INFLUENCE OF THE INTERACTIVE EFFECT OF MANAGERS’ POWER AND STABILITY ON TRANSFORMATIONAL LEADERSHIP BEHAVIOUR AND SUB- ORDINATES’ SUPPORT FOR CHANGE INITIATIVES

(1)

INFLUENCE OF THE INTERACTIVE EFFECT OF MANAGERS’ POWER AND STABILITY ON TRANSFORMATIONAL LEADERSHIP BEHAVIOUR AND

SUB-ORDINATES’ SUPPORT FOR CHANGE INITIATIVES

Master Thesis, Msc Human Resource Management Msc BA, specialisation Change Management

University of Groningen, Faculty of Economics and Business

April 19, 2015

REMCO VAN LAAR

Studentnumber: 1858556 Aquamarijnstraat 175 9743 PD Groningen tel: +31 (0)6-50806350 e-mail: r.van.laar.1@student.rug.nl Supervisor at university: 1st supervisor dr. J. (Jennifer) Jordan 2nd supervisor

(2)

INFLUENCE OF THE INTERACTIVE EFFECT OF MANAGERS’ POWER AND STABILITY ON TRANSFORMATIONAL LEADERSHIP BEHAVIOUR AND

SUB-ORDINATES’ SUPPORT FOR CHANGE INITIATIVES

ABSTRACT

Mangers try to implement changes successfully and support for these changes from their sub-ordinates is a prerequisite for that. This research explores how transformational leadership behaviour influences subordinates’ support for change and how this leadership behaviour is influenced by the power of a manager and the stability of that power. By drawing on studies from various literature streams, this research proposes that power and stability interact to af-fect stress, transformational leadership behaviour subordinates’ change support such that those managers with unstable power show more stress, resulting in less transformational lead-ership behaviour which leads to less change support from subordinates compared to stable power. Even though the hypotheses were not supported, a significant relationship was found suggesting that stable power in certain situations leads to less transformational leadership compared to unstable power. The discussion of this research provides explanations for this and other findings and gives clear suggestions for future research.

Keywords: power, stability, stress, transformational leadership behaviour, subordinates’

change support, manager-subordinate relationship.

ACKNOWLEDGEMENT

(3)

TABLE OF CONTENT

1. Introduction ... 4

1.1 Change Support ... 4

1.2 Variables and Conceptual Model ... 5

1.3 Theoretical and Practical Relevance ... 7

1.4 Outline of Research ... 8

2. Theory ... 9

2.1 Power ... 9

2.2 Stability of Power ... 10

2.3 Transformational Leadership Behaviour ... 11

2.4 The Interactive Effects of Power and Stability on Stress and Transformational Leadership Behaviour ... 13

2.5 Change Support ... 15

2.6 Transformational Leadership Behaviour ... 16

2.7 Relationship between Transformational Leadership Behaviour and Change Support .. 17

2.7.1 Identifying and Articulating a Vision. ... 18

2.7.2 Providing an Appropriate Model. ... 18

2.7.3 Fostering the Acceptance of Group Goals. ... 18

2.7.4 High Performance Expectations. ... 19

2.7.5 Providing Individualized Support. ... 19

2.7.6 Intellectual Stimulation. ... 20

3. Methodology ... 21

3.1 Research Design ... 21

3.2 Population and Sample ... 21

3.3 Data Collection ... 22 3.4 Measures ... 22 3.4.1 Variables. ... 22 3.4.2 Control Variables. ... 24 3.5 Data Analysis ... 24 4. Results ... 26 4.1 Data Reduction ... 26

4.2 Descriptive Statistics and Correlations ... 30

4.3 Hypotheses... 31

5. Discussion ... 38

5.1 Overview of Research and Contributions ... 38

5.2 Summary of Findings and Theoretical Implications ... 39

5.2 Limitations and Further Research ... 41

5.4 Practical Implications ... 42

6. Conclusion ... 43

7. References ... 44

Appendix A: Overview of Questionnaire Items ... 54

(4)

1. INTRODUCTION

This part of the research introduces an issue of interest to research and proposes a model with five variables to address this issue. The conceptual model is shortly introduced and the contri-butions of this research to literature and practice are outlined. This section is completed with an overview of this research.

1.1 Change Support

Organizational changes can take many forms (Burnes, 2009). However, success is not guaran-teed and estimations about projects that fail are as high as two-thirds (Beer & Nohria, 2000). It is important for organisations to understand why so many change initiatives do not reach the intended results (e.g. Burnes, 2005; Ford, Ford, & D'Amelio, 2008). A part of the explana-tion is that the role of subordinates in the implementaexplana-tion process is often underestimated (e.g. Armenakis, Harris, & Mossholder, 1993) and that change can only be successful when subor-dinates integrate the change in their practices (e.g. Jones, Jimmieson, & Griffiths, 2005) or is at least strongly dependent on subordinates’ support for the change (e.g. Choi, 2011). This research focuses on conditions under which subordinates do so.

(5)

1.2 Variables and Conceptual Model

In an organisational context, managers use their power over subordinates to reach the goals of the organisation and implement specific changes. Having power however, is not without con-sequence. Power is likely to change the behaviour of managers towards their subordinates, perceived as leadership behaviour. Some suggest that having power leads to exhibiting more anti-social behaviour (Keltner, Gruenfeld, & Anderson, 2003), some present evidence of more pro-social behaviour (Overbeck & Park, 2001, in Galinsky, Gruenfeld, & Magee, 2003), and some suggest that having power can lead to either anti- or pro-social behaviour, depending on the situation (Galinsky et al., 2003). Some aspects of the relationship between power and leadership behaviour of managers are not well understood and there is no consensus yet about the nature of the relationship which suggests the possibility of a moderator (Maner & Mead, 2010). The stability of the managers’ power seems to have the potential to shed some light on this relationship.

Even though power is usually conceptualised as a static concept (e.g. Fast, Gruenfeld, Sivanathan, & Galinsky, 2009; Galinsky et al., 2003) within organisations and in a context of organisational change, power is seldom stable. Power can be viewed as stable and not likely to change or unstable and likely to change (e.g. Jordan et al., 2011) and from this it follows that power and stability seem related and probably influence leadership behaviour together. In line with other studies (e.g. Jordan et al., 2011; Sherman et al., 2012; Williams, 1997), this research proposes that power and stability interact to affect leadership behaviour of managers.

In this research, the influence of this interaction effect on transformational leadership behaviour is studied. Six types of behaviour of managers are identified as transformational: Identifying and Articulating a Vision, Providing an Appropriate Model, Fostering the Accep-tance of Group Goals, High Performance Expectations, Providing Individualised Support, and Intellectual Stimulation (Podsakoff, MacKenzie, Moorman, & Fetter, 1990). A central aspect of these six transformational leadership behaviours, is their focus on a positive manager-subordinate relationship which is evident from their close link to positive relational behav-iours that focus on enhancing the quality of and satisfaction with a relationship (Gable, Gon-zaga, & Strachman, 2006) and include assurances to maintain the relationship (Brandau-Brown & Ragsdale, 2008).

(6)

al., 2003; Maner, Gailliot, Butz, & Peruche, 2007) and are highly motivated to retain their positions of power when these are threatened (i.e. are unstable) (Bugental, 2000; Maslow, 1937). As a result, research across different settings, shows that those with unstable power experience more stress than those with stable power (Manuck, Marsland, Kaplan, & Williams, 1995; Sapolsky & Share, 1994; Scheepers & Ellemers, 2005).

Moreover, academic literature on the behavioural effect of experienced stress is fairly consistent. Stress leads to anti-social behaviour and causes, especially for close relationships, withdrawn behaviour, anger and hostility towards the other party that harm relationships and lower their overall quality (Bodenmann, Meuwly, Bradbury, Gmelch, & Ledermann, 2010; Buck & Neff, 2012; Lewandowski, Mattingly, & Pedreiro, 2014; Smith & Carlson, 1997). It is likely that stress also negatively influences the manager-subordinate relationship which can be intensive, long lasting and build on liking, loyalty and professional respect (van Dam, Oreg, & Schyns, 2008); similar to high-quality relationships as described by leader-member exchange theory (Graen, 2004; Graen & Uhl-Bien, 1995; van Dam et al., 2008).

Viewing the exchange between managers and subordinates as a high-quality relation-ship, this research proposes that managers who experience stress will behave more negatively towards their subordinates and managers who do not experience stress will behave more posi-tively towards their subordinates. Since transformational leadership behaviour is similar to positive relational behaviours, this research predicts that subordinates view their managers as less transformational when they behave more negatively towards subordinates as a result of experienced stress.

Now, the relationship between transformational leadership behaviour and subordinates’ support for change will be addressed. Research on the success of change implementation has identified support for change as an important success factor (e.g. Choi, 2011). Support for change is a positive judgement about a specific change resulting from several attitudinal con-structs of subordinates (Choi, 2011). Attitudes of subordinates towards change are shaped by situational variables like leadership behaviour (Choi, 2011; Whelan-Berry, Gordon, & Hin-ings, 2003).

(7)

little empirical evidence is available about the influence of transformational leadership behav-iour on subordinates’ support for specific changes (Burke, 2010; Herold et al., 2008).

The organisational change approach addresses the effectiveness of more specific change behaviours but does not relate them to broader theories of leadership (Herold et al., 2008; Kotter, 1996). Moreover, the two approaches have not yet been sufficiently integrated (Bass & Riggio, 2006; Burke, 2010). Some attempts are made to integrate these two approaches (e.g. Carter et al., 2012; Groves, 2005) and link transformational leadership behaviour of managers during change to their subordinates’ commitment to those changes (e.g. Bycio, Hackett, & Allen, 1995; Herold et al., 2008; Seo et al., 2012; Yu, Leithwood, & Jantzi, 2002).

The current research takes a similar perspective, but links transformational leadership behaviour of managers to subordinates’ support for change initiatives. When the effects of transformational leadership behaviour on subordinates are analysed, it becomes clear that these behaviours that can be typified as considerate, supportive and positive create subordi-nates’ buy-in to the change and develop a sense of ownership (Abrell-Vogel & Rowold, 2014; Groves, 2005). The willingness to be part of the change is enhanced and subordinates actively involve themselves in the change process, resulting in positive attitudes towards the change a high intention to support it (Carter et al., 2012). In conclusion, this research proposes that transformational leadership behaviour of mangers has the potential to positively influence subordinates’ support for a specific change initiative.

FIGURE 1: Conceptual model

1.3 Theoretical and Practical Relevance

(8)

behaviour that encompasses different attitudinal constructs of subordinates and in doing so advances our understanding as suggested by Choi (2011). Thirdly, this research contributes to the literature by bridging different research streams. First of all, literature on power and lead-ership is brought together and the concept of power is broadened by conceptualising it as ei-ther stable or unstable. Furei-thermore, literature from both leadership style research and organ-isational change literature is used to bridge the gap between transformational leadership be-haviour and support for specific changes. Overall, this research shows that the cross-level context of the manager-subordinate relationship can be a means to relate antecedents of lead-ership behaviour to the downstream effects of this behaviour.

From a practical point of view, this research is relevant because it adds to the under-standing of how stability interacts with power, especially since the effects might interfere with the wish of the organisation to change continuously. This is important since organisations are becoming more and more flexible, reacting to an increasing number of stimuli from the envi-ronment and trying to change continuously. Furthermore, the influence of transformational leadership behaviour on subordinates’ change support has important implications for those who can influence leadership behaviour of managers during change (i.e. consultants or change agents) or managers themselves who can consciously try to change their leadership behaviour in order to create more support for their change initiatives.

1.4 Outline of Research

(9)

2. THEORY

This theoretical framework will discuss both the influence of power and stability of that power on managers’ leadership behaviour through a stress reaction and the effects of manag-ers’ leadership behaviour towards subordinates’ support for change initiatives. It is proposed that transformational leadership has the potential to increase subordinates support for change initiatives and therefore will this research focus on transformational leadership behaviour of managers.

2.1 Power

This research, focusing on the manager-subordinate relationship in organisations, will use the one-dimensional view of power1 (Lukes, 1974, p. 11). In the organisational context of this reseach, power is exerted by managers as authority in the name of the organisation by indi-viduals over other indiindi-viduals, based upon their relative position in the organisational hierar-chy (Jones, 2010, p58, 60). In line with this reasoning, power is defined as the ability to pro-vide and withhold valued resources or administer punishments (Emerson, 1962; Fiske, 1993). Managers are supposed to use their power to manage subordinates and make necessary changes in the status quo to reach the goals of the organisation, which can include business as usual or specific change initiatives that managers need or want to implement. Having power however, is not without consequences. Power is likely to change the behaviour of managers towards their subordinates. In the organisational context of this research with a focus on the manager-subordinate relationship, the behaviour of managers towards their subordinates is perceived as leadership behaviour.

Research about the influence of power on leadership behaviour suggests both anti- and pro-social behavioural effects. Having power over others might trigger approach behaviour where one is more focussed on oneself, tries to influence others more for their own ends, de-valuates others opinions and keeps a psychological distance from the one’s with less power (Keltner et al., 2003; Kipnis, 1972). If this would be the case, the power difference of manag-ers over their subordinates influences their behaviour towards subordinates in a way that sup-ports the ends of the manager and does not takes into account on the needs of, nor the rela-tionships with subordinates.

1

(10)

On the other hand, having power over others is also associated with being goal-oriented and taking action to solve social dilemmas (Galinsky et al., 2003); in situations where the re-sponsibility of the welfare of individuals is at stake, those individuals with power over others will be likely to actively associate with and personalise to them (Brewer, 1988; Overbeck & Park, 2001, in Galinsky et al., 2003). More in general, when an individual with power over someone else feels responsible for that person, he or she is likely to assist this person (Brewer, 1988; Overbeck & Park, 2001, in Galinsky et al., 2003). These effects of having power are pro-social and focus on working together to get things done or looking out for those with less power. If having power would lead to these effects, managers would use their power over subordinates to work towards organisational goals and implement changes. Furthermore, managers have a responsibility for their subordinates that can result in exhibiting behaviours to assist them or work together with them, investing in the relationship with them.

Even though the relationship between power and behaviour has been hypothesised by previous research, aspects of this relationship, like factors that influence managers to use their power for themselves (i.e. anti-social) or for the group (i.e. pro-social) are not well understood (Maner & Mead, 2010) and there is no consensus yet about the nature of this relationship. Some suggest that having power leads to exhibiting more anti-social behaviour (Keltner et al., 2003), some present evidence of more pro-social behaviour (Overbeck & Park, 2001, in Ga-linsky et al., 2003), and some suggest that having power can lead to either anti- or pro-social behaviour, depending on the situation (Galinsky et al., 2003). This lack of consensus suggests that there might be a moderator in the relationship between power and leadership behaviour of managers (Maner & Mead, 2010). The stability of the managers’ power seems to have the potential to shed some light on this relationship.

2.2 Stability of Power

(11)

Following this definition and the conceptualisation of power in this research, power and stability seem to be related. Other studies have suggested this relationship between power and stability as well. Jordan et al. (2011) propose that power and stability interact to affect risk-taking behaviour through a stress reaction of those with unstable power. Williams (2014) sug-gest that the effects of power are influenced by positional instability to result in more self-serving behaviour as opposed to other-centred behaviour. Sherman et al. (2012), have not included stability measures in their research that relate power and stress, but acknowledge the importance of stability as a potential moderator in this relationship and the limitations of not including stability. In line with these studies, this research proposes as well that power inter-acts with stability to affect leadership behaviour of managers. Before this interaction is focus-sed on in more detail, first a basis is provided for the leadership behaviour of managers.

2.3 Transformational Leadership Behaviour

Behaviour of managers towards their subordinates to reach organisational goals or implement change initiatives is experienced over time by subordinates as leadership behaviour. A defini-tion of leadership that fits the organisadefini-tional focus of this research and the one-dimensional view of power, is presented by Yukl (2013, p. 23): “leadership is the process of influencing others to understand and agree about what needs to be done and how to do it, and the process of facilitating individual and collective efforts to accomplish shared objectives”. For reasons explained below, this research focuses on transformational leadership behaviour and how this behaviour is influenced by the interactive effect of power and stability through stress and in-fluence subordinates’ support for change.

Transformational leadership behaviour of managers moves subordinates beyond imme-diate self-interests through idealized influence (charisma), inspirational motivation, intellec-tual stimulation or individualized consideration (e.g. Bass & Avolio, 1990; Bass, 1999). Over time, the dimensions of transformational leadership are adapted by several authors. This re-search will use the dimensions as proposed by Podsakoff, MacKenzie, Moorman and Fetter (1990), who considered six types of behaviours as transformational:

• Identifying and Articulating a Vision: Behaviour on the part of the manager aimed at identifying new opportunities for the organisation and developing, articulating, and in-spiring subordinates with his/her vision of the future;

(12)

• Fostering the Acceptance of Group Goals: Behaviour on the part of the manager aimed at promoting cooperation among subordinates and getting them to work together toward a common goal;

• High Performance Expectations: Behaviour that demonstrates the manager’s expecta-tions for excellence, quality, and/or high performance on the part of subordinates; • Providing Individualised Support: Behaviour on the part of the manager that indicates

that he/she respects subordinates and is concerned about their personal feelings and needs;

• Intellectual Stimulation: Behaviour on the part of the manager that challenges subordi-nates to re-examine some of their assumptions about their work and rethink how it can be performed, (adapted from Podsakoff et al., 1990).

(13)

Appropriate Model) or they are inspired to do the same and following the vision of the man-ager (i.e. Identifying and Articulating a Vision).

As can be seen from these examples, positive relational behaviours show the importance of and the wish to maintain and improve relationships (Johnson et al., 2008; Lewandowski et al., 2014; Weigel & Ballard-Reisch, 1999) and are strongly related to the relationship oriented transformational leadership behaviours. Therefore, managers that exhibit positive relationship behaviours can be said to be transformational. Next, the effects of the interaction between power and stability are examined and will be linked to transformational leadership behaviour.

2.4 The Interactive Effects of Power and Stability on Stress and Transformational Lead-ership Behaviour

Several studies have proposed stress as a mediator through which power and stability interact to affect behaviour (e.g. Jordan et al., 2011; Sapolsky, 2005; Scheepers, 2009). This research proposes stress as a mediator as well based on the following reasoning. Powerful individuals are likely to have access to more social and material rewards (Keltner et al., 2003; Maner et al., 2007). Therefore it is logical that powerful individuals are highly motivated to retain posi-tions of power when these are threatened (i.e. are unstable) (Bugental, 2000; Maslow, 1937). This reasoning is supported by research of non-human species, where the powerful experience the greatest stress-related physiology because of the potential loss of access to resources and prospective mates when the hierarchy is unstable (Manuck et al., 1995; Sapolsky & Share, 1994). Furthermore, since powerful individuals are likely to have a high status, this logic is supported by research of Scheepers & Ellemers (2005) who show that people in unstable high-status groups are subject to greater threat-based stress, relative to stable high-status groups. Across different settings experience those with unstable power more stress than those with stable power. This research propose that managers as well will experience more stress when their power is unstable compared to when their power is stable. Next, the effect of ex-perienced stress on transformational leadership behaviour of managers is explained.

(14)

one’s self by behaving friendly towards others and will at best be perceived as positive for a short period of time, but are not likely to benefit a manager-subordinate relationship in the long run.

It is likely that the negative effects of stress also apply to the manager-subordinate rela-tionship, which has informal, social aspects as well. Moreover, manager-subordinate relation-ships can be an intensive, long lasting and are build on liking, loyalty and professional respect (van Dam et al., 2008). Also, according to leader-member exchange (LMX) theory, managers do not have similar relationships with all their subordinates on a dyadic level (Graen, 2004; Graen & Uhl-Bien, 1995). Managers have a high-quality relationship with some subordinates and a low-quality relationship with others (van Dam et al., 2008). Especially high-quality, close relationships seem susceptible to the results of stress. Viewing the exchange between managers and subordinates as a high-quality relationship, this research proposes that ers who experience stress will behave more negatively towards their subordinates and manag-ers who do not experience stress will behave more positively towards their subordinates. Since transformational leadership behaviour is similar to positive relational behaviours, this research predicts that subordinates view their managers as less transformational when they behave more negatively towards subordinates as a result of experienced stress.

Based on the reviewed literature, I propose that power and stability have an interactive effect on transformational leadership behaviour:

Hypothesis 1 (H1): Power and stability will interact to affect transformational ship behaviour such that those with unstable power show less transformational leader-ship behaviour than those with stable power.

Also, I propose that the effects on transformational leadership behaviour are mediated by stress when power is unstable. This is formulated in two separate hypotheses. First, I hypothe-size that power will result in stress when it is unstable as opposed to stable power.

Hypothesis 2 (H2): Power and stability will interact to affect stress such that those with unstable power show more stress than those with stable power.

(15)

Next, the influence transformational leadership behaviour on subordinates’ support for change initiatives will be discussed.

2.5 Change Support

Organizational changes can take many forms. Changes can be small-scale or large-scale, con-tinuous or disconcon-tinuous, cultural or technological and can focus on individuals, groups, units or are organisation wide (Burnes, 2009). Many of those organizational change initiatives, however, do not bring the intended results. Estimations about the projects that are considered as failures are as high as two-third (Beer & Nohria, 2000) and some suggest this failure rate to be even higher (Burnes, 2005, 2009). The problem here lies not in the change initiatives themselves, but in their implementation (Klein & Sorra, 1996; Kotter, 1995; Kotter, 1996). The change plans that managers implement might be sound, but other factors interfere with the success of these change initiatives.

Increasingly, it is stressed that the central role of individuals (i.e. subordinate) is often underestimated by the change agents (i.e. managers) (Armenakis et al., 1993; George & Jones, 2001; Greenhalgh, Robert, Macfarlane, Bate, & Kyriakidou, 2004; Isabella, 1990; Lau & Woodman, 1995). This has resulted in a more micro-level focus with an essential role for individuals (Judge, Thoresen, Pucik, & Welbourne, 1999; Porras & Robertson, 1992; Tetenbaum, 1998). Change can only be successful when subordinates integrate the changes in their practices (Jones, Jimmieson, & Griffiths, 2005; Meyer, Srinivas, Lal, & Topolnytsky, 2007; Weeks, Roberts, Chonko, & Jones, 2004) or is at least strongly dependent on subordi-nates’ acceptance of and support for these changes (Armenakis et al., 1993; Armenakis & Harris, 2009; Fedor, Caldwell, & Herold, 2006; Smollan, 2006).

(16)

contin-uum. Subordinates can express behaviour that is more towards supporting the change or more towards resisting the change initiatives of the managers. This implies that when subordinates show less resistance, they are more supportive towards the change and when subordinates show more resistance behaviour, this is equal to less support for a change.

The attitudes of subordinates towards organisational change are shaped by situational variables. Situational variables can influence different subordinates differently because it is likely that aspects of a change initiative are perceived in different ways by individuals (Judge et al., 1999; Lau & Woodman, 1995). One of the situational factors that can influence subor-dinates’ attitudes towards a change initiative is leadership behaviour (Choi, 2011; Whelan-Berry et al., 2003). Next, the influence of transformational leadership behaviour on subordi-nates’ support for change will be discussed.

2.6 Transformational Leadership Behaviour

Two approaches are utilized to understand how managers’ leadership behaviour can influence subordinates’ support for change initiatives (Herold et al., 2008). On the one hand, does the leadership style approach presume that some leadership styles are better suited for certain change situations and this approach links leadership styles to organisational change in general, but does not address specific changes (Herold et al., 2008). This stream of literature mainly sees transformational leadership as effective during organisational change (Bass, 1999; Bass & Riggio, 2006; Conger, 1999; Herold et al., 2008; Pawar & Eastman, 1997; Shamir & How-ell, 1999), because such leadership is, “at its core, about issues around the processes of trans-formation and change” (Bass & Riggio, 2006, p. 225). Transtrans-formational leadership behaviour is already positively related to organisational commitment and OCB2 in general (Judge & Pic-colo, 2004; Lowe, Kroeck, & Sivasubramaniam, 1996; Yu et al., 2002), and particularly dur-ing times of organizational change (Bass et al., 2003; Carter et al., 2012; Herold et al., 2008; Nemanich & Keller, 2007; Shin et al., 2012). Furthermore, transformational leadership behav-iour is suggested to decrease cynicism about organisational change (Bommer, Rich, & Rubin, 2005; Wu, Neubert, & Yi, 2007). However, even though the importance of leadership during change is often highlighted (e.g. Gill, 2003; Kotter, 1996), little empirical evidence is avail-able about the influence of transformational leadership behaviour on subordinates’ change support (Burke, 2010; Herold et al., 2008), because change support relates more to specific change situations than organisational change in general.

2

(17)

On the other hand does the organisational change approach propose specific change behaviours to be most effective in leading certain aspects of a change implementation (e.g. creating and communicating an appealing vision or providing support during the change) (Herold et al., 2008; Kotter, 1996). The organisational change approach does not link the spe-cific change behaviours to broader theories of leadership (Herold et al., 2008) and the two approaches have not yet been sufficiently integrated (Bass & Riggio, 2006; Burke, 2010). More recently however, some attempts are made to integrate these two approaches (e.g. Carter et al., 2012; Groves, 2005) and link transformational leadership behaviour of managers during change to their subordinates’ commitment to those changes (Seo et al., 2012). The cur-rent research takes a similar perspective, but links transformational leadership behaviour of managers to subordinates’ support for change initiatives. Some studies make a distinction between the type of change (planned versus emergent) (Van der Voet, 2014) or the phase of the change (beginning phases versus later phases) (Seo et al., 2012) when researching this relationship, but these contingencies are beyond the scope of this research.

Transformational leadership behaviour can be conceptualised both on an individual level and on a unit level (Herold et al., 2008; Kark & Shamir, 2002). On an individual level, managers, exhibiting transformational leadership behaviour, send different signals to their subordinates, who interpret these differently (Herold et al., 2008). These different signals are based on each subordinate’s unique capabilities and needs while consulting, coaching and guiding them separately during the implementation of change initiatives (Carter et al., 2014). On a unit level, researchers have conceptualised transformational leadership behaviour exhib-ited by managers as signals that are shared and experienced similarly by all subordinates of a manager (e.g. Hackman & Oldham, 1980; Kark, Shamir, & Chen, 2003). This study will use the individual level conceptualisation of transformational leadership behaviour of managers because of the explicit manager-subordinate relationship that this research focuses on (see also the high-quality relationship of LMX theory) and the fact that situational factors like leadership, influence subordinates’ change support behaviour differently (Judge et al., 1999; Lau & Woodman, 1995). The last part of this theory section discusses on an individual level the relationship between transformational leadership behaviour of managers and subordinates’ support for specific changes.

2.7 Relationship between Transformational Leadership Behaviour and Change Support

(18)

that relates towards specific changes has mainly focussed on commitment to that change. Commitment, compared to other attitudinal constructs, is suggested to be most closely related to support for change (Herscovitch & Meyer, 2002; Meyer et al., 2007; Seo et al., 2012) be-cause transformational leadership behaviour activates several motivational processes which lead to the transformation of subordinates’ self-interests toward commitment to change targets (Shamir, House, & Arthur, 1993). Transformational leadership behaviour is already related to more commitment of subordinates in specific change situations (e.g. Bycio et al., 1995; Herold et al., 2008; Yu et al., 2002). Based on this stream of research, transformational lead-ership behaviour of managers is now related to support behaviour of subordinates for specific change initiatives.

The first set of behaviours of transfor-mational managers consists of identifying new opportunities and developing, articulating and inspiring others with a change vision of the future (Podsakoff et al., 1990). By communicating a positive, appealing, and emotional change vision to subordinates, combining organisational goals (i.e. change goals) with subordinates’ higher order values, managers raise the attractive-ness (Ford & Ford, 1994; Shamir et al., 1993) of these goals (Abrell-Vogel & Rowold, 2014). When subordinates can relate to the vision for a change, it becomes part of their self-concept; they identify with the vision and will work hard to support the change (Shamir et al., 1993). Therefore, a transformational change vision is likely to increase subordinates’ support for the change initiatives.

With this second set of transformational leadership behaviours, managers set an example for their subordinates to follow that is consis-tent with their espoused values (Podsakoff et al., 1990). Transformational managers are seen as role models that lead by example by communicating their own values and following these in their day-to-day behaviour (Abrell-Vogel & Rowold, 2014). In doing so, they show evi-dence of high integrity and reinforce their subordinates’ trust in them (Dirks & Skarlicki, 2004; Podsakoff et al., 1996). By guiding subordinates through a change and strengthen their trust in the manager to act according communicated values even in times of change, managers are seen as a role model (Abrell-Vogel & Rowold, 2014). Trust in managers can increase sub-ordinates’ commitment to and readiness for the change, decreasing cynicism about change (Abrell-Vogel & Rowold, 2014; Qian, 2008; Rafferty & Simons, 2006) and will thus have a positive effect on subordinates’ support for change initiatives.

Thirdly, behaviour shown by man-agers that is aimed at promoting cooperation among their subordinates and getting them to

2.7.1 Identifying and Articulating a Vision.

2.7.2 Providing an Appropriate Model.

(19)

work together toward a common goal is identified as another set of transformational leader-ship behaviours (Podsakoff et al., 1990). Working together and working towards common goals (i.e. the vision) is important for the success of a change initiative. By highlighting the importance of team spirit to achieve the challenging goals of the vision and by referring to past team success in achieving comparable goals, transformational managers accomplish that, according to social identity theory, team membership is viewed as positive by their subordi-nates and becomes part of their self-concept (Abrell-Vogel & Rowold, 2014; Tajfel & Turner, 2004). Uncertainty of a change situation results in stress for subordinates, similar to stress resulting from unstable power of managers. A positive self-concept is necessary for subordi-nates to cope with this stress and be able to spot the positive consequences of change (Judge et al., 1999). As a result, subordinates will support the change when they have spotted the positive consequences associated with it.

The next category of transformational lead-ership behaviours entails managers’ behaviour that demonstrates to followers the expectations for excellence, quality and high performance (Podsakoff et al., 1990). By expressing high performance expectations, transformational managers show trust in their subordinates and in their capability to meet these expectations (Abrell-Vogel & Rowold, 2014) and deal with challenges that arise during change (Vakola, Tsaousis, & Nikolaou, 2004). This trust to deal with these challenges increases subordinates self-efficacy (Eden, 1990; Yukl, 1989), which positively affect subordinates’ change readiness (Cunningham et al., 2002), coping with change (Judge et al., 1999), openness to change (Wanberg & Banas, 2000) and commitment to change (Herold et al., 2008). In conclusion, trusting subordinates to solve the challenges of a change initiative is likely to result in more support for that specific change.

Another set of behaviours that is identified as transformational, is behaviour on the part of the manager that indicates respect for subordi-nates and concern about their personal feelings and needs (Podsakoff et al., 1990). As stated before, uncertainty of a change situation can be stressful for subordinates. Their perceived capability of dealing with these challenging situations plays an important role in coping with stress (Folkman & Lazarus, 1984) and influences positive interpretations of these situations. Individualized Support of managers offers support and guidance to subordinates and in doing so reduces their stress level (Rowold & Schlotz, 2009), resulting in more individual resources to effectively cope with the change situation and a more positive interpretation of the change, leading to more change support.

2.7.4 High Performance Expectations.

(20)

Finally, behaviour on the part of the manager that challenges subordinates to re-examine some of their assumptions about their work and rethink how it can be performed is seen as transformational as well (Podsakoff et al., 1990). Subordi-nates are continuously challenged by certain behaviours and are likely to “focus on their tasks instead of on external worries and concerns” (Shin & Zhou, 2003, p. 704). This can lead to creativity among followers (Jung, 2001; Shin & Zhou, 2003), seizing opportunity to posi-tively reframe past experiences (Ford, 2002). Subordinates that receive continuously Intellec-tual Stimulation behaviours from their managers are therefore able to independently rethink negative beliefs of a change initiative and to reframe problems that arise as challenging op-portunities that they want to seize and thus support (Abrell-Vogel & Rowold, 2014).

These types of considerate, supportive, and positive behaviours on the part of the man-ager create subordinates’ buy-in to the change and develop a sense of ownership (Abrell-Vogel & Rowold, 2014; Groves, 2005). The willingness to be part of the change is enhanced and subordinates actively involve themselves in the change process, resulting in positive atti-tudes towards the change a high intention to support it (Carter et al., 2012). In conclusion, this research proposes that transformational leadership behaviour of mangers has the potential to positively influence subordinates’ support for a specific change initiative. The following hy-pothesis is formulated:

Hypothesis 4 (H4): More transformational leadership behaviour of managers will lead to more support from subordinates for change initiatives.

In short, this theoretical framework discussed the influence of power and stability on trans-formational leadership behaviour of mangers and the influence of this behaviour on subordi-nates’ support for a specific change. A final hypothesis is formulated that includes all previ-ous predicted relationship:

Hypothesis 5 (H5): Power and stability of managers will interact to affect stress, trans-formational leadership behaviour and subordinates’ change support such that those managers with unstable power show more stress, resulting in less transformational leadership behaviour which leads to less change support from subordinates as opposed to managers with stable power, who show less stress, resulting in more transforma-tional leadership behaviour which leads to more change support from subordinates.

The next section explains the methodology that is used to test this theoretical framework.

(21)

3. METHODOLOGY

In this section the research design, the sample and the procedure for data collection of this study will be discussed. Also the measures for the variables and control variables and the phases of the data analysis will be explained.

3.1 Research Design

The conceptual model of this research was tested with a quantitative research design. For this theory testing research, data was collected by means of two questionnaires; one to be com-pleted by managers and the other by the direct subordinates of those managers. Managers who participated in this study were required to have at least two subordinates, who they managed directly and with whom they were in frequent contact.

3.2 Population and Sample

The companies participating in this study were selected randomly; first, via my professional network and second, randomly chosen institutions throughout The Netherlands. First 49 re-quests for participation in this study were sent out to my professional network, comprising a wide variety of companies. This did not yield enough results, 27% of the managers that were approached, agreed to participate (13 managers) and only 62% of those completed the ques-tionnaire and had at least two subordinates complete their quesques-tionnaire as well (8 managers). Second, requests to participate in this study were sent to approximately 400 governmental institutions throughout the Netherlands and approximately 150 managers in some of the de-partments of a large professional education institute in the Netherlands. In response to this, 61 managers (approximately 10%3) were willing to participate in this study, and 79% (48 man-agers) of them completed the questionnaire and provided at least two subordinates who com-pleted out their questionnaire as well. In total, 56 managers participated in this study and each manager provided at least two subordinates to participate as well. Some managers provided more subordinates (up to 8). Two subordinates of one manager completed only two comple-mentary half’s of the questionnaire. Their results were excluded from the analyses, because it could not be traced that it was the same subordinate who first completed part of the question-naire and later on finished it. In total, 155 subordinates completed the questionquestion-naire.

3

(22)

3.3 Data Collection

The data was collected through two online questionnaire’s, created in the program Qualtrics. A direct link to the questionnaire was sent to the managers and they received a unique code to be used when completing the questionnaire to assure anonymity. The managers could com-plete the questionnaire in their own time and were reminded when they had not comcom-pleted the questionnaire after a week and thanked for their participation when they finished the ques-tionnaire. During the questionnaire, the managers were asked to provide the email address of two to five subordinates. Next, a personalised4 questionnaire was created for the subordinates of that manager and a direct link to this questionnaire was sent to them. The subordinates also received reminders to complete the questionnaire.

3.4 Measures

Well-known, validated and reliable multi-item scales were used to measure the five variables of the conceptual model of this research. These scales were translated into Dutch, the main language of the participating managers and subordinates. The scales were first translated into Dutch and then back-translated to English by someone else. Differences between the original scales and the back-translations were discussed and the Dutch translation was adapted until agreement was reached. The questionnaire for the managers included three variables of the conceptual model: Power, Stability and Stress. The questionnaire for the subordinates in-cluded two variables of the conceptual model: Transformational Leadership Behaviour of the manager and Change Support of the subordinate. Furthermore, both questionnaires contained questions about personal demographics. Most questions were statement questions, where managers or subordinates could rate to what extent they agreed with statements on a 7-point Likert scale ranging from 1 (totally disagree) to 7 (totally agree). Sometimes, other ways of measurement were used in the questionnaires; these are specified below if applicable. An overview of all questions that measured the variables of the conceptual model, including the control variables, is presented Appendix A.

The variable Power was measured with a validated measurement scale containing eight items; the Sense of Power scale developed by Anderson and Galinsky (2006). This measurement scale assesses whether managers think that they have power result-ing from their position or function in the organisation. An example of a statement (SoP_3) of this scale is: “When I am working, I can get others to do what I want.” Furthermore, the actual

4

The name of the manager was inserted in several places in the subordinate questionnaire when the subordinates were asked to think about their manager and answer questions about him/her.

(23)

power of managers was measured by means of four questions. These items addressed manag-ers’ (1) position in the hierarchy with a slide bar from 0 (bottom) to 100 (top); (2) their level of seniority with a 5-point Likert scale ranging from 1 (starter) to 5 (very senior); (3) the number of subordinates they supervise with 6 answer possibilities (none, 1-5, 6-10, 11-15, 16-20, and more than 20); and (4) the organisational level they belong to with a 3-point Likert scale with 1 being lower management, 2 middle management and 3 top management.

The variable Stability was measured with a validated measurement scale containing twelve items; the Stability of Power scale developed by Deery et al. (1994). This scale com-prises four subscales: General Stability, Promotion, Demotion and Job Position Security. The General stability subscale had three items that measured whether managers perceive their cur-rent position as stable or likely to change. An example of a statement (Stab_GS_3) of this subscale is: “I feel my position in the organization is stable.” The Promotion subscale also had three items that measured whether managers think they will be promoted soon within the or-ganisation. One of the statements (Stab_P_1) of this subscale is: “There is a possibility that I could be promoted to a higher position.” The third subscale of the Stability of Power scale is Demotion with three items. This subscale is the opposite of the Promotion subscale and as-sessed managers’ perception of being demoted soon. An example statement (Stab_D_3) from this subscale is: “I think I will be demoted to a lower position in the near future.” The fourth subscale, Job Position Stability, also contained three items. This subscale measured how safe managers think their position in the organisation is. An example of one of the items of this subscale (Stab_JPS_2) is: “My job position is secure.” Furthermore, the variable Stability is measured with the following question: “How stable is your position within the organisation?” This question had a slide bar as answer option ranging from 0 (very unstable) to 100 (very stable)5.

The variable Stress was measured with a validated measurement scale of four items; the Job Stress scale developed by Dubinsky, Yammarino and Jolson (1994). This scale measured the extent to which managers experience stress during their work. An example of a statement (Str_1) of this scale is: “I feel a great deal of stress because of my job.”

The validated measurement scale of three items, developed by Podsakoff et al. (1996)6, measuring Intellectual Stimulation was used as a proxy to measure the variable Transforma-tional Leadership Behaviour. This subscale measured to what extent managers stimulate their subordinates to re-examine their assumptions about their work and rethink how it can be

5

For data analysis purposes, the data resulting from this item was re-coded into a 7-point Likert scale.

(24)

formed. An example of a statement of this subscale is: “My manager has stimulated me to think about old problems in new ways.”

The last variable, Change Support, is measured by three of the four subscales of a vali-dated measurement scale developed by Oreg et al. (2008); Resistance to Change. The three subscales are Routine Seeking with five items, Emotional Reaction with four items and Short Term Focus with four items as well. The subscale Routine Seeking measured whether subor-dinates prefer change or like things to stay the same. An example (RtC_RS_1) from this scale is: “I generally consider changes to be a negative thing.” The Emotional Reaction sub-scale described the emotional reaction that subordinates had towards change. An example (RtC_ER_3) of this scale is: “When things don’t go according to plans, it stresses me out.” The third subscale is Short Term Focus. This subscale assessed subordinates’ stance towards change with regard to future perspectives. An example (RcT_STF_3) of this subscale is: “When someone pressures me to change something, I tend to resist it even if I think the change may ultimately benefit me.”

Age and Gender, both from the managers and the subordi-nates, are used as standard control variables. For Hypotheses 1, 2 and 3, the Age and Gender of the managers were used as control variables because the leadership behaviour of the man-agers was the focal target in these hypotheses. In Hypothesis 4, the Age and Gender of the subordinates were used as control variables, because the support for change of subordinates is the main interest of this hypothesis. For Hypothesis 5, all four variables are used as control variables. An overview of all the items is presented in Appendix B.

3.5 Data Analysis

After the data collection, the data was cleaned, erroneous and incomplete results were ex-cluded from further analysis and the data file of the managers was merged with the data file of the subordinates7. The data analysis consisted of three phases. First, principal component tor analyses and Chronbach’s Alpha’s were used for data reduction. Two rotated varimax fac-tor analyses were performed, one for all the items completed by the managers (Power, Stabil-ity and Stress) and another for all the items completed by the subordinate (Transformational Leadership Behaviour of the manager and Change Support of the subordinate). An eigenvalue of 1 was maintained when performing the factor analyses. Missing values were replaced with means because there were almost no missing values in the dataset. Furthermore, small

7

The two data files were merged into a Long Data File, as opposed to a Wide Data File. This means that the data of a certain manager was insert next to the data of each of his/her subordinates.

(25)

(26)

4. RESULTS

In this section, the results of the data analysis are presented. First, data reduction is discussed. Next, descriptive statistics and correlations between the different variables and control vari-ables are presented. Finally, the regression analyses of the hypotheses are discussed.

4.1 Data Reduction

Three variables of the conceptual model (Power, Stability and Stress) are measured in the managers’ questionnaire. All items measuring these variables are input for a principal compo-nent factor analysis in order to see the underlying relationships between the items (Child, 1990). Some items are excluded from further analysis for a variety of reasons, for example: a commonality value that was too low; a tendency to load alone in a component; or a strong tendency to load in components with items that measured different variables. Table 1 gives an overview of the items that are excluded during the principal component factor analysis and the reasons for exclusion.

(27)

TABLE 1

Items Excluded during Factor Analysis Item Reason for exclusion from further analysis

SoP_1 - loads negatively and item cannot be reversed - loads with Stab_GS items

SoP_2 - loads with Stab_D items

- loads, besides the primary component with items measuring Power, into other components (with items measuring Power and items measuring Stability) with a difference <0,3 SoP_3 - KMO adequacy is 0,497 (0,549 after exclusion)

SoP_4 - has a low communality value (<0,5) - loads with Stab_D items

- loads, besides the primary component with items measuring Power, into another component (with items measuring Power) with a difference <0,3

SoP_5 - has a low communality value (<0,5)

- loads, besides the primary component with items measuring Power, into another component (with items measuring Stability) with a difference <0,3

SoP_6 - loads with Stab_D items

SoP_7 - loads in a component with no other items SoP_8 - loads negatively and item cannot be reversed Pow_2 - KMO adequacy is 0,497 (0.549 after exclusion) Pow_3 - loads negatively and item cannot be reversed

- loads in a component with no other items

- many items measuring Stability load, besides their primary component, with Pow_3, with a difference <0,3

Stab_GS_1 - loads, besides the primary component with items measuring Stability, into another component (with items of the Stab_P subscale) with a difference <0,3

Stab_GS_2 - loads, besides the primary component with items measuring Stability, into another component (with items of the Stab_P subscale) with a difference <0,3

Stab_D_1 - even though this item forms with Stab_D_3 an acceptable component, the Cronbach’s Alpha is 0,607

Stab_D_2 - loads, besides the primary component with items measuring Stability, into another component (with items measuring Stability) with a difference <0,3

Stab_D_3 - even though this item forms with Stab_D_1 an acceptable component, the Cronbach’s Alpha is 0,607

Stab_JPS_3 - loads with Pow_1 and Pow_4

- loads negatively and item cannot be reversed

- loads, besides the primary component with items measuring Stability, on other components (with items of the Stab_GS subscale and with items of the Stab_P subscale) with a difference <0,3

Str_2 - has a low communality value (<0,5)

TABLE 2

Rotated Component Matrix and Communalities

Items Communalies Component1

Extraction 1 2 3 4

StabScGrouped ,775 ,858 Stab_GS_3 ,724 ,839 Stab_JPS_2 ,718 ,837

(28)

Stab_P_3 ,858 ,915 Stab_P_1 ,849 ,915 Stab_P_2 ,659 ,773 Str_1 ,807 ,895 Str_3 ,764 ,862 Str_4 ,746 ,824 Pow_1 ,870 ,902 Pow_4 ,854 ,883 Notes:

Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization.

1_{Rotation converged in 5 iterations.}

TABLE 3

Cronbach’s Alpha’s and Variables for Power, Stability and Stress Variable name Cronbach’s Alpha Items of which the variable is composed

Pow_1 - Pow_1

Pow_4 - Pow_4

Stab ,817 StabScGrouped, Stab_GS_3, Stab_JPS_1, Stab_JPS_2 Stab_P ,851 Stab_P_1, Stab_P_2, Stab_P_3

Stress ,828 STR_1, STR_3, STR_4

Two variables of the conceptual model (Transformational Leadership Behaviour of the man-ager and Change Support from the subordinate) are measured in the subordinates’ question-naire. All items measuring these variables are input for a principal component factor analysis in order to see the underlying relationships between the items (Child, 1990). Some of the items are excluded from further analysis for reasons similar to those mentioned above. Table 4 gives an overview of the items that are excluded during the principal component factor analysis and the reasons for exclusion.

(29)

of 0,7 (Nunnally et al., 1967) and combined with the low eigenvalue of this component (1,10) the decision is made to exclude both RcT_STF_3 and RcT_STF_4 from further analysis. Cronbach’s Alpha’s are calculated of the three components (Tlb_IS: α = 0,94; RcT_ER: α = 0,79; RcT_RS: α = 0,73) and subsequently three new variables are created (see Table 6 for an overview of all variables for Transformational Leadership Behaviour and Change Support).

TABLE 4

Items Excluded during Factor Analysis Item Reason for exclusion from further analysis

RtC_RS_4 - loads in a component with no other items RtC_RS_5 - has a low communality value (<0,6)

- loads into multiple components with items from the RtC scale with a difference <0,3 RtC_STF_1 - has a low communality value (<0,5)

- loads into multiple components with items from the RtC scale with a difference <0,3 RtC_STF_2 - has a low communality value (<0,5)

- loads into multiple components with items from the RtC scale with a difference <0,3 RtC_STF_3 - even though this item loads with RcT_STF_4 into a acceptable component, the Cronbach’s

Alpha is 0,607

RtC_STF_4 - even though this item loads with RcT_STF_3 into a acceptable component, the Cronbach’s Alpha is 0,607

TABLE 5

Rotated Component Matrix and Communalities

Items Communalies Component1

Extraction 1 2 3 TLB_IS_2 ,901 ,948 TLB_IS_1 ,857 ,920 TLB_IS_3 ,830 ,905 TLB_IS_4 ,818 ,904 RtC_ER_2 ,703 ,803 RtC_ER_4 ,594 ,767 RtC_ER_3 ,600 ,753 RtC_ER_1 ,605 ,736 RtC_RS_1 ,639 ,795 RtC_RS_2 ,677 ,302 ,764 RtC_RS_3 ,576 ,729 Notes:

Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization.

(30)

TABLE 6

Cronbach’s Alpha’s and Variables for

Transformational Leadership Behaviour and Change Support Variable name Cronbach’s Alpha Items of which the variable is composed

Tlb_IS ,940 Tlb_IS_1, Tlb_IS_2, Tlb_IS_3, Tlb_IS_4 RcT_ER ,793 RcT_ER_1, RcT_ER_2, RcT_ER_3, RcT_ER_4 RcT_RS ,727 RcT_RS_1, RcT_RS_2, RcT_RS_3

4.2 Descriptive Statistics and Correlations

Table 7 and 8 present descriptive statistics of all the variables and control variables that are included in this study. Some aspects of these descriptive statistics stand out. The answers given for Pow_1 range from 70 to 100 (M = 87,31; SD = 8,81), while the question measured the hierarchical position of managers from 0 to 100. On the contrary, Pow_4 that asks the level of management of managers (with three answer options: lower, middle, or top manage-ment), shows that the power of managers has a mean of 2,07 and a standard deviation of 0,79. Next, some of the descriptive statistics of the control variables are highlighted. The age of the managers ranged from 31 to 64 (M = 50,95; SD = 6,76) and the age of the subordinates ranged from 20 to 64 (M = 44,28; SD = 10,45). Thus, the average age of the sample is rela-tively high. The second control variable, gender shows that there is a relarela-tively equal distribu-tion between man and women in this study. Regarding the managers, 32 were male (57,1%) and 24 (42,9%) were female. The gender of the subordinates was even more equally distrib-uted, 78 were male (51,3%) and 74 were female (48,7%).

(31)

TABLE 7

Descriptive Statistics and Correlations Manager

Variable Mean SD mAge mGen Pow_1 Pow_4 Stab Stab_P Stress

mAge 50,947 6,760 - mGen 1,421 0,495 ,056 - Pow_1 87,309 8,812 ,206* -,106 - Pow_4 2,066 0,786 ,016 -,089 ,746*** - Stab 5,487 1,170 ,284*** ,095 ,344*** ,316*** - Stab_P 4,656 1,393 ,091 -,099 ,208* ,230** -,010 - Stress 3,884 1,394 -,048 -,047 ,038 ,186* ,001 -,018 - Notes: N = 152

mGen: 1 = Male; 2 = Female

Significance levels (2-tailed): p < 0,1 = †_{; p < 0,05 = *; p < 0,01 = **; and p < 0,001 = ***.}

TABLE 8

Descriptive Statistics and Correlations Subordinate

Variable Mean SD sAge1 sGen Tlb_IS RtC_RS RtC_ER

sAge1 44,282 10,454 - sGen 1,487 0,501 ,016 - Tlb_IS 5,628 1,102 ,059 ,186* - RtC_RS 2,531 1,152 ,020 -,095 ,003 - RtC_ER 2,987 1,059 -,005 -,137† -,144† ,440*** - Notes: N = 152 (1_{N(sAge) = 149)}

sGen: 1 = Male; 2 = Female

4.3 Hypotheses

(32)

Table 11), the regression results show a marginal significant negative main effect of Pow_4 on Tlb_IS (b = -0,17; SE = 0,95; t(151) = -1,78; p < 0,1). Stab has no main effect on Tlb_IS and there is also no interaction effect in this analysis. Furthermore, a marginally significant positive main effect is again found between mGen and Tlb_IS (b = 0,35; SE = 0,18; t(151) = 1,90; p < 0,1). Finally, Hypothesis 1 is tested with the variables Pow_4 and Stab_P (see Table 12). The regression results show no main effects of the predicting variables and no main ef-fects of the control variables, only a marginal significant negative interaction effect (b = -0,17; SE = 0,10; t(151) = -1,70; p < 0,1). In conclusion, Hypothesis 1 is not supported by any of the regression analyses. Even though Pow_1 and Stab have a significant interaction effect and Pow_4 and Stab_P have a marginally significant interaction effect on Tlb_IS, these re-sults are the opposite of what the Hypothesis 1 predicts.

TABLE 9

Regression Analysis of Hypothesis 1 Dependent variable: Tlb_IS

Predictor variables Model 1 Model 2 Model 3

Control variable mAge ,004 (,013) ,005 (,014) ,003 (,014) mGen ,362* (,180) ,336† (,184) ,337† (,184) Independent variable Pow_1 - -,082 (,097) -,081 (,097) Stab - ,040 (,098) ,029 (,102) Pow_1 x Stab - - -,039 (,097) R2 ,028 ,032 ,034 ∆ R2 ,028 ,005 ,001 Notes:

B coefficients (unstandardised) are reported in Italic, between parentheses the standard error

TABLE 10

Control variable

mAge ,004 (,013) ,007 (,014) ,002 (,013)

(33)

Independent variable Pow_1 - -,061 (,093) ,001 (,093) Stab_P - -,049 (,092) ,027 (,092) Pow_1 x Stab_P - - -,309** (,100) R2 ,028 ,033 ,093* ∆ R2 ,028 ,006 ,059 Notes:

TABLE 11

Control variable mAge ,004 (,013) ,002 (,014) ,000 (,014) mGen ,362* (,180) ,323† (,182) ,348† (,184) Independent variable Pow_4 - -,153 (,094) -,170† (,095) Stab - -,067 (,098) ,045 (,101) Pow_1 x Stab - - -,080 (,086) R2 ,028 ,045 ,051 ∆ R2 ,028 ,018 ,006 Notes:

TABLE 12

(34)

Notes:

Hypothesis 2 predicts that power and stability will interact to affect stress such that those with unstable power show more stress than those with stable power. Multiple regression analyses with the different variables are conducted to test this hypothesis. The first regression analysis tests the interactive effect of Pow_1 and Stab on Stress (see Table 13). This regression shows no significant main effects or interaction effects. The second regression analysis to test H2 (see Table 14) includes Pow_1 and Stab_P as predictor variables. This regression shows no significant results as well. Testing H2 with Pow_4 and Stab as predictor variables (see Table 15), the regression shows a significant positive main effect of Pow_4 on Stress (b = 0,24; SE = 0,12; t(151) = 2,03; p < 0,05) No other main effects or interaction effects are found. Finally, Hypothesis 2 is tested with the predictor variables Pow_4 and Stab_P (see Table 16). This regression results showed again only a significant positive main effect of Pow_4 on Stress (b = 0,28; SE = 0,12; t(151) = 2,39; p < 0,05). In conclusion, Hypothesis 2 is not supported by any of the four regression analyses.

TABLE 13

Regression Analysis of Hypothesis 2 Dependent variable: Stress

Control variable mAge -,009 (,017) -,012 (,018) -,018 (,018) mGen -,125 (,230) -,112† (,235) -,109 (,235) Independent variable Pow_1 - ,060 (,124) ,064 (,123) Stab - ,009 (,126) -,037 (,130) Pow_1 x Stab - - -,162 (,123) R2 ,004 ,006 ,018 ∆ R2 ,004 ,002 ,012 Notes:

(35)

TABLE 14

Control variable mAge -,009 (,017) -,011 (,017) -,014 (,017) mGen -,125 (,230) -,116 (,234) -,092 (,234) Independent variable Pow_1 - ,070 (,119) ,105 (,122) Stab_P - -,039 (,117) ,004 (,122) Pow_1 x Stab_P - - -,176 (,132) R2 ,004 ,007 ,019 ∆ R2 ,004 ,003 ,012 Notes:

TABLE 15

Control variable mAge -,009 (,017) -,007 (,017) -,011 (,018) mGen -,125 (,230) -,063 (,230) -,014 (,232) Independent variable Pow_4 - ,277* (,119) ,244* (,120) Stab - -,069 (,125) -,113 (,128) Pow_1 x Stab - - -,160 (,108) R2 ,004 ,040† ,054 ∆ R2 ,004 ,036 ,014 Notes:

TABLE 16

Control variable