and a survey link when they agreed to participate. The invitation also contained the request to have the survey completed by the counterpart, either their leader or one of their followers. Each participant was given an individual code that had to be entered before starting the survey. This code was also sent to their counterpart, and was later used to match the leader and follower answers.

Due to the limited time available for data collection, convenience sampling was mainly used to find respondents. Furthermore, attention was drawn to the research via the distribution of a digital flyer through social media networks of the students, which also resulted in some voluntary applications. A small incentive was offered to the participants to increase the response rate: the chance to win a movie voucher (one per 10 dyads was available).

The survey was sent out to 348 persons in total, of which 296 responses were collected, resulting in a response rate of 85%. The regular response rate in organizational individual survey research is 52,7% (Baruch & Holtom, 2008). Our particularly good response rate can be explained by the multiple reminders that were sent out, and by

personally addressing people to participate in the survey. Also, the participants themselves increased the response rate, because they often motivated the other person of the dyad to complete the questionnaire as well. 8,9% of the responses was incomplete or invalid,

(29)

important that the sample is large enough to represent a general population. The sample of 135 dyads enabled me to perform a valid analysis.

3.2 Participants

The population for this study is managers and employees of organizations in the Netherlands. The sample consisted of 135 leader-follower dyads that work at least three days per week and work together for at least two months. The gender distribution was almost equal (49,6% male, 50,4% female), which is remarkable, since the general male-female ratio for managers in The Netherlands is 76% versus 24% (Central Bureau for Statistics, 2015). The main reason for this difference is that female leaders were actively approached in convenience sampling, because three of the other students wanted to analyze the moderation effect of gender in their conceptual model. The leaders’ age ranged from 22 to 69 years (M = 43.50 SD = 11.55). Followers were mainly female (60%) and ranged in age from 18 to 66 years old (M = 34.51,

SD = 12.16). The majority of the respondents completed higher education (leaders: 78,5%,

followers: 71,4%). The number of years respondents worked in their position ranged from just a few months, to almost 42 years (in months: M = 66.39 SD = 83.74). Most dyads worked together for a reasonably long time, the average was almost 3 years (in months: M = 34.9, SD = 48,25).

3.3 Measures

Variables measured in the survey were Dialogical leadership, leader vulnerability and LMX (rated by leaders) and perceived leader effectiveness (rated by followers). All questions were recorded using a 7-point Likert scale for the sake of uniformity in measuring the variables. The scale ranged from 1 (strongly disagree) to 7 (strongly agree).

(30)

Leader vulnerability. Leader Vulnerability was measured using a 7-item scale from

Van Bunderen (2018). One of the questions is: “I NEVER show a sign of weakness to my team(s)”. This item was reverse coded. The reliability of the scale proved very good (α = .85)

Dialogical leadership. For dialogical leadership, I developed a scale myself in

cooperation with Prof. van Loon (Tilburg University), who is the founder of the concept, and my thesis supervisor. The scale is a self-measure of leaders’ application of dialogue in leader-follower communication, based on a table where dialogical conversation techniques are distinguished from normal conversation (Van Loon, 2017 p.79). This table is also represented in the theoretical background of this thesis (Figure 2).

The scale concentrates on the more practical side of Dialogical leadership, that I call ‘outer dialogue’. Leaders for example have to leave their comfort zone, show real

involvement, create an open-minded attitude and question themselves. Constructive dialogue takes time, because different perspectives should be brought to the table, many questions have to be asked and leaders should accept that there is not always one solution. As such, Van Loon emphasizes that Dialogical leadership is especially useful when analyzing ‘wicked problems’: complex dilemmas that are extremely difficult or even impossible to solve, and grow in the attempt to do this (Rittel & Webber, 1973). One of the statements of the scale is: “I like to hear and explore others’ opinions, instead of ‘pushing’ my own view”.

Participants were first asked to read a small text, in which ‘wicked problems’ were explained. They were then asked to imagine themselves, together with the follower who was their counterpart in the survey, in a team conversation or meeting, addressing such a problem. They had to indicate on a 7-point Likert scale, how likely they would choose a conversation style that fits the concept of dialogue as discussed above.

(31)

The initial scale already showed good reliability (α = .79). Analysis showed that the internal consistency could be improved substantially (more than .01) when one of the items was removed (Field, 2013). The statement corresponding to this item was: “My subordinates cannot exactly predict how I will think and behave, as this is dependent on the interaction and the conversation”. Deleting this statement can be supported by theory, as ‘unpredictability’ is not a main characteristic of choosing dialogue. Given the table on which the scale was built (figure 2), most items are about leaders questioning themselves and opening up for other opinions, while creating an open-minded atmosphere for discussion. Being unpredictable is not a necessary technique to engage in dialogue, when the leader is already opening up to other opinions. Also, the individual correlation of this statement with the total scale was only .57. With the item removed, the reliability of the Dialogical leadership scale improved to a Cronbach’s alpha of .81.

LMX. The scale used for Leader-Member Exchange is the SLMX-MDM scale by

Liden & Maslyn (1998). The scale consists of 12 items, divided in four dimensions: affection, loyalty, contribution and professional respect. An example of a question is: “I like my

subordinate very much as a person”. As the scale proved sufficiently reliable (α = .80), the answers were averaged into one score for analysis.

Perceived leader effectiveness. The scale on perceived leader effectiveness was

derived from Van Knippenberg and Van Knippenberg (2005). The scale consisted of seven items. An example of a question is: “My leader/manager motivates me to exert myself for the team”. The scale proved extremely reliable (α = .94).

Control variables. To reduce possible bias of the regression results, control variables

were included in the analysis. We controlled for leader gender, leader age and the duration of the working relationship between leader and follower.

(32)

Gender is often used as a control variable in leadership research, because there are substantial gender differences in the way leaders behave (Eagly & Johnson, 1990). It is even more applicable to this study, because I presume that emotional intelligence is an important contributing factor in the application of both Dialogical leadership and leader vulnerability. For instance, research shows that transformational leadership is found to be significantly dependent on emotional intelligence, including qualities as empathy and self-awareness (Mandell & Pherwani, 2003). These qualities are also important for Dialogical leadership and leader vulnerability. In the same study, emotional intelligence scores of females were higher than those for males. Hence, it is interesting to investigate if the results stand in the analysis when adjusted for gender.

A second control variable is age. Interestingly in research on transformational

leadership, age moderates the positive relation between transformational leadership and team performance, such that older leaders tend to be more effective. When the age of the leader comes closer to the mean of that of the followers in the team, the effects of transformational leadership decrease (Kearney, 2008). Dialogical leadership is also a positive leadership style that shares characteristics with transformational leadership. In this point of view, it is

interesting to investigate if the results persist when corrected for age of the leader.

As a third and final control variable, the length of the working relationship between leader and follower will be used as to make certain that it is the frequency of interaction rather than the length of relationship that influences how Dialogical leaders are perceived by their followers.

3.4 Predictions

As indicated before, it is expected to find a significant positive relationship between leader vulnerability and perceived leader effectiveness (H1) and a positive relationship between

(33)

Dialogical leadership and leader vulnerability (H2). Consequently, a mediated effect of Dialogical leadership on perceived leader effectiveness through leader vulnerability is expected (H3) and a moderation effect of LMX on the relationship between Dialogical leadership and leader vulnerability (H4). Finally, the whole moderated mediation model is tested at once (H5), expecting that Dialogical leadership is positively related to perceived leader effectiveness through leader vulnerability when LMX is of high quality, but not when LMX quality is low.

(34)

4. Results

This section of the thesis will present the results of the data analysis in detail. Descriptives, correlations and Cronbach’s alpha values are visualized in Table 1.

Table 1: Descriptive statistics and correlations

Variables M SD 1 2 3 4 5 6 7 1. Gender (L) 0.50 0.50 - 2. Age (L) 43.50 11.55 -.20* - 3. Working together (months) 34.90 1.00 -.01 .41** - 4. Dialogical leadership 3.74 0.54 -.15 .27** .17 (.81) 5. Leader vulnerability 5.02 0.91 .15 .22* -.00 .35** (.85) 6. Perceived leader effectiveness 5.47 1.00 .09 -.08 -.03 .15 .11 (.94) 7. LMX 5.63 0.54 .05 -.02 .08 .21* .14 .30** (.80)

Note. N = 135. Gender was coded as (0) = male, (1) = female. Reliabilities (Cronbach's α) are reported between brackets on the diagonal. * p < .05, ** p < .01.

To test the hypotheses, hierarchical regression analysis was used, while taking into account the control variables (Field 2013). To confirm the results, the regression was

replicated using Hayes’ PROCESS model for SPSS (Hayes, 2013). The results are presented in Table 2.

(35)

Table 2: Results regression analyses

Leader vulnerability Perceived leader effectiveness

Variable Step 1 Step 2 Step 3 Step 1 Step 2 Step 3 Step 4

Controls Age .31** .24** .24** -.06 -.08 -.07 -.08 Gender .21* .24** .24** .08 .08 .08 .07 Relation -.13 -.16 -.16 -.01 -.05 -.05 -.04 Dialogical leadership .34** .34** .13 .12 .11 LMX .07 .07 .27** .27** .27** Dialogical lead. x LMX .01 -.06 -.06 Leader vulnerability .04 R2 _.10 _.22 _.22 _.01 _.12 _.12 _.12 Adjusted R2 _.08 _.18 _.18 _-.01 _.08 _.08 _.07 Change in R2 _.10** _.12** _.00 _.01 _.11** _.00 _.00 Overall F 4,68 9,67 .03 .53 7.65 .45 .22 df 3, 131 2, 129 1, 128 3, 131 2, 129 1, 128 1, 127

Note. N = 135. Standardized beta coefficients are presented. * p < .05, ** p < .01.

Hypothesis 1 proposed that leader vulnerability positively relates to perceived leader effectiveness. This hypothesis was not supported (β = .04, t = .47, p = .64, Adj. R2_{= .07).}

Hypothesis 2 predicted that Dialogical leadership positively relates to leader

vulnerability. The results give support for this hypothesis (β = .34, t = 3.99, p < .001, Adj. R2

= .12). Dialogical leadership has a statistically significant positive relationship with displayed leader vulnerability14_.

14_{Confirmatory factor analysis was performed because this study made use of multiple variables that were measured from the leaders’ point} of view. Each of the items of Dialogical leadership, leader vulnerability and LMX loaded onto the corresponding factors and there was no cross loading onto the factors of other variables. Hence, the variables can be considered as separate constructs (Brown, 2014).

(36)

In Hypothesis 3, I proposed that leader vulnerability mediates the relationship between Dialogical leadership and perceived leader effectiveness. Since Hypothesis 1 was not supported, testing this hypothesis has become superfluous. However, the results are displayed here for procedural purposes. To test for mediation, PROCESS model 4 with 5000 repetitions was executed (Hayes, 2013). There was no evidence for a direct relationship between the dependent and the independent variable of the conceptual model (b = 0.17; Bias and accelerated 95% CI: -0.02, 0.09). Hence, there is no direct effect between Dialogical leadership and perceived leader effectiveness. Also, the indirect effect of Dialogical leadership on perceived leader effectiveness via leader vulnerability was statistically

insignificant (b = 0.02; Bias and accelerated 95% CI: -0.03, 0.09). Concluding, Hypothesis 3 is rejected.

Hypothesis 4 predicted LMX as a moderator of the relationship between Dialogical leadership and leader vulnerability. Moderation was tested by using regression analysis, after centralizing the independent and moderator variables, and creating a new interaction term (Field, 2013). The results are presented in Table 3, and show that there is no significant moderation effect (p = .75). Therefore, Hypothesis 4 is also rejected.

Table 3: Results moderation analysis of LMX

B SE t Sig. Moderation (Constant) 2.06 .73 2.83 .01** X: DIAL .79 .19 4.08 .00** W: LMX_C .11 .14 .80 .43 DIAL*LMX .09 .29 .32 .75 N=135 * p < .05, ** p < .01.

(37)

Finally, in Hypothesis 5 I proposed a moderated mediation model, such that Dialogical leadership is positively related to perceived leader effectiveness through leader vulnerability when LMX is of high quality, but not when LMX quality is low. PROCESS macro for SPSS (model 7) with 5000 repetitions was used to test the whole moderated mediation model at once, including control variables (Hayes, 2013). The results showed no evidence for a moderated mediation effect when LMX was low (b = 0.02; Bias and accelerated 95% CI: -0.03, 0.09), moderate (b = 0.02; Bias and accelerated 95% CI: --0.03, 0.08) or high (b = 0.02; Bias and accelerated 95% CI: --0.03, 0.09). Hence, as expected when looking at the rest of the results, hypothesis 5 is also rejected.

To summarize, no statistical evidence was found for the conceptual model, since no significant mediation or moderation effects returned out of the analysis. Hence, hypotheses 1, 3, 4 and 5 are not supported. On the other hand, evidence was found for a significant relation between Dialogical leadership and leader vulnerability.

(38)

5. Discussion

This study started with the notion that leaders of today can be more effective by opening up to followers and identify with them on an emotional level (Judge et al., 2004; Ladkin & Taylor, 2010; Wang et al., 2011). Expressing vulnerability can be beneficial to managers, building better leader-follower relationships by being an antecedent of trust. As revealing their weaknesses is a difficult step to take, especially when in a powerful position (Brecher, 2017; Claeys, 2017), the practical concept of Dialogical leadership was introduced in this research to be an antecedent of leader vulnerability. Grounded in psychology, this relational leadership technique connects leaders’ self-awareness with taking action through dialogue (Van Loon, 2017). It was theorized that through applying Dialogical leadership, leaders would render themselves more vulnerable, and consequently be perceived as more effective by their followers.

The findings of my research show that Dialogical leadership and leader vulnerability do not significantly influence perceived leader effectiveness. There are several theoretical explanations for this, which will be discussed below. Thereafter, I will also identify methodological reasons for the absence of significant results, before I discuss the positive findings that also resulted of the analysis.

First of all, the little research that exists on the subject all implies that leader vulnerability is a dynamic concept, dependent on the congruence of both the inner ‘self’ of leaders and their actual behaviour (Ladkin & Taylor, 2010; Brecher, 2017; Claeys, 2017). As a result, the vulnerability that is expressed by leaders can be very different from what is felt by their followers. It might very well be the case that the vulnerability that leaders in the survey indicated they displayed, was not actually perceived by their followers as such, and did not contribute higher discerns of effectiveness. This is in agreement with the multi-staged model of leader vulnerability by Claeys (2017): leader vulnerability requires managers to

(39)

balance between their inner emotions and the needs of the context, and therefore go through multiple stages, where only in the fourth and final stage the expressed vulnerability is actually perceived by others. Research on authenticity also emphasizes this difficulty of transferring from leader to follower (Gardiner, 2011), which seems to be valid for vulnerability as well.

A second possible explanation is the influence of contextual factors. The success of leader vulnerability may be very contingent on the environment. Claeys (2017) argues that the transformation of vulnerability into effective behaviour is dependent on environmental elements that influence followers, such as political skills and psychological safety. Political skill is the ability to persuade and influence others (Mintzberg, 1983). As most work environments are political by nature (Ahearn, Ferris, Hochwarter, Douglas & Ammeter, 2004), real effectivity of leaders is not determined by their cognitive abilities alone, but also heavily on their political skill (Spencer & Spencer, 1993). Indeed, research shows that perceptions of leaders’ political skill positively influences their perceived effectiveness (Douglas & Ammeter, 2004). Another potential contextual factor is the absence of

psychological safety. Psychological safety is a condition for one to show vulnerability and take risk, as a result of trust (Edmondson, 1999). It has been recently proven as an influencer of perceived leader effectiveness on a team level in health care and education (Edmondson, Higgins, Singer & Weiner, 2016).Given the above arguments, it might be the case that possible (lack of) political skill of leaders or perception of psychological safety in the organization of respondents influenced the results of our study in a way that the relation between leader vulnerability and perceived effectiveness of leaders was influenced.

Third, the lack of a direct effect between vulnerability and perceived effectiveness of leaders might result from the fact that the two concepts are theoretically quite distant from each other. Leader vulnerability is mostly researched in broader contexts of trust and

(40)

authenticity, building high-quality relationships. These high-quality relationships in their turn lead to measurable outcomes, such as perceived leader effectiveness and follower

performance. For instance, in line with earlier research, the data analysis of this study showed a positive correlation between LMX and perceived leader effectiveness. This finding was not further discussed, because it was not part of the conceptual model. To my knowledge,

scholars have not found direct causality between leader vulnerability and follower ratings of leader effectiveness, and this study is no exception.

There are also methodological explanations for not finding significant results. Firstly, multiple response biases can occur from the way respondents were recruited and data was collected. We used convenience sampling, a form of non-probability sampling using

respondents that were close at hand. As people are more likely to take a survey together with a colleague when the working relation is good, it is possible that the motives of the

candidates were quite alike, leading to a very homogeneous sample (Furnham, 1986). The high mean scores on LMX (5.63 / 7) and perceived leader effectiveness (5.46 / 7) in the data (table 1) ratify this explanation even more. The respondents generally rated their working relationship and effectiveness of their leader as very good. Another form of response bias is giving socially desirable and evasive answers in surveys, common in self-report research (Warner, 1965; Kalton & Schuman 1982). This is even more likely when the respondent is to evaluate a person that he or she has a close working relationship with. Although the responses were anonymous and participants were informed about this prior to taking the survey,

socially desirable answers or evasive answers are feasible.

Despite the insignificant findings for the research model as a whole, the results did reveal a significant relationship between Dialogical leadership and leader vulnerability. The direction of causality did not become clear from the data analysis, because there was no correlation with the other variables of the conceptual model. The results thus point in the

(41)

direction that there must be a third factor explaining both constructs, for example a character trait from the leader. I propose that self-esteem could be this confounding variable, and I will argue why.

To start with, it was derived from the theory that both Dialogical leadership and leader vulnerability require people to open up, question themselves and leave their comfort zone. To be able to do this, leaders need to be emotionally stable and capable. Self-esteem, defined as a person’s overall self-evaluation of his or her worth (Mehdizadeh, 2010), makes the leader believe to be capable from the ‘self’. Self-esteem is highly correlated with the Big Five personality traits (Robins, Tracy, Trzesniewski, Potter & Gosling, 2001; Wang, Jackson, Zhang & Su, 2012) and emotional stability in particular (Judge & Bono, 2001). Leaders high in self-esteem are able to connect self-awareness with the courage of taking action. Hence, self-esteem fosters openness and can be a precursor to both leader vulnerability and

Dialogical leadership. Further research could explore self-esteem in relation to Dialogical leadership and leader vulnerability, for instance by using Coopersmith's (1965) self-esteem scale. Additionally, the Big five personality factors scale (Goldberg, 1999) could be included in the research, to study general effects of personality on these constructs.

The lack of evidence for causality raises the question if Dialogical leadership indeed serves as practical leadership behaviour to express vulnerability, or if leaders that are already find it easy to render themselves vulnerable, are more willing to apply this leadership style. Although the answer to this question remains unknown, it can be argued that for leaders who do not know how to render themselves vulnerable, Dialogical leadership could be a good starting point. As it so happens, the success of leader vulnerability seems to highly depend on the interactions between leader and follower. Vulnerability can be an antecedent of trust, enabling a leader to build a better relationship with employees, but only when this is being perceived by the follower, as indicated by the model of Claeys (2017). Relational leadership