Exploring personal initiative by examining the role of managers as an antecedent of this proactive behavior : focusing on the mediating role of job satisfaction and the moderating role of role overload

Exploring personal initiative by examining the role of

managers as an antecedent of this proactive behavior:

focusing on the mediating role of job satisfaction and the

moderating role of role overload

Marlijn J. J. Sijtsma

June 2016

MSc Business Administration – Leadership and Management

Master Thesis

Statement of originality

This document is written by Student M.J.J. Sijtsma 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.

ABSTRACT

Personal initiative is a important concept in today’s turbulent environment. This study further explored this whole concept by considering antecedents of this component of proactive behavior. Surveys were distributed to 81 teams with a total of 301 employees and 81 managers. The purpose of this study was to investigate the role a manager has in increasing or decreasing personal initiative. This study elaborated on this by considering the underlying mechanisms of the relationship between leadership and personal initiative; job satisfaction as a mediator, and role overload as a moderator. Regression analysis with PROCESS shows that empowering leadership, job satisfaction and role overload have a direct effect on personal initiative. The relationship between empowering leadership and personal initiative was partly mediated by job satisfaction. Furthermore, directive leadership has a direct effect on job satisfaction, and an indirect effect on personal initiative through job satisfaction. No support is found for the direct effect of directive leadership on personal initiative, and no support is found for the moderating effect of role overload. Still the additional analysis shows that job satisfaction up to a moderate level of role overload increases personal initiative, when managed by an empowering leader. This study extents the literature a bit concerning the importance of leadership in increasing personal initiative, and the role of job satisfaction and role overload is its entirety.

Key words: personal initiative, empowering leadership, directive leadership, job

TABLE OF CONTENT

1. Introduction 5

2. Theoretical background 8

2.1 Empowering leadership vs. directive leadership 10

2.2 Personal initiative 12

2.2.1. The relationship between empowering vs directive 14 leadership and personal initiative

2.3. Job satisfaction 16

2.3.1. The relationship between empowering vs directive 16 leadership and job satisfaction

2.3.2. The relationship between job satisfaction and personal 19 initiative

2.3.3. The mediating role of job satisfaction 20

2.4. Role overload 22

2.4.1. The moderating role of role overload 23

3. Research method 24

3.1. Procedure 25

3.2. Sample 26

3.3. Measures 27

4. Data analysis and results 29

4.1. Data analysis 29 4.2. Results 30 4.2.1. Descriptive statistics 30 4.2.2. Mediation 31 4.2.3. Moderation 35 4.2.4. Additional analysis 37 5. Discussion 40 5.1. Theoretical implications 41 5.2. Practical implications 44 5.3. Future research 46 5.4. Limitations 47 6. Conclusion 49 7. References 50

1. INTRODUCTION

Proactive behavior is an increasingly important component of job performance. One element of proactive behavior is personal initiative (Crant, 2000). It is argued that personal initiative will be very important in tomorrows’ jobs because more personal initiative will be required through global competition, faster rates of innovation, new production concepts, and changes in the job concept (Frese & Fay, 2001; Ilgen & Pulakos, 1999). Personal initiative is necessary to develop knowledge and skills in order to be competitive on the global market. Rapid innovation requires individuals to come up with creative ideas and convert them into products or services. New production concepts, such as just-in-time approach, increases individual responsibilities to make decisions on their own. Furthermore, part-time and temporary jobs increase, so people are assigned to projects and work outside their tight structure. Employees should and rely on own decisions, and show personal initiative (Frese & Fay, 2001). Therefore, organizations nowadays should be concerned about this type of proactive behavior, which is characterized by its self-starting nature, its proactive approach, and being persistent by achieving a goal to overcome difficulties (Frese et al., 1996; Frese et al., 1997).

Much research has been done on this topic, still, there is much left over for further research. Many unanswered questions remain of which some will be answered in this study. What can managers do to stimulate personal initiative? What kind of leadership characteristics influence personal initiative? Are there other emotional needs that should be fulfilled before personal initiative can take place? These are questions raised by looking at an example of two teams in a big Dutch retail organization.

The manager of team X is providing much direction and guidance to the team members; telling them wat to do and sets deadlines. On the other hand, the manager of team Y providing

more autonomy and self-control to the team members. The team members of team X were less satisfied with their job than the team members of team Y and demonstrated more negative feelings towards the job and organization. Furthermore, team members of team Y seems to show more personal initiative than team X; they came up with new ideas and new projects they wanted to work for. Team X is performing better than team Y; the quarterly results show better numbers. For that reason it is suggested that personal initiative, in combination with higher job satisfaction, has a positive influence on team performance. In the end, this is also beneficial for the organization when teams are performing better. In the Dutch retail organization employees are working hard. Team members were most of the time overloaded with tasks, e-mail, meetings, and deadlines. The question arises whether this role overload affects the relationship between how the manager behaves and the extent to which an employee is showing personal initiative. Role overload can be characterized by too many demands place on the employee and that they experience too little time to fulfill them (Baruch et al., 1985). It can also be classified as time pressure. According to Baer & Oldham (2006), time pressure is to what extent an employee experiences that he or she has enough time to fulfill tasks, or need to work harder than usual.

Based on the Self-Determination Theory and Cognitive Evaluation Theory this study examines the relationship between empowering versus directive leadership and personal initiative (Deci & Ryan, 2011). Autonomy, self-control and self-determination are important antecedents of intrinsic motivation, which is part of personal initiative. Empowering and directive leaders differ on the autonomy and controlling aspect, therefore it is expected that this undermines or stimulates intrinsic motivation and have a great influence on personal initiative.

Specifically, the purpose of the research is twofold. First, it aims to enlighten the understanding of the relationship between leadership and personal initiative shown by team

members. Furthermore, this study examines the mediating role of job satisfaction in the relationship between leadership and personal initiative. Second, this study seeks to explore the moderating effect of role overload in the relationship between leadership and personal initiative. The expected relationships between the variables are displayed in figure 1.1 and 1.2. The former shows the research model with empowering leadership as the independent variable, while the latter shows the same research model with directive leadership as the independent variable.

This paper is organized into five sections. The first section will provide a theoretical background. The different concepts tested in this study will be explained, and how these concepts are interrelated. The second section will explain the methodology used to test the proposed hypotheses. The third section will present how the analysis is done and the results of the study. The fourth section will provide a discussion, including theoretical and practical implications, suggestions for future research, and limitations. Lastly, this paper will end with a conclusion.

This article examines the mechanisms underlying the relationship between two leadership styles that differ in the extent they put control on the team members, and personal initiative, by closely looking at job satisfaction and role overload. The findings should provide a better understanding of the effect of directive and empowering leadership, which can serve as a basis for practical implications in order to increase personal initiative in a team, which may lead to better team and organizational performance.

Figure 1.1 Research model 1

Figure 1.2 Research model 2

2. THEORETICAL BACKGROUND

In order to understand the different concepts in the model a theoretical framework is provided in the next section. The self-determination theory is set as a basis to build further on. Furthermore, this study will look at past theories and concepts to acquire more knowledge about the different relationships between the concepts in the research models. This section will provide the necessary theory to understand the expected relationships.

Self-Determination Theory

In this study the foundation lies in the Self-Determination Theory (SDT), which is an approach to human motivation and personality, and differentiates motivation in terms of autonomy and being controlled (Deci & Ryan, 2011). It evolved out of research that has been done on intrinsic and extrinsic motivation. In this study the focus will be on intrinsic motivation which is considered as a prototype of freedom and self-determination. SDT is a theory that views human beings as being proactive with an intrinsic functioning which might be influenced by the social context (Deci & Ryan, 1985). Self-determination can be enhanced or undermined by external factors (Deci & Ryan, 2011). Cognitive Evaluation Theory (CET) is a sub-theory of the SDT developed by Deci & Ryan (1985). This theory explains variability in intrinsic motivation in terms of social and environmental factors. According to CET people should experience self-determined behavior for intrinsic motivation. This requires support for autonomy and competence (Reeve, 1996). To increase proactive behavior, such as personal initiative, an employee should be intrinsically motivated (Isen & Reeve, 2005). According to CET the controlling aspect of directive leaders that pressures people to think, feel, and behave in a certain way does not provide autonomy, and therefore undermines intrinsic motivation (Deci & Ryan, 2011). On the other hand, empowering leaders provide autonomy (Bennis & Townsend, 1997) and self-control to employees, thereby enhancing intrinsic motivation. Therefore, this study assumes a positive relationship between empowering leadership and personal initiative, and a negative relationship between directive leadership and personal initiative. The next section will elaborate on the CET by clarifying all the relations which emanate mainly from self-determination and intrinsic motivation.

2.1 Directive leadership vs. empowering leadership

Team leadership is a dynamic process that is especially about social problem solving (Fleishman et al., 1991). This can be accomplished by managers through four generic responses 1) information search and structuring 2) information use in problem solving 3) managing personnel resources 4) managing material resources. The tasks within a team are accomplished by the team members, still the leader is responsible to ensure that all tasks are accomplished (Burke et al., 2006). Different leadership styles can be distinguished; for example, transformational, transactional, participative, democratic, empowering, and directive leadership style. Researchers made a distinction between all these leadership styles, because one style might be effective in one situation but not in the other. Not every leadership style is appropriate for each situation and one particular style might result in certain organizational behavior, such as personal initiative. This study will elaborate on empowering and directive leadership styles, because this is the clearest reflection of the contradiction between empowering employees versus exercising control over them.

Pearce et al. (2003) suggest that empowering leadership is aligned with: supporting, managing conflict and team building, developing and mentoring, delegating, and consulting. Empowering leaders encourage opportunity thinking, self-rewards, self-leadership, participative goal setting, and teamwork. Leaders can show empowering leadership by empowering their employees and create empowered teams. Argote (1999) emphasized knowledge sharing as a critical element of empowered teams.

Arnold et al. (2000) distinguished five different dimensions of empowering leadership. The first is leading by example, which means leader’s behavior that shows commitment to his or her work and the work of the team members. The second is coaching by teaching team members and let them become self-reliant. This also includes making suggestions about

performance improvements. The third dimension is participative decision making where input is used from the team members in making decisions. This includes also stimulating team members to express their own ideas and opinions. The fourth dimension is informing, which refers to what extent the leader shares information about the mission and philosophy with the team members. The fifth one is concern, which reflects to what extent the leader cares about the team members’ well-being. The sixth dimension is interacting with the team, which includes working closely with the team members and keep an eye on what is going on. Empowering leadership is recognized in the literature as having a positive influence on team performance (Srivastava et al., 2006). Team members in empowered teams have more autonomy, self-direction, and control over their work environment. Managers in this case should promote empowerment and support the team. Furthermore, they should provide information and resources, encourage self-reinforcement, provide social and emotional encouragement, build trust and openness, and encourage self-goal setting.

Fiedler (1989) described directive leadership as a framework for decision making and action that is provided by the leader to the team members in alignment with the vision of the leader. In order to increase goal attainment, the directive leader provides guidance and direction to the team members. Directive leaders are monitoring and managing teams (Druskat & Weeler, 2003). Jung & Avolio (1999) found that directive leaders are focusing on the future by providing value-laden goals. This should add meaning to the team member’s actions to reach this goals.

Pearce et al. (2003) argued that directive leaders provide an initiating structure, which means that goals are attained while minimizing role ambiguity and conflict. Directive leaders assign tasks to the team members, specify the way how tasks should be carried out, set clear channels for communication, and initiate and organize the team activity. Pearce et al. (2003) aligned behaviors from the Yukl (1987) typology with directive leadership and found that this

type of leadership is aligned with: organizing, monitoring, informing, setting clear roles and goals, and problem solving.

These two leadership styles are distinguished because they enable us to measure to what extent providing direction and guidance to team member will lead to certain employee behavior or affection. On the one hand directive leaders that provide specific guidance to achieve goals, and the means of it, and also sets performance standards (House, 1971). On the other hand, empowering leaders who provide responsibility and authority for decision making to team members (Hollander, 2012). Giving employees the support they need. It is a form of delegation, but better described as sharing broader responsibilities (Mills & Ungson, 2003). Two leadership styles in the end of the continuum, regarding autonomy and employees’ self-control.

2.2 Personal initiative

Personal initiative is defined as work behavior that is self-starting and proactive, that overcomes barriers to achieve a goal (Frese et al., 1996; Frese et al., 1997; Frese & Fay, 2001). The employee has often been viewed as somebody to be socialized into the job and into the company culture (Van Maanen, 1976). Employees were more passive, engaging in behaviors as a response to the stimuli in their environments, under control of the manager (Locke & Latham, 2002). Frese and Fay (2001) argued that in this case the goals are set by the organization or manager, and not by the employee. Providing training to the employee should ensure that they know what to do. In turn feedback is provided by the manager. This concept is very reactive in nature. More research is now arguing in favor of proactive behavior of the employee.

Grant & Ashford (2008) described the development from reactive behavior to proactive behavior in the literature. Examples of active behavior are personal initiative (Frese

et al., 1996, 1997), voice (Van Dyne & LePine, 1998), taking charge (Morrison & Phelps, 1999), and role breadth self-efficacy (Parker, 1998, 2000). Personal initiative is a new and different approach to organizational behavior and performance. Traditional non-active concept assumes that an organization can change the individual, but that the individual cannot change the workplace. The concept personal initiative is a new development in which researchers started focusing on participation of employees in their own socialization process (Frese and Fay, 2001).

According to Frese and Fay (2001), personal initiative composes of three elements; self-starting, pro-active and persistent. Self-starting means that an employee is doing something, without being told by the manager to do so. An employee sets their own goals, rather than accepting the goals assigned by the manager. Setting their own goals is based on personal ideas ore regarding an idea or project that is known. The second element is proactivity which means that personal initiative has a long-term focus. The long-term focus means that employees are looking for things that may occur in the future and to respond proactively to them. Personal initiative goes further than being only proactive, because it requires persistence when taking initiative to reach a certain goal. This third element can also be described as overcoming barriers. Personal initiative can be taken by an employee when a process, procedure, or task is changed. People can be resistant to change in first instance and may refuse to adapt to it. Therefore, persistence is needed to overcome this resistance and barriers.

Ilgen and Pulakos (1999) argue that future work will require a higher degree of personal initiative as a result of global competition, the faster rate of innovation, new production concepts, and changes in the job concept. Global competition will rely not only on company level, but also on the individual level. In order to remain competitive on the work market, employees continuously have to develop knowledge and skills. Faster rate of

innovation requires faster implementation of creative ideas. Personal initiative is important in developing creative ideas and to transform them into concrete products/services. New production concepts require increased responsibility for production, service and quality issues (Murphy & Jackson, 1999; Wall and Jackson, 1995). Personal initiative should be taken by employees to make decisions on their own. Lastly, changes in job contracts resulted in an increase in temporary work, where job security is decreasing (Bridges, 1995). People are more assigned to projects than to jobs. Working outside a rigid structure requires relying on own decisions, so personal initiative being important.

2.2.1 The relationship between empowering vs directive leadership and personal initiative Pearce et al. (2003) argues that empowering leaders encourage self-leadership, that they support their team members, and delegate responsibilities tot hem. Bennis & Townsend (1997) found that empowering leaders provide autonomy to employees. Research shows that autonomy is an important antecedent of proactive behavior among employees (Crant, 2000). Hackman (1987) described autonomy as the degree to which an employee experiences freedom and feels independent. Employees that perceive more autonomy at work experience a sense of ownership over their work and have more direct impact on outcomes. Employees are taking more responsibility, especially when they want to overcome barriers (Crant, 2000). This feeling of job control leads to higher personal initiative (Amabile et al., 1996).

Furthermore, empowering leaders stimulate self decision making and carrying out work without supervision and intervention of the manager (Bass, 1985). This gives the employee feel having control over the situation. Employees who perceive they have high control over their work feel more responsible to develop new ideas, in order to solve problems (Ohly et al., 2006). Furthermore, job control stimulates employees to experiment in the workplace and generate and enhance new ideas. When the employee feels having control over

his or her job he or she will show personal initiative. Furthermore, job control can be perceived as more challenging, and therefore results in greater personal initiative. Autonomy and self-control are important antecedents of personal initiative. Empowering leaders provide this autonomy and self-control to employees. Therefore, the following is proposed:

H1. Empowering leadership is positively related to personal initiative

On the other hand, a directive leader is characterized by someone who gives guidance to employees, let them know what is expected, and sets rules and regulations which they have to follow (House & Mitchell, 1975). Pearce et al (2003) argue that directive leaders assign goals to team members, rather than let the employees set the goals by themselves. In general, proactive goal generation is about envisioning and planning, thereby changing the self or the work environment. This is a process which is self-initiated, in which employees are setting the goals by themselves, rather than assigned by the leader (Srivastava et al., 2006). Proactive behavior is about self-initiated and future-oriented action that aims to change and improve the situation, and also are able to change oneself, rather than initiated and planned for by the manager (Parker et al., 2006). Controlling events seems to undermine intrinsic motivation as part of personal initiative (Deci & Ryan, 1980). Amabile (1996) found that students who where highly controlled by their leader show less personal initiative. So a leader providing no support for personal initiative is able to hinder this kind of proactive behavior by setting goals, initiate goal achievement, and exercising control over employees, rather than stimulating self-leadership. Therefore the following is proposed:

2.3 Job satisfaction

Job satisfaction has received much attention in research; thousands of studies done on job satisfaction can be found in the journals about organizational behavior. Locke (1969, p. 316) argues in his paper ‘What is job satisfaction’ that job satisfaction is ‘the pleasurable emotional state resulting from the appraisal of one’s job as achieving or facilitating one’s job values’. He also described job satisfaction as complex emotional reactions to the job, experiencing pleasure as well as displeasure with certain jobs or tasks. Job satisfaction is seen as an outcome of the relationship between what somebody expects or wants from the job, and what is actually perceived what it brings. Brief (1998) described job satisfaction as the attitude someone has regarding the job. Spector (1997) considered job satisfaction as a global feeling people have about the job, or different attitudes people have regarding various aspects of the job. Job satisfaction simply is ‘the degree to which people like their jobs’ (Spector, 1997, p. VII). It can be related to job payment, promotion possibilities, co-workers, supervisor, and the nature of the work itself. A job cannot be treated as a single entity as it composes of several tasks, overall job satisfaction is the sum of all the elements the job consists out of. In this study three elements of job satisfaction are considered; overall satisfaction with the job, satisfaction with the work itself, and satisfaction with the organization the employee works for.

2.3.1 The relationship between empowering vs directive leadership and job satisfaction Leadership is considered in the literature as an important determinant of job satisfaction. Researchers found a positive relationship between leadership and job satisfaction (e.g. Stogdill, 1963; House et al., 1971). Leaders adopt different attitudes and behaviors that

influence the employee feelings. Many researchers suggest that the main job of a leader is to ensure that all tasks and team maintenance are completed (Hackman & Walton, 1986).

Empowering leaders delegate authority and provide responsibility to employees (Hollander, 2012). This is consistent with the assumption that empowering leaders provide autonomy to employees (Bennis & Townsend, 1997). Autonomy seems to have a positive influence on job satisfaction (Hackman & Oldham, 1976). Job autonomy has been identified as one of the most important determinants of job satisfaction (Hackman & Oldham, 1980; Parker & Wall, 1998). Employees who have more autonomy at work are more satisfied as a result of greater self-regulation (Vroom, 1964).

Furthermore, empowering leaders stimulate participative goal-setting (Pearce et al., 2003). Locke & Schweiger (1979) described participation as joint decision making that also includes delegation by the manager. Oldham & Cummings (1996) found that employees are most creative when they are supervised by a manager in a participative way and are not highly controlled. Furthermore, London & Larsen (1999) investigated that managers not controlling the employee empower self-development. When employees are participating in decision making and planning they have a clear understanding of the goals and plans for implementation. When employees are able to participate in the long-term strategy they are able to influence their working environment (Hickson, 1986). Miller & Monge (1986) argue that participation of employees in decision making might increase job satisfaction. Empowering leaders apply more participative approaches, which enhance the employee’s feelings of personal control. Spector (1986) found that employees who perceived higher levels of control were more satisfied.

Next to participative decision making, providing self-control and autonomy, an empowering leader gives support to the team (Pearce et al., 2003). Billingsley & Cross (1992) found that job satisfaction is influenced by leadership support; the higher the manager is

supporting the team member the greater the job satisfaction. Empowering leaders stimulate personal initiative by joint decision making and providing support, self-control and autonomy. Therefore, the following is proposed:

H3. Empowering leadership is positively related to job satisfaction of the employee

A directive leader provides more guidance and direction (House, 1971), and setting clear roles and goals (Pearce et al, 2003). They make decisions by themselves, give instructions and commands to employees, and expect their compliance (Howell & Costley, 2001). Directive leaders are seen as defective decision makers. Furthermore, teams with a directive leader advance are less information seeking from team members and therefore decrease team member’s input (Peterson, 1997).

Nowadays, employees view their jobs more as personal fulfillment. Directive leadership seems not to fit in this expectation (Sims & Manz, 1996). There is more a climate of self-determination which means that employees are able to initiate and regulate their own behavior (Deci & Ryan, 1985). Employees should feel they have psychological ownership through participation in employee ownership initiatives. Employee ownership integrates the goals of the employee and the organization and foster the feeling that employees have control in the organization (Pearce et al, 2001). Deci et al. (1986) found that controlling behavior demonstrated by the leader was negatively related to employee’s satisfaction. Employees that experience self-control and autonomy have more positive attitude towards the job than when they are highly controlled. Control by the leader fosters processes with more rigid rules and regulations, this involves greater tension and pressure, and therefore a more negative attitude towards the job (Deci & Ryan, 1987).

In todays jobs employees desire to have autonomy and self-control in their job. Directive leaders are controlling employees, rather than stimulating this self-leadership. Therefore, the following is proposed:

H4.. Directive leadership is negatively related to job satisfaction of the employee

2.3.2 The relationship between job satisfaction and personal initiative

Much research has been done on the outcomes of job satisfaction or dissatisfaction. Satisfaction with co-workers, the work itself, pay, promotion, reduce the chance an employee will quit and leave the organization (Farris, 1971). On the other hand, dissatisfied employees are more likely to leave the organization. Lambert et al., (2001) found that job satisfaction has a large direct effect on turnover intent. Furthermore, the meta-analysis of McShane (1984) shows that employees who are less dissatisfied are more likely to be absent at work.

Roznowski & Hullin (1992) argued that data about job satisfaction among employees is the most informative data a manager can have to predict employee behavior. Ilies & Judge (2005) found that positive effects of an employee enhances optimism regarding the attainment of goals. This optimism of achieving goals may lead to personal initiative, for example. Personal initiative is self-starting and proactive in nature, has a long-term focus and is goal directed and action oriented (Frese & Fay, 2001). When work is satisfying an employee is more likely to show personal initiative to improve further the work situation (Sonnentag, 2003). Individual motivation is likely to enhance personal initiative (Isen & Reeve, 2005). An employee is intrinsically motivated, for example, when it engages in activities which are satisfying. Intrinsic motivation can be used to understand personal initiative (Frese & Fay, 2001). Motivation stems from a desire to achieve something (Ryan & Decy, 2000). An employee who is energized and also activated to achieve a certain goal can be classified as

motivated. Intrinsic motivation results from the motivation to do something because it is interesting and joyable. Positive feelings towards the job can promote a more responsible long-term focus (Isen & Reeve, 2005). Furthermore, it motivates employees to set more difficult and challenging objectives (Ilies & Judge, 2005).

Intrinsic motivation is an important element of personal initiative. An employee is intrinsically motivated when the work is satisfying. Employees will show personal initiative, by setting difficult goals and having a long-term focus, when he or she is satisfied with the job. Therefore, the following is proposed:

H5. Job satisfaction is positively related to personal initiative

2.3.3 The mediating role of job satisfaction

Intrinsic motivation does not result in detaching oneself from work when difficulties arise, but motivates the employee to demonstrate personal initiative (Sonnentag, 2003). To understand motivation, we refer to the motivational states which are ‘can do’ and ‘reason to’, and ‘energized to’ (Parker et al., 2006). ‘Can do’ motivation refers to self-efficacy, control appraisals and attributions, and also the perceived cost of the action. Frese & Fay (2001) argue that to show personal initiative individuals should feel they control the situation. The reasons behind this are that people who feel having control over the situation are likely to maintain a strong sense of responsibility. Secondly, ‘reason to’ motivation reflects whether an employee feels able to improve things at work and see value to show personal initiative in the future. For self-initiated goals, the ‘reason to’ is not very clear. However, when the goals are externally given, ‘reason to’ is more clear. Griffin et al. (2007) argue that employees may show more personal initiative in situations where goals are not clearly externally specified and that attainment is not linked to rewards.

Research shows that managers play an important role in motivating people at work. Taylor et al. (2008) found out for example that teachers who are motivating providing more autonomy support, more instrumental help, and more involvement of the students. Managers showing this empowering leadership stimulate employees to shown self-leadership (Neck, 2006). Behavior is only intrinsically motivated when an employee is satisfied. When an employee experiences enjoyment, satisfaction, and pleasure he or she can be intrinsically motivated (Frese & Fay, 2001). Sonnentag (2003) found that satisfied employees are more likely to engage in proactive behavior. Only when the employee has the feeling of having control over work he can translate personal initiative in actual behavior. Work motivation is part of personal initiative; in order to show personal initiative an employee should be intrinsically motivated. Intrinsic motivation is part of personal initiative. An empowering leader can stimulate intrinsic motivation, and therefore personal initiative, by providing autonomy, support and self-control to employees. An employee can only be intrinsically motivated when he or she is satisfied with the job. Therefore the following is proposed:

H6. The positive relationship between empowering leadership and personal initiative can be explained by the extent to which the employee is satisfied with the job

A directive leader provides direction and gives orders. The manager showing this kind of behavior exercises power over the employees. This way of providing direction to employees impedes self-leadership and autonomy. Controlling, rather than autonomy and supportive leadership behavior, undermines intrinsic motivation (Deci et al., 1980). Furthermore, it seems that students who were highly controlled were showing less initiative (Amabile et al., 1996). Directive leaders decrease intrinsic motivation and hinder personal initiative by exercising control over employees. In order to be intrinsically motivated an

employee needs to be satisfied. However, directive leaders put greater tension and pressure on employees (Deci & Ryan, 1987), what results in a more negative feelings about the job. Therefore the following is proposed:

H7. The negative relationship between directive leadership and personal initiative can be explained by the extent to which the employee is satisfied with the job

2.4 Role overload

If demand on the job exceeds the capacity it can be classified as role overload (Wickens & Hollands, 2000). To what extent an employee perceives time pressure because of the number of this obligations can be defined as role overload (Reilly, 1982). According to Baer & Oldham (2006) time pressure is to what extent an employee experiences that he or she has enough time to fulfill tasks, or need to work harder than usual. Employees have certain commitments and responsibilities in their job. Role overload can be characterized by too much work and too little time get things done (Baruch et al., 1985). It is part of role conflict, which means that role demands are higher than the capabilities and resources available for it (House, 1980). Role overload reduces a person’s ability to control the work environment and that reduces the individual’s ability to work efficiently (Fried et al., 1998).

Generally, research perceives role overload as an important job stressor or stress predictors. Role ambiguity, role conflict, and role overload are all three considered as stressors. Role ambiguity takes place when the expectations for a given job are unclear or incomplete (Lichtman & Hunt, 1971). Role conflict refers to ‘the extent to which a person experiences pressures within one role that are incompatible with the pressures that arise within another role’ (Kopelman et al., 1983, p. 201). Concerning role overload, the distinction can be made between intellectual demands and time pressure demands (Karasek, 1979). In this study the focus lies on the latter, including also the complexity of tasks.

2.4.1 The moderating role of role overload

Existing literature suggests that empowering leadership increases personal initiative shown by team members, by providing autonomy (Crant, 2000). Furthermore, employees feeling self-controlled show proactive behavior through idea generation and problem solving (Ohly et al., 2006). Le Pine et al. (2005) argued that high demand in the job is a challenge-related stressor. Role overload seems to increase personal initiative when employees are trying to cope with high demands (Fay & Sonnentag, 2002). Past research found that job demands, such as time pressure are related to higher personal initiative (Fay & Sonnentag, 2002; Sonnentag, 2003).

Proactive behavior of an employee can be explained by looking at job control and job demands. Job control refers to the freedom an individual has to carry out certain tasks (Jackson et al., 1993). Lazarus & Folkman (1984) stated that a situation is perceived as challenging when it provides a personal gain. Only if these demands can be fulfilled be with personal effort (Skinner & Brewer, 2002). Parker et al. (2006) suggest that job control enhances self-efficacy and increases the felt responsibility, and therefore increases personal initiative. Empowering leaders provide room for self-control, so that self-fulfillment can take place. When the situation places higher demands on the employees, they should do more to be self-starting, pro-active, and persistent (Frese & Fay, 2001). Employees should focus more on long-term goal achievement (Fay et al., 2001) and be persistent to overcome the barrier of role overload. Role overload is a suboptimal work situation, so this triggers employees to be proactive (Frese & Zapf, 1994). An empowering leader is providing more autonomy and gives more freedom of self-control to the employee, providing responsibility and authority for decision making to the team members (Hollander, 2012). Ohly and Fritz (2010) argue that employees will show more personal initiative when job control and time pressure are both

high. Empowering leaders give responsibility to employees and provide more space for self-leadership. Especially when role overload is high and more work comes down to the employees they will show more personal initiative in order to overcome this barrier. Therefore, the following is proposed.

H8. Role overload strengthens the positive relationship between empowering leadership and personal initiative

Directive leaders assign goals to team members (Pearce et al., 2003), and provide guidance to employees (House & Mitchell, 1975). This controlling situation undermines intrinsic motivation (Deci & Ryan, 1980) and decreases personal initiative (Amabile, 1996). However, role overload requires employees to show more personal initiative to cope with time pressure (Ohly & Fritz, 2010). When role overload is high managers will also place more demand on individuals. Probably directive leaders still set clear goals and provide direction to employees (Yukl, 1987). However, it is expected that employees up to a certain point will increase personal initiative, and go beyond the manager’s expectations, in order to overcome this barrier of role overload (Crant, 2000). Therefore the following is expected:

H9. Role overload reduces the negative relationship between directive leadership and personal initiative

2. Research method

In the following section the research methodology will be presented. First, the design of the study is described, explaining in detail the procedure of data gathering. Secondly, the sample for this study is exemplified. Finally, the measures are given for each variable, including items and scales.

3.1 Procedure

The data was collected by three students and some part-time students of the University of Amsterdam (UvA), under the supervision of Phd Candidate R.E. van Geffen. Teams from different Dutch organizations were asked to participate in this study. The sample consists of two subsamples. The first subsample is composed by the three students, and the second subsample is composed by 42 part-time students. Surveys were developed and distributed online via LimeSurvey. Together, four surveys were developed; the first two surveys designed for the first moment of measurement and the second two surveys were designed for the second moment of measurement. Each pair of surveys is made up of one edition designed for the manager, and the other edition designed for the team members.

The surveys were distributed in Dutch or in English, depending on the preference of the participant. The items and scales are used from other studies written in English, so the surveys were translated into Dutch. This has been done by R.E. van Geffen and another expert in this field. Matching codes were used to link the completed surveys of the team members to the completed survey of the corresponding manager.

The participants received an e-mail with a personalized link to Lime Survey and were requested to complete the surveys online. The survey is introduced by a short explanation about the project and the voluntary and confidentiality of participation. The manager had to answer questions about the team and about work experiences. The survey distributed to the team members covered topics such as behavior at work, leadership behavior of the manager, and also their own work experiences. The surveys for the second moment of measurement also included demographic information.

The first two surveys were distributed on 11 February 2016. A reminder was send one week later. The second survey was distributed four weeks later, so data collection for survey

one was closed. This second survey was distributed in order to measure effects or changes over time. After one week the data collection for survey two closed as well.

3.2. Sample

The sample consists of 81 teams, including one manager and three or more team members. Purposive sampling is used to acquire teams from different organizations in the Netherlands. The teams work for different Dutch organizations, including multinationals, small companies, non-for profit organizations, and governmental organizations. The surveys are distributed to 382 participants, of which 81 managers and 301 team members. The amount of team members in each team is on average 3.72. The first subsample is composed by three students and consists of 36 teams. In this subsample 190 participants are Dutch speaking and 25 participants are English speaking. The other subsample is composed by 42 part-time students of the pre-master Business Administration at the University of Amsterdam (UvA). Snowball-sampling, a non-probability sampling technique, is performed by these students by asking their own team, and one other team in the organization they work for, to participate in this study. This comprises of 45 teams of the entire sample.

Concerning the survey of the first moment of measurement which was distributed to the team members and manager, 50 managers completed the survey, and 288 team members. In this study the team member’s ratings are only important, because it is the perception of the team member that matters. The overall response rate for the first measurement is 88.50%.

For the second moment of measurement the surveys were again distributed to 382 participants. For this time slot only 25 managers completed the survey and 83 team members. The response rate was 28.27%.

3.3 Measures

The surveys were developed based on the variables that should be measured. The three students and thesis supervisor all used the same dataset, so the surveys consists of 24 variables; 15 variables included in the employee survey and 9 variables included in the manager survey. For this study only the variables at stake were considered. Each variable has corresponding items in a range from 3 to 12 items. The items are developed by different authors, and combined to create one survey. All items are measured on a seven-point Likert scale; completely disagree (1), disagree (2), somewhat disagree (3), do not agree/do not disagree (4), somewhat agree (5), agree (6), completely agree (7). This study used 5 variables that consist of 32 items. An overview with all the variables including the corresponding items can be found in the appendix.

Leadership is divided by the two variables ‘directive leadership’ and empowering leadership’. Directive leadership is measured with a 5-item scale, developed by Zhang et al. (2011). Items that are included: ‘My manager asks me to follow standard rules and regulations, ‘my manager lets me know what is expected of me’ and ‘my manager tells me how the work needs to be done’. Cronbach’s alpha for the scale is .73. Empowering

leadership is measured by 12 items which are developed by Zhang & Bartol (2010). Three

examples of items that were included: ‘My manager allows me to do my job my way’, ‘My manager makes many decisions together with me’ and ‘My manager makes it more efficient for me to do my job by keeping the rules and regulations simple. All items concerning directive and empowering leadership covered topics about employee autonomy, self-regulation/self control, and participative decision making. Cronbach’s alpha for the scale is .93.

Personal initiative is measured with a 7-item scale developed by Frese et al. (1997). It

barriers (Frese & Fay, 2001). Therefore, items that are included: ‘usually I do more than I am asked to do’, ‘I am particularly good at realizing ideas’ and ‘I use opportunities quickly in order to attain my goals’. Cronbach’s alpha for the scale is .86.

Job satisfaction is measured with 3-item scale developed by Gladstein (1984). The

three items that are included: ‘I am satisfied with my present job’, ‘I am pleased with the work I do’ and ‘I am very satisfied with working in this organization’. Generally, this covers the definition of Spector (1997) concerning job satisfaction: the global feeling an employee has bout the job or the different attitudes he or she has towards the job. Cronbach’s alpha for the scale is .84.

Role overload is measured with a 5-item scale developed by Tarafdar et al. (2007).

Items that are included are: ‘I never seem to have enough time to do my actual work’, ‘I often do have to do more work than I can handle’ and ‘I often work beyond actual or official working hours’. These items can be classified as ‘role overload’, where an employee has to do more than the time available (Wickens & Hollands, 2000). Cronbach’s alpha for the scale is .85.

In the end of the survey of the second time measurement the employees and managers were asked, on a voluntarily basis, to fill in their gender, age, tenure, education and sector. In the survey is this stated by the following; ‘I am a women/man/other or rather not disclose’, ‘In what year were you born?’, ‘How long (in years) have you had your current position?’ ‘What is your highest, completed, level of education?’ and, only specified for managers, ‘In which sector/field do you work?’.

The data will be analyzed by performing simple regression analysis to examine the strength of relationships between the variables. Furthermore, PROCESS (Hayes, 2013) will be used to test direct effects and test whether the mediation models and moderation models are supported.

4. DATA ANALYSIS AND RESULTS

The next sections will present the analysis of the collected data and the results of this analysis. The procedure for analyzing the data is described in detail. Furthermore, the results will provide descriptive statistics and explain how mediation and moderation is done. The effects over time are measured over a period of four weeks. An additional analysis is done with the mediating variable job satisfaction and moderating variable role overload.

4.1 Data analysis

Analysis of the data has been done with the IBM SPSS Statistics 23 for Mac. One counter-indicative item of directive leadership has been recoded. Prior to analyzing the data the scale reliabilities were checked. One item of directive leadership has been removed in order to increase reliability and have a Cronbach’s Alpha above .70. The scales are reliable, being greater than .70. Missing data is not replaced. In order to get the average of all the items of each variable, new variables have been computed putting together the different items. Outliers were detected and one respondent has been removed from the data. This removal leads to a non-significant correlation between directive leadership and personal initiative and a substantially lower correlation coefficient. A normality check was done by looking at the normal probability plots and looking at the kurtosis and skewness coefficients. Data is normally distributed, except for empowering leadership and job satisfaction. The first assumption concerning multicollinearity issues is not violated. All VIF values are below the threshold of 3 so no multicollinearity exists. No assumptions of linear regressions were violated; linearity and additivity of the relationship between personal initiative and the other independent variables exists. Homogeneity of variances exists. Statistical independence of the errors exists because the Durban Watson score is almost 2. Furthermore, the error is normally distributed looking at the the histogram and P-P plot.

Transforming the data with a log transformation did no good at all; correlation coefficients and significance levels changed dramatically. Therefore, bootstrap is conducted to ensure that the model will produce reliable results even if the normality assumption is not met (Field, 2013). This bootstrap facility makes use of re-sampling with replacement from the original sample by 1000 times.

First, a regression analysis is performed. Bivariate correlations were used in order to look at the strength of the relationship between the variables. Secondly, PROCESS (Hayes, 2013), an add-on for SPSS, is used to test the direct effects and to test whether the model is mediated or moderated by job satisfaction and role overload. PROCCESS uses ordinary least squares to estimate the unknown parameters in linear regression, and is able to test mediating and moderating models. Thirdly, PROCESS is is used to conduct repeating measures and test effects over time. No effects over time are found so this is not further exemplified in this paper. Finally, an additional analysis is conducted to test whether role overload functions as a moderator in the relationship between job satisfaction and personal initiative.

4.2 Results

4.2.1 Descriptive statistics

Means, standard deviations, Cronbach’s Alpha and correlations are presented in table 4.1. The results shows us that managers score high on directive leadership (M = 5.21) and also high on empowering leadership (M = 5.40), as indicated by their team members. Furthermore, on average employees score high on personal initiative (M = 5.75), are they highly satisfied (M = 5.71), and experience a high level of role overload (M = 4.56). Control variables are not taken into account, because the sample size will be reduced from 301 to 73 participants. The amount of participant from which demographic information is known is too small to generalize the outcomes to the target population, therefore, they are not included in this study.

Cases are excluded pairwise in order to have a greater sample size. Therefore, the sample size (N) varies for each correlation (table 4.1). The descriptive statistics find support for correlations between some variables. Empowering leadership is positively correlated with directive leadership (r = .46, p<.01), with personal initiative (r = .29, p<.01), and with job satisfaction (r = .43, p<.01). Directive leadership is positively correlated with job satisfaction (r = .18, p<.01). Job satisfaction is positively correlated with personal initiative (r = .27, p<.01). Lastly, role overload seems to be positively correlated with personal initiative (r = .13, p<.05).

The proposed hypotheses with direct effects, the mediating variable job satisfaction and moderating variable role overload are analyzed separately and the results will be presented in the following section.

4.2.2. Mediation

PROCESS, developed by Andrew F. Hayes (2013), is an add-on for SPSS to test mediation, moderation and conditional process analysis. Ordinary least squares is used to test direct and indirect effects. The simple mediation model is used in PROCESS to test the direct and indirect effect of directive and empowering leadership on personal initiative, through job satisfaction. Y is the outcome variable personal initiative, M is the mediating variable job satisfaction, and X are the independent variables empowering leadership or directive leadership.

Table 4.1 Mean (M), standard deviation (SD), sample size (N), Pearson correlation coefficients of

the variables and Cronbach's Alpha

Variable M SD 1 2 3 4 5 1. Directive leadership (N = 258) 5.21 .94 (.73) 2. Empowering leadership (N = 259) 5.40 .94 .46** (.93) 3. Personal initiative (N = 281) 5.75 .65 .12 .29** (.86) 4. Job satisfaction (N = 250) 5.71 1.01 .18** .43** .27** (.84) 5. Role overload (N = 250) 4.56 1.30 .11 .03 .13* .02 (.85) Pearson correlations; *p<.05 **p<.01

The output generated by the PROCESS application is presented in table 4.2 and 4.3. Bootstrapping for indirect effects is used, by re-sampling 1000 times. The percentile-based confidence intervals are 95%.

Research Model 1

In research model 1 the direct effects of empowering leadership, job satisfaction and personal initiative are tested. Furthermore, the indirect effect is measured through the mediating effect of job satisfaction. The direct effect of empowering leadership on personal initiative is B = .15 (p<.01). This means that when the other factor job satisfaction remains the same, the effect of empowering leadership on personal initiative is .15. In other words, when empowering leadership increases by one unit, the change in personal initiative is .15. With a confidence interval of 95% (.06 to .28) is this direct effect statistically different from zero. Hypothesis 1 is supported. The direct effect of empowering leadership on job satisfaction is B = .46 (p<.01). This means that when empowering leadership will increase by one unit, the change in job satisfaction is .46. So hypothesis 3 is supported. The direct effect that job satisfaction has on personal initiative is B = .12 (p<.01). When job satisfaction increases by one unit the change in personal initiative is .12. Hypothesis 5 is supported.

The indirect effect of job satisfaction in the relationship between empowering leadership and personal initiative is B = .05 (p<.01). This results from the fact that empowering leadership increases job satisfaction, which in turn results in a higher level of personal initiative. In other words, personal initiative will be affected by .05 when empowering leadership remains unchanged and job satisfaction changes by the factor it would change when empowering leadership increases by one unit. With a confidence interval of 95% this indirect effect of .05 is statistically significant from zero (.01 to .10). This means

that the relationship between empowering leadership and personal initiative can be partly explained by job satisfaction. Hypothesis 6 is supported.

The total effect that empowering leadership has on personal initiative is B = .20 (p<.01). This total effect is equal to the sum of the direct and indirect effect (AB + C). This means that when empowering leadership increases by one unit, personal initiative will be changed by .20 through the direct effect of empowering leadership and the indirect effect through job satisfaction. With a confidence interval between .12 to .28 this effect is statistically different from zero.

Research Model 2

In research model 2 the direct effects of directive leadership, job satisfaction, and personal initiative are tested. Furthermore, the indirect effect is measured through the mediating effect of job satisfaction. The direct effect of directive leadership on personal initiative is B = .03 (p = .25). This means that when the other factor job satisfaction remains the same, the effect of directive leadership on personal initiative is .05. In other words, when directive leadership increases by one unit, the change in personal initiative is .05. With a confidence interval of 95% (-.03 to .13) is this direct effect not statistically different from zero. Beforehand, a negative direct effect of directive leadership on personal initiative was expected. So hypothesis 2 is not supported. The direct effect of directive leadership on job satisfaction is B = .19 (p<.01). This means that when directive leadership will increase by one unit, the change in job satisfaction is .19. It was expected beforehand that directive leadership will lead to a decrease in job satisfaction. So hypothesis 4 is not supported. The direct effect that job satisfaction has on personal initiative is B = .17 (p<.01). When job satisfaction increases by one unit, the change in personal initiative is .17. Hypothesis 5 is supported again.

The indirect effect of job satisfaction in the relationship between directive leadership and personal initiative is is B = .03 (p<.01). With a confidence interval of 95% this indirect effect of .03 is statistically significant from zero (.01 to .07); This means that there exists an indirect effect between directive leadership and personal initiative through job satisfaction. It was expected that directive leadership will be negatively related to personal initiative, which would be mediated by job satisfaction. Therefore, hypothesis 7 is not supported.

The total effect that directive leadership has on personal initiative is B = .08 (p = .06). This total effect is equal to the sum of the direct and indirect effect (AB + C). This means that when directive leadership increases by one unit, personal initiative will be changed by .08 through the direct effect of empowering leadership on personal initiative and the indirect effect through job satisfaction. With a confidence interval between -.004 to .17 this effect is not statistically different from zero.

4.2.3. Moderation

PROCESS, developed by Hayes (2013), is used to test the moderating effect of role overload. The simple moderation model is used to test the conditional effect of role overload concerning the relationship between directive and empowering leadership and personal initiative. Y is the outcome variable personal initiative, M is the moderating variable role overload, and X are the independent variables empowering leadership or directive leadership. The output generated by the PROCESS application is presented in table 4.4 and 4.5. Bootstrapping for indirect effects is used by re-sampling 1000 times. The percentile-based confidence intervals are 95%.

Table 4.3 Testing mediation model 2

Consequent

M (JobSat) Y (PI)

Antecedent Coeff. SE p Coeff. SE p

X (Directive Leadership) a .19 .07 <.01 c' .05 .04 .25 M (Job Satisfaction) --- --- --- b .17 .04 <.01 constant i1 4.71 .36 <.01 i2 4.54 .29 <.01 R2 = .03 R2 = .08 F(1.25) = 8.15, p = 0.005 F(2.25) = 10.73, p<.01 Consequent M (JobSat) Y (PI)

Antecedent Coeff. SE p Coeff. SE p

X (Empowering Leadership a .46 .06 <.01 c' .147 .04 <.01 M (Job Satisfaction) --- --- --- b .117 .05 <.01 constant i1 3.21 .34 <.01 i2 429 .27 <.01 R2 = .18 R2 = .11 F(1.25) = 55.41, p<.01 F(2,25) = 15.53, p<.01

Research Model 1

The regression coefficient for empowering leadership (X) and the moderating variable role overload (M) is B = .03 (p=.38). This is not statistically different from zero. The results are presented in table 4.5.The relationship between empowering leadership and personal initiative is not dependent on the level of role overload. Hypothesis 8 is not supported.

Research Model 2

The regression coefficient for directive leadership (X) and the moderating variable workload (M) is B = -.01 (p = .75). This is not statistically different from zero. The results are presented in table 4.6.The relationship between directive leadership and personal initiative is not dependent on the level of role overload. Hypothesis 9 is not supported.

4.2.5. Additional analysis

No relationship has been found for the moderating variable role overload between leadership and personal initiative. It appears that no significant curvilinear relationship exists between role overload and personal initiative. However, this study supports the existing literature that role overload has a positive influence on personal initiative. In order to measure another possible effect of role overload in the model another research model was developed and tested (see figure 4.1). The moderating variable role overload is added to investigate the potential effect of this variable on the relationship between job satisfaction and personal initiative. This model is tested two times with empowering leadership and directive leadership acting as the independent variables.

Table 4.4 Testing moderation model 1 Coeff. SE t p Intercept i1 5.01 .73 6.85 <.01 Empowering Leadership (X) b1 .09 .13 .70 .48 Role Overload (M) b2 -.08 .17 -.50 .61

Role Overload x Empowering

leadership (XM) b3 .03 .03 .87 .38

R2 = .10

F(3.25) = 9.19, p<.01

Table 4.5 Testing moderation model 2

Coeff. SE t p

Intercept i1 4.93 .63 7.83 <.01

Directive Leadership (X) b1 .11 .12 .93 .36

Role Overload (M) b2 .10 .15 .69 .49

Role Overload x Directive

leadership (XM) b3 -.01 .03 -.32 .75

R2 = .03

Empowering leadership shows no significant interaction effect for this model. The interaction coefficient is B = -.01 (p = .64) with a confidence interval between -.07 and .04. However, the different values of role overload show a significant moderating effect at a low and moderate value of role overload. The different levels of role overload are measured on a seven-point Likert scale, from disagree (1) to completely agree (7). The low value of role overload is the mean minus one standard deviation, the moderate value is the mean, and the high value is the mean plus one standard deviation. Low role overload at a value of 3.25 has an effect of B = .06 (p<.01) with a confidence interval between .01 and .13. Moderate role overload at a value of 4.56 shows an effect of B = .05 (p<.01) with a confidence interval between .01 and .12. High role overload at a level of 5.86 shows no significant moderating effect (B = .04) with a confidence interval between -.02 to .13. This means that job satisfaction up to a certain level of role overload increases personal initiative. However, when role overload is high, job satisfaction of the employee does not play a role anymore.

Table 4.6 Conditional effect of role overload at all values

Value Coeff. P

Low 3.25 .06 <.01

Moderate 4.56 .05 <.01

High 5.86 .04 >.01

Directive leadership shows no significant support for this model. The interaction coefficient is B = -.01 (p = .78) with a confidence interval between -.06 and .05.

Figure 4.1 Research model; additional analysis

5. DISCUSSION

This study focuses on personal initiative as a component of proactive behavior which is required more and more in tomorrows’ jobs because of global competition, faster rates of innovation, new production concepts, and changes in the job concept (Frese and Fay, 2001; Ilgen & Pulakos, 1999), which could increase job performance (Crant, 2000). The purpose of this study was to examine the relationship between directive versus empowering leadership and personal initiative, mediated by job satisfaction or moderated by role overload.

The collected data was analyzed with PROCESS (Hayes, 2013) and the results shows us that only empowering leadership has a significant positive influence on personal initiative, which is partly mediated by job satisfaction. Despite no direct relationship exists between directive leadership and personal initiative an indirect effect takes place through job satisfaction. No support is found for the moderating effect of role overload. The repeated measures show no significant effect over time. Additional analysis shows that role overload has a moderating effect on the relationship between job satisfaction and personal initiative, up to a moderate level of role overload.