• No results found

The Relationship between Management Control Systems and Motivation, affected by different Generations

N/A
N/A
Protected

Academic year: 2021

Share "The Relationship between Management Control Systems and Motivation, affected by different Generations"

Copied!
60
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

The Relationship between Management Control

Systems and Motivation, affected by different

Generations

By

Durk Reijenga

S3370259

Faculty of Economics and Business (FEB) Accounting & Control

Abstract

To motivate, attract and retain (valued) employees, organisations should understand their work motivation. To align motivation of employees with the organisation’s objectives and strategies, an MCS can be used. Most organisations are not likely to address the different motivations of different generations in their MCS. This study investigates how MCS affect the motivation of employees, and how this relationship is affected by Baby Boomers, Generation X and Millennials. Regression analyses are used to analyse survey data from 315 employees from four higher professional educational organisations. A significant relationship between Belief Systems and more autonomous types of motivation were found. Furthermore, there is no moderating effect of generations on the relationship between MCS and motivation. Thus, when organisations change their strategies to motivate, attract and retain their employees, by changing their MCS they should consider less about different generations.

Keywords

Management Control Systems, The Continuum of Motivation, Simons’ Levers of Control, Generations, Baby Boomers, Generation X, Millennials.

Supervision

P. van Veen-Dirks and W.G. de Munnik Date: 20-01-2020

(2)

2

TABLE OF CONTENT

1. INTRODUCTION ... 4

1.1 Scientific Contribution ... 6

2. LITERATURE REVIEW AND HYPOTHESES DEVELOPMENT ... 7

2.1 Management Control Systems ... 7

2.2 Motivation ... 9

2.3 The relationship between MCS and Motivation ... 11

2.4 The relationship between MCS and Motivation, affected by Generations ... 13

3. METHODOLOGY ... 19

3.1 Respondents ... 19

3.2 Survey Instrument ... 19

3.3 Early and Late Response ... 20

3.4 Variables ... 20

3.4.1 Dependent Variable – Motivation ... 20

3.4.2 Independent Variable – MCS ... 20

3.4.3 Moderating Variable – Generations ... 21

3.4.4 Control Variables – Strategy, Culture and Structure ... 21

3.5 Data Analysis ... 23

4. RESULTS ... 25

4.1 Sample Characteristics ... 25

4.2 Descriptive Statistics and Correlation Analysis ... 26

4.3 Regression Analysis ... 28

4.3.1 Model 1 - Control Variables ... 29

4.3.2 Model 2 – Main Effect of MCS on Motivation ... 29

4.3.3 Model 3 - Moderating Effect of Generations ... 32

4.4 Robustness Checks ... 33

4.4.1 Results separated by Type of Employment ... 33

4.4.2 Age as a Moderating Effect ... 33

4.4.3 Autonomous and Controlled motivation as Dependent Variables ... 34

4.4.4 Positive Controls and Negative Controls as Independent Variables ... 34

4.5 Additional Analyses ... 35

4.5.1 Results separated by Organisations ... 35

4.5.2 The Moderating Effect of Generations including Departmental Size ... 36

4.5.3 Results split by Gender ... 36

5. CONCLUSION ... 38

5.1 Discussion ... 39

5.2 Limitations and future research ... 40

REFERENCES ... 41

APPENDIX A – COMPARISON OF THE CRONBACH ALPHA’S ... 50

APPENDIX B – EARLY AND LATE RESPONSE ... 51

(3)

3

APPENDIX D – TEST OF HOMOSCEDACTICITY ... 53 APPENDIX E – ROBUSTNESS CHECKS ... 54 APPENDIX F – ADDITIONAL ANALYSES ... 58

(4)

4

1. INTRODUCTION

Recently, a post on Facebook from EY1, a big accounting firm, went viral: “At this company, you can take 12 weeks off or temporarily work part-time if you wish”. Does EY attract the new

working generation, ‘Millennials’, with their flexible way of working? Over the last months, newspapers published articles about how to attract, retain and motivate Millennials. Perkins (2019) announced in Forbes that Millennials' tolerance for organisations decreased in 2019 and is still decreasing. Millennials are disrupting the world of work by protesting, quitting their jobs, and looking elsewhere for ways to make a living (Perkins, 2019). Data from the Central Bureau for Statistics (2019) shows that more than 50% of the working population consists of Millennials. Because Millennials do not have much loyalty to an employer, organisations should consider the possible methods that could motivate, attract and retain them (Petroulas, Brown & Sundin, 2010).

Motivation can be defined as the needs, desires, feelings and ideas that direct our behaviour toward a goal (Maslow, 1943). Motivation is a complex phenomenon of human behaviour; since individuals have different needs and are motivated by different internal and external factors (Bobar & Caperman, 2008). This study concerns motivation in the working environment. The motivation of the Millennials to work differs from other generations (Deal, Stawiski, Graves, Ruderman, Gentry & Weber, 2013). Millennials could be more driven by intrinsic motivators as (1) flexibility, (2) social relationships and the (3) opportunity to develop, compared to other generations (Kultalahti & Viitala, 2014). Similarities and differences between the three generations (Baby Boomers, Generation X and Generation Y) on the motivation to work can originate from their different experiences and background during life (Ballone, 2007). Those different experiences may lead to different preferences in characteristics, values and work ethics and can influence the way of being motivated (Ballone, 2007; Glass, 2007; Kerr & Heyns, 2018).

To increase the change that motivation, behaviours and decisions of employees align with the organisation’s objectives and strategies, a Management Control Systems (MCS) can be used. Merchant and Van der Stede (2012, p. 290) defined an MCS as all devices or systems managers

use to ensure that the behaviours and decisions of their employees are consistent with the organisation’s objectives and strategies. Most of the existing MCS are designed with 1 EY, previously known as Ernst & Young.

(5)

5

Baby Boomers or Generation X employees in mind (Petroulas et al., 2010). Petroulas et al. (2010) investigated the preferences of generations for the design and use of MCS and found that specific generational characteristics have an impact on their preferences for the MCS design. However, Young, Du, Dworkis & Olsen (2015) stated that most organisations are not likely to adequately address the different motivations of different generations in their MCS, especially if the companies do not understand the motivations underlying their employees’ behaviour (Young et al., 2015).

Most studies investigate the relationship between MCS and motivation (Baerdemaeker & Bruggeman, 2015; Van der Kolk, van Veen-Dirks & ter Bogt, 2018), but do not take into account the influence of generations on this relationship. By knowing how generations affect the relationship between MCS and motivation, management can use this information to improve the achievement of organisational and individual goals and motivate their employees. When employees are motivated, they are possibly more engaged with their work and therefore, this could increase the chance that organisational performance will improve (Abioro, 2013). Therefore, investigating the influence of generations on the relationship between MCS and motivation could result in findings that contribute to new management insights. This results in the following research question:

How do Management Control Systems affect the motivation of employees, and how is this relationship affected by different generations?

In the subsequent section, the scientific contribution to the existing literature of this study is mentioned. After that, the literature review and, consequently, the used methodology to achieve the results are presented. The final section provides a conclusion of the findings and a discussion, including the limitations of this study.

(6)

6

1.1 Scientific Contribution

Existing literature describes the characteristics of different generations, the relationship between different generations and motivation and the relationship between MCS and motivation (Hart & Brossard, 2002; Arsenault, 2004; Glass, 2007; Ballone 2007; Kerr & Heyns, 2018; Van der Kolk et al., 2018; Baerdemaeker & Bruggeman, 2015). These studies do not investigate the influence of generations on the relationship between MCS and motivation. This study narrows down the literature gap by taking into account different generations by investigating the relationship between MCS and motivation. This study could also contribute to the insights that each generation has different characteristics and work values, and therefore implicitly has preferences for different types of MCS as suggested by Petroulas et al. (2010).

This research has next to the theoretical implications, also an implication for society. By investigating the influence of generations on the relationship between MCS and motivation, this study aims at contributing to management insights to the way how they look at different generations and how they can pay attention to the motivation of employees. Furthermore, it may also help organisations to prevent them from losing employees. Because, when employees are more intrinsic motivated, they are less inclined to move to another job (Rosli & Hassim, 2017). By getting insights into how different generations possibly influence the relationship between MCS and motivation, management could adapt their strategy to motivate and retain employees. If employees are more engaged with their work, this could increase the chance that organisational performance will improve (Abioro, 2013).

(7)

7

2. LITERATURE REVIEW AND HYPOTHESES DEVELOPMENT

This chapter contains a literature review which is the basis for the development of the hypotheses. First, I analyse the different MCS types and the four ‘levers’ of control that are used in this research. Second, I investigate the theory behind the continuum of motivation. After that, I investigate the relationship between MCS and the motivation of employees and how different generations, as a moderating role, will affect this relationship. The derived hypotheses are presented within the different sub-paragraphs.

2.1 Management Control Systems

MCS describe the framework which could influence the behaviour of employees by applying the right methods and techniques with the purpose that the objectives of the organisation will be achieved (Franssen, Arets, van Loon & Vaassen, 2015). Until the fifties of the last century, the ‘control of organisations’ was mainly focused on cost determination and financial control (Haustein, Luther & Schuster, 2014). Anthony (1965) was the first who introduced the definition of Management Control as “the process by which managers ensure that resources

are obtained and used effectively and efficiently in the accomplishment of the organisation’s objectives.” (Anthony, 1965, p. 635). Based on the definition of Anthony (1965), nowadays,

more than twenty definitions of Control Systems are mentioned in the existing literature (Haustein et al., 2014). For example, the frameworks Merchant (Object of Control Framework), Anthony (Management Planning and Control Systems), Otley (The Performance Management Framework) and Simons (Levers of Control).

Simons (1995) introduced the Levers of Control framework (LoC) to help managers to implement an organisational strategy. This concept focuses on the individual itself and aims at giving direction for organisational behaviour and stimulating creative innovation (Simons, 1995). Simons’ LoC has gained a prominent position in Management Control literature and is among the top 3 most influential textbooks in the field of MCS worldwide (Strauß & Zecher, 2013; Kruis, Speklé & Widener, 2016).

Simons’ Levers of Control

Simons (1995) suggested that MCS need to provide information to think about new strategies due to the changing environment and the possible impact this will have on organisational strategies and goals. Within other MCS frameworks, the organisational strategies and goals are considered ‘fixed’ (Koekkoek & Corbey, 2017). The current increasingly diverse and changing

(8)

8

workforce has to deal with three different generations, which can result in changing strategies and goals and are therefore not considered as ‘fixed’ (Young et al., 2015). The MCS of organisations that are considered as ‘fixed’ are possibly not likely to adequately address the changing strategies and goals (Young et al., 2015). Simons’ LoC explicitly takes the principle into account that the goals of an organisation may change as a result of an adjustment of the strategy (Simons, 1995). The MCS framework of Simons consists of four levers: (1) Belief Systems, (2) Interactive Control Systems, (3) Boundary Systems and (4) Diagnostic Control Systems (Simons, 1995).

Belief Systems are focussing on a strategic level of an organisation; the managers want to share

the mission and vision throughout the organisation (Kruis et al., 2016). Further, Interactive

Control Systems are used to promote communication and influence organisational learning and

the emergence of new ideas and strategies (Simons, 1995). Organisational learning takes place at the individual level and needs to be shared by open communication across all levels (Argyris, 1977). Belief Systems and Interactive Control Systems are controls that motivate, reward, guide and promote learning and new ideas within the organisation (Tessier & Otley, 2012). Those controls do not limit the behaviour of the employee but aim at improving an enjoyable working environment (Tessier & Otley, 2012). Those controls stimulate inspiration, innovation and the feeling of freedom of choices which could be perceived as ‘positive’ by employees (Tessier & Otley, 2012). Therefore, Belief Systems and Interactive Control Systems are named positive

controls. Simons (1995) compares the positive controls to the yang of the Chinese philosophy

representing sun, warmth and light.

Simons (1995) compares the negative controls, Boundary Systems and Diagnostic Control Systems, to the opposite yin force, representing darkness and cold. Boundary Systems are used to set limits on opportunity-seeking behaviour and are used to strengthen the strategy by recognising behavioural threats are dealing with this in for example codes of conduct (Simons, 1995; Kruis et al., 2016). Further, Diagnostic Controls are focussing on the monitoring function of control and are used to motivate, monitor, and reward achievement of specified goals (Simons, 1995; Bedford & Malmi, 2015). Boundary Systems and Diagnostic Control Systems are controls that coerce, punish, prescribe and control (Tessier & Otley, 2012). Because it possibly limits the freedom of behaviour of the employee by setting boundaries and monitoring, this could decrease the feeling of freedom of choices. This could be perceived as ‘negative’ by employees (Tessier & Otley, 2012). While the word negative has ‘bad’ connotations, negative

(9)

9

controls are not defined as ‘bad’ controls. Instead, they are seen as important as positive controls (Simons, 1995).

Figure 1: Simons’ Levers of Control Framework (Simons, 1995).

2.2 Motivation

Motivation is a complex phenomenon of human behaviour; since individuals have different needs and are motivated by different internal and external factors (Bobar & Caperman, 2008). To explain the human behaviour of motivation different motivational theories are known, like the Maslow’s hierarchy of needs, the Herzberg’s motivation and the Self Determination Theory (SDT) (Haque, Haque & Islam, 2014). The SDT is widely used in research regarding motivation in different areas, for example, within education and the work environment (Reeve, 2002; Deci, Olafsen & Ryan, 2017). Deci and Ryan (1985) introduced the SDT and stated that underlying to motivation lays three basic needs: autonomy, relatedness and competence (Ryan & Deci, 2000). Van den Broeck, Vansteenkiste, De Witte, Soenens and Lens (2010) defined the three basic needs as follows: (1) Autonomy represents individuals’ inherent desire to feel volitional

and to experience a sense of choice and psychological freedom when carrying out an activity

(p. 972); (2) relatedness is defined as individuals’ inherent propensity to feel connected to

others, that is, to be a member of a group, to love and care and be loved and cared for (p. 972);

(3) competency is defined as individuals’ inherent desire to feel effective an interacting with

the environment (p. 972).

The SDT distinguishes two types of motivation, namely intrinsic motivation and extrinsic motivation (Deci & Ryan, 2008). They define intrinsic motivation as an engagement in certain

(10)

10

activities because the individual finds it inherently interesting and enjoyable and that the activities are performed in the absence of external rewards (Ryan & Deci, 2000). The motivation derives from the person himself and is motivated to learn something and to develop himself (Deci & Ryan, 2008). Extrinsic motivation occurs when a person is motivated by external rewards (for example a high salary or a lease car) or to avoid punishment (Ryan & Deci, 2000).

The continuum of motivation

Because a human being is not only driven by just one type of motivation due to different goals or desires, Ryan and Deci (2000) introduced the concept of the continuum of motivation. The continuum means that every person has a score on every form of motivation ranging from nonself-determined (controlled motivation) to self-determined (autonomous motivation) (Deci & Ryan, 2008). Within the continuum, the SDT defines different subtypes of extrinsic motivation. The subtypes are seen as falling along the controlled-to-autonomous continuum and are based on the degree that extrinsic motivation has been internalised (Deci & Ryan, 2000; Gagné & Deci, 2005; Kerr & Heyns, 2018). Internalisation describes the process of employees’ taking in values, attitudes, or regulatory structures (Gagné & Deci, 2005). As a result, the extrinsic motivation of behaviour is shifted into more autonomous types of motivation whereby the influence of an external reward is no longer needed (Gagné & Deci, 2005). The SDT defines four subtypes of extrinsic motivation: External Regulation, Introjected Regulation, Identified Regulation and Integrated Regulation. At which the External Regulation is the less internalised type, and Integrated Regulation is the most internalised type of extrinsic motivation. The SDT suggests that the internalisation of extrinsic motivation is determined by the degree to which people can satisfy the three basic needs (Gagné, Forest, Aubé, Morin & Malorni, 2010). When those basic needs are more fulfilled, this could lead to more internalisation of the extrinsic motivation, and therefore the employee could become more autonomous motivated (Gagné & Deci, 2005).

Figure 2 illustrates the different subtypes of motivation in the continuum. On the right side of the continuum, there are more autonomous types of motivation. The most autonomous type of motivation is intrinsic motivation. On the left side, there are more controlled types of motivation. Amotivation is the least self-determined type of motivation. Amotivation is the contrary of intrinsic motivation and is described as the total lack of intention and motivation

(11)

11

(Gagné & Deci, 2005; Kerr & Heyns, 2018). The individual is entirely non-autonomous (Deci & Ryan, 2008).

Figure 2: The self-determination continuum showing different types of motivation, the regulatory style, the locus of causality and the source of regulation.

2.3 The relationship between MCS and Motivation

In the existing literature, the relationship between MCS and motivation is described by using different concepts of MCS and motivation. In this research, the relationship between MCS and motivation is addressed by building upon Simons LoC to define the concept of MCS and the concept of the continuum of Ryan and Deci (2000) to define motivation.

The positive controls, Belief Systems and Interactive Control Systems, are defined as controls that reward, motivate, guide and promote learning and new ideas within an organisation (Tessier & Otley, 2012). Belief Systems aim to create a shared mission and vision among employees and Interactive Control Systems promote communication, new ideas and influence organisational learning (Simons, 1995). By sharing an organisational mission and vision, the purpose of Belief Systems, it can facilitate goals at the employee level. When individuals know what goals to fulfil, they will consider the possibility of achieving this goal (Liberman & Trope, 1998). Achieving those goals can be stimulated by guiding and promoting learning, supported by Interactive Control Systems. Stimulating employees to do what they are good at, using their talents and being able to develop them further, can increase the feeling that they can complete

(12)

12

their goals with the right knowledge and skills. Goals that are viewed as desirable and possible to achieve can contribute to the employees’ feeling of competence.

Interactive Control Systems support discussing employees’ goals in frequent dialogue with their managers. If their managers respond it can increase the employees’ feeling that the organisation care about them (Widener 2007). Besides, when employees and organisation share the same mission and vision, employees will be more likely to feel related to the organisation (An & Ayayi, 2018). Sharing a mission and vision stimulates a feeling of solidarity, which can increase the relationships between employees and strengthen the feeling that employees care about each other (An & Ayayi, 2018). Therefore, the desire for interaction, connection and belonging could enable the feeling for relatedness. When employees work in respect to the shared mission and vision of the organisation, they could perceive that they should not think about the activities that they perform or that the activities fit within the organisational frameworks. This can support an increased feeling of autonomy. This feeling can be further stimulated by giving employees the freedom to discuss and introduce new ideas and strategies.

In summary, the positive controls likely help to fulfil employees’ feeling of relatedness, competence and autonomy. Moreover, this suggests that Belief Systems and Interactive Control Systems can affect the working environment and contribute to the degree of fulfilling their basic needs, which in turn can affect the degree of internalisation. The more fully extrinsic motivation will be internalised, the more autonomous motivated the employee could be (Gagné et al., 2010). Therefore, I suggest that Belief Systems and Interactive Control Systems are positively related to more autonomous types of motivation, resulting in the following hypothesis:

H1: Belief Systems and Interactive Control Systems are positively related to more autonomous types of motivation.

The negative controls, Boundary Systems and Diagnostic Control Systems, are described as controls that coerce, punish, prescribe and control employees (Simons, 1995). Boundary Systems are used to set limits on opportunity-seeking behaviour. Diagnostic Control Systems are focussing on the monitoring function to determine differences between goal and reality in order to assess whether goals have been achieved or not. Those negative controls limit the behaviour of freedom of the employee by setting boundaries and monitoring and could decrease the feeling of competence. Because those systems prescribe what employees should do,

(13)

13

employees could interpret this as if they are not suitable to think and determine what the right activities are. Employees could also have the feeling that their managers enforce tasks on them due to predefined goals and tasks and could limit employees’ freedom of choices. Constraining the behaviour of an employee may lead to a decreased feeling of autonomy (Deci et al., 2017). I do not expect an influence of the negative controls on the feeling of relatedness.

In summary, by using negative controls, the employees’ feeling for autonomy and competence might be less fulfilled. Moreover, this suggests that Boundary Systems and Diagnostic Control Systems can affect the working environment and decreases the fulfilling of their basic needs. A decrease in the fulfilment of the basic needs can result in less internalisation of extrinsic motivation. Therefore, I expect that another mechanism will motivate the employees by using negative controls, for example, external rewards or avoiding punishment. Therefore, I suggest that Boundary Systems and Diagnostic Control Systems are positively related to more controlled types of motivation, resulting in the following hypothesis:

H2: Boundary Systems and Diagnostic Control Systems are positively related to more controlled types of motivation.

2.4 The relationship between MCS and Motivation, affected by Generations

A generation is a group of people who share similar birth years and experiences of important life events at significant developmental phases (Kupperschmidt, 2000). In this research, I investigate how generations affect the relationship between MCS and motivation. Twenge and Campbell (2009) provide evidence that a generation is a meaningful psychological variable, as it captures the culture of an individual’s formation during a specific period. Based on the study of Twenge and Campell (2009), I use the concept ‘generation’ as a variable and the timeframes as defined by them. Three different generations are subject to this research, namely: Baby Boomers (1946-1964), Generation X (1965-1981) and Generation Y (1982-1999). The fourth generation, Generation Z (2000-2012), is not included in this research because Generation Z is barely represented in the current working environment (Hart, 2006; Dimock, 2019).

The study of Ballone (2007) suggests that there are differences between generations concerning the way of thinking and being motivated. For example, Baby Boomers prefer promotion, but Generation X prefer flexibility (Ballone, 2007). To better understand the motivation of different

(14)

14

generations, it is important to understand the characteristics of the different generations. Differences between generations can originate from their experiences and background during life (Ballone, 2007). Different experiences may lead to differences in characteristics, values and work ethics (Ballone, 2007; Glass, 2007). Characteristics and work values are typically very stable over lives and feature a generation (Smola & Sutton, 2002; McGuire, By & Hutchings, 2007). Generational characteristics do not form stereotypes but reflect on average, a member of one generation compared to another generation in general. There will always be exceptions that can be influenced by family, religion or schooling (Noble & Schewe, 2003).

The characteristics of a generation may influence members’ attitudes towards organisations, work values and desires, and management choices (Zemke, Rains & Filipczak, 2000; Applebaum, Serena & Shapiro 2004; Sayers, 2006; Westerman & Yamamura, 2007). To increase the change that motivation, behaviours and decisions of employees align with the organisation’s objectives and strategies, an MCS can be used (Merchant & Van der Stede, 2012). Petroulas et al. (2010) stated that different generations prefer differently designed MCS. Therefore, I investigate if the relationship between MCS and motivation is affected by each generation (Baby Boomers, Generation X and Millennials) separately.

Baby Boomers

The Baby Boomers are the generation born directly after World War II and are defined as the ones born between 1946 and 1964 (Twenge & Campbell, 2009). Their parents had financial challenges growing up during World War II. Petroulas et al. (2010) suggest that Baby Boomers find money and promotion important due to the financial challenges their parents experienced. Baby Boomers are assigned as employees with a strong work ethic to achieve money and promotions. They tend to be status-conscious, careful of authority and career-focused. Hence, they are aware of their social status (Ballone, 2007). Money, promotions and social status are examples of external rewards. Behaviour that is driven by external rewards has less possibility of internalisation of extrinsic motivation and therefore, get driven by more controlled types of motivation. Besides referring to social status, status can also refer to a professional position. To achieve a certain status within an organisation, an employee’s good performance has to be recognised by others. The employee gets a reward due to the achievement for good performance by, for example, promotion. Reward achievements are types of Diagnostic Control Systems because they can monitor if goals are achieved. Because Baby Boomers prefer reward

(15)

15

achievements, this suggests that Baby Boomers are more driven by Diagnostic Control Systems to get driven by more controlled types of motivation.

Baby Boomers are also characterised as careful of authority, which means that a person has respect for authority (Ballone, 2007). Authority is defined as the power or right to give orders and make decisions. When their managers impose activities, this likely decreases Baby Boomers’ feeling of choice and psychological freedom. By limiting an employee’s feeling of freedom and choices this could decrease the need for autonomy and therefore the fulfilment of the basic needs. This can result in less possibility of internalisation of extrinsic motivation. Since Boundary Controls limits the choice and behaviour of employees, I expect that Baby Boomers are more driven by Boundary Controls and therefore by more controlled types of motivation. Furthermore, some employees of the Baby Boom generation are close to retirement and want to slow down. Therefore, I expect that they are less driven by the emergence of new ideas and strategies and learning (Petroulas et al., 2010). This is an example of Interactive Control Systems.

In summary, Baby Boomers are being more characterized than other generations as very career-focused and careful of authority and less by the emergence of new ideas and strategies and learning. This suggests that they could be more driven by Diagnostic Control Systems and Boundary Systems, the negative controls. Therefore, I expect Baby Boomers to prefer Boundary Systems, Diagnostic Control Systems and more controlled types of motivation. This leads to the following hypothesis:

H3: The relationship between Boundary Systems, Diagnostic Control Systems and more controlled types of motivation is stronger for Baby Boomers compared to Millennials and

Generation X. Generation X

The group of people born between 1965 and 1981 are called Generation X (Twenge & Campbell, 2009). They have a strong preference for training and skill development, and they prefer to have things less formalised (Petroulas et al., 2010). Having things less formalised contradicts with the purpose of Diagnostic Control Systems, which suggest that employees are monitored by achieving specified goals. By having things less formalised, it contradicts with the purpose of Boundary Systems as well. Those negative controls set limits on the behaviour or freedom of the employee. Because Generation X does not prefer constraints on their

(16)

16

behaviour, I suggest that Generation X is less driven by negative controls and the controlled types of motivation. Even so, if Generation X prefers to have things less formalised, it suggests that they want to experience some freedom of choices. Freedom of choices possibly increases the fulfilment of the basic need for autonomy.

Generation X thinks self-improvement and challenges are essential and prefer to work alone rather than in a team (Petroulas et al., 2010). This could suggest that Generation X does not prefer shared values with other employees’, because they will have their focus on improving themselves instead of being interested in their colleagues. Creating shared values among employees is an example of Belief Systems. Because Generation X prefers to work alone rather than in a team, this suggests that Generation X is less stimulated by the use of Belief Systems and less driven by the feeling of relatedness. Self-improvement and challenges are achieved by learning and this is an example of Interactive Control Systems. Their preference for training and skill development likely increases the ability to complete a task with the right knowledge. Completing a task suggest that this can satisfy the basic need for competence.

In summary, by trying to satisfy the basic needs of competence and autonomy, the internalisation of extrinsic motivation is likely to become more autonomous motivated. Prefer to have things less formalised contradicts with the purpose of Diagnostic Control Systems and with the purpose of Boundary Systems as well. By preferring to work alone rather than in a team, this suggests that Generation X is less stimulated by the use of Belief Systems. Because Generation X prefers learning, this suggests that Generation X is more driven by Interactive Control Systems to get driven by more autonomous types of motivation. Therefore, by satisfying the basic needs and preferring Interactive Control Systems, I suggest that Generation X are more driven by more autonomous types of motivation. This leads to the following hypothesis:

H4: The relationship between Interactive Control Systems and more autonomous types of motivation is stronger for Generation X compared to Baby Boomers.

I except that the suggested relationship is stronger for Generation X compared to Baby Boomers because I expect that Baby Boomers are more related to the negative controls and more controlled types of motivation. Moreover, I expect a stronger relationship between the more

(17)

17

positive controls and more autonomous types of motivation for Millennials than for Baby Boomers.

Millennials

Generation Y refers to the generation born between 1982 and 1999 (Twenge & Campbell, 2009). Employees from Generation Y are mentioned as ‘Millennials’; the term is found by Strauss and Howe (Puffer, 1991). The study of Kultalahti and Viitala (2014) found that Millennials are more focusing on intrinsic than extrinsic motivators like (1) social relationships, (2) flexibility and the (3) opportunity to develop, than non-Millennials. Millennials prefer a social and good relationship with their colleagues and supervisors. This suggests that they feel more connected to their colleagues when they interact with them and are seen as a member of a group. This could result in satisfying the basic need for relatedness. The Millennial generation also prefers flexibility; this indicates that they are more driven by autonomy because they prefer to have a sense of choice and psychological freedom when carrying out an activity. Being driven by autonomy suggest that Millennials do not prefer negative controls like Boundary Systems and Diagnostic Control Systems, because it can limit their behaviour. When focussing on opportunities to develop, they have a strong preference for training and skill development (Petroulas et al., 2010). Completing a task with the right knowledge suggest that this can satisfy the basic need for competence. Interactive Control Systems are used to stimulate learning.

The Millennial generation has to deal with environmental and social sustainability issues, more than previous generations (Buahene & Kovary, 2003). A firm that has embedded environmental and social sustainability issues in their mission and vision and open communicates about it, is more likely to attract, motivate and retain Millennials (Petroulas et al., 2010). This suggests that Millennials prefer organisations that share a clear vision and mission throughout the organisation. Sharing a clear vision and mission are examples of Belief Systems.

In summary, by trying to satisfy the basic needs of competence, relatedness and autonomy, the internalisation of extrinsic motivation is likely to become more autonomous motivated. Being driven by autonomy suggest that Millennials do not prefer negative controls like Boundary Systems and Diagnostic Control Systems, because it can limit their behaviour. Because Millennials prefer an organisation that shares their vision and mission and invest in learning, this suggests that Millennials are more driven by Belief Systems and Interactive Control Systems. Therefore, by satisfying the basic needs and by preferring Belief Systems and

(18)

18

Interactive Control Systems, I suggest that Millennials are more driven by more autonomous types of motivation. This leads to the following hypothesis:

H5: The relationship between Belief Systems, Interactive Control Systems and more autonomous types of motivation is stronger for Millennials compared to Baby Boomers and

Generation X.

I expect that the suggested relationship is stronger for Millennials compared to Baby Boomers and Generation X, because of their expected fulfilment of the basic needs and their stronger preferences for the more positive controls.

The hypotheses, as presented in this chapter, result in the conceptual model as shown in Figure 3. This conceptual model is used as a framework for this study.

(19)

19

3. METHODOLOGY

In this study, quantitative research is the underlying method to answer the research question, which is done by a survey-study. First, I state the respondents, the survey instruments and the variables. Finally, data analyses are presented.

3.1 Respondents

The dataset used in this study is obtained from my supervisor and not collected by myself. The data are based on survey responses from four higher professional educational organisations. These organisations are relevant for this research because higher professional educational organisations use MCS to align employee’s behaviour and motivation with organisational strategies and objectives (Guenther & Schmidt, 2015). All employees in this sample have a master’s degree.

Distribution of the web-based survey occurred through mailings. The survey was sent out in August 2017 till December 2017 and the user language was English and Dutch. Reminders were sent to the respondents, but this varied between the four different organisations. One individual response was removed due to incompleteness, resulting in 315 usable individual responses.

3.2 Survey Instrument

The data used in this study is collected with the Qualtrics tool by applying convenience sampling. By using a survey, it is easy to obtain information describing the characteristics of a large sample of individuals (of different generations) relatively quickly (Ponto, 2015). Besides, the use of surveys is cost-effective, and data can be easily collected from a large number of respondents from different generations (Van der Stede, Young & Chen, 2005). In general, the higher the number of respondents, the more accurate the information that is derived from the survey will be (Van der Stede et al., 2005).

The survey is based on different validated surveys and uses a seven-point Likert scale question (Bedford & Malmi, 2015; Gagné et al., 2015; Bedford, Malmi & Sandelin, 2016; Kruis et al., 2016). Using validated surveys lowers the measurement error (Ponto, 2015). Research points out that seven response categories provide high reliability, internal consistency, and validity, and are preferred by respondents (Preston & Colman, 2000). One of the questions was adjusted because it used an interval from 1 till 12 instead of the seven-point Likert scale. A Cronbach's

(20)

20

Alpha (!) is used to measure the internal consistency between different questions of a set of scales. When combining different sub-questions, the reliability of the Cronbach Alpha’s has to be above .70 to be respectable (0.65-0.70 is minimally respectable; > 0.8 is very good). A summary of all the Cronbach’s of the different constructs is presented in Appendix A.

3.3 Early and Late Response

To test the early and late response, I divided the group into two groups based on the average starting date. By conducting a t-test on the early and late respondents, I found no significant differences in the scores of interest (Appendix B). Late respondents can be a proxy for non-response. Therefore, I conclude that a non-response bias is not a concern in this study.

3.4 Variables

3.4.1 Dependent Variable – Motivation

Prior studies have used a variety of measures for measuring employees’ motivation (Deci & Ryan, 2000; Gagné & Deci, 2005; Gagné et al., 2010; Van der Kolk et al., 2018). This study used the validated survey, The Multidimensional Work Motivation Scale, of Gagné et al., (2015). The MWMS is based on the continuum of the SDT framework and measures work motivation in the field of organisational behaviour (Deci & Ryan, 1985). They divided motivation into four different types that represent the continuum of motivation: External Regulation, Introjected Regulation, Identified Regulation and Intrinsic Regulation. They did not include amotivation to create a brief and practical survey and to focus on only active types of motivation. Because Integrated Regulation is complicated to distinguish from Identified Regulation, they omitted Integrated Regulation (Vallerand, Pelletier, Blais, Brière, Senecal, & Vallières, 1993). They used three survey-items for each of the subscales of motivation partially based on Vallerand et al. (1993). They examined the structure of the MWMS in a group of 3435 employees in seven different languages and nine countries. The response scale had seven points ranging from 1 (strongly disagree) to 7 (strongly agree). The responses to the items were averaged to produce the measure of External Regulation (! = .87); Introjected Regulation (! = .86); Identified Regulation (! = .82) and Intrinsic Regulation (! = .91).

3.4.2 Independent Variable – MCS

The MCS used in this research is based on Simons’ LoC framework (1995). The LoC consist of four control systems: Belief Systems, Boundary Systems, Diagnostic Control Systems and Interactive Control Systems (Simons, 1995). This study relies on the validated survey

(21)

21

administered on 217 business unit managers in the Netherlands in the recently published study of Kruis et al. (2016) to measure the Belief Systems and Boundary Systems. Belief Systems are measured with four questions about the organisation’s core values and mission statement, taken from Widener (2007). The responses to the items were averaged to produce the measure of Belief Systems. The scale was reliable (! = .85). Boundary Systems are measured with four questions about the organisation’s code of business conduct, also taken from Widener (2007). The response scale is a Likert-type 7-point semantic scale. The responses to the items were averaged to produce the measure of Boundary Systems and was reliable (! = .86).

To measure Diagnostic Control Systems and Interactive Control Systems this study relies on the validated survey, recently tested on 400 respondents from accounting firms in Australia, in the study of Bedford and Malmi (2015). Diagnostic Control Systems is measured through five items that are based on Henri (2006), Widener (2007) and the descriptions of Simons (1995). The responses to the items were averaged to produce the measure of Diagnostic Control Systems and was highly reliable (! = .97). Interactive Control Systems is measured through five items based on the formative measurement model outlined by Bisbe, Batista-Foguet and Chenhall (2007). The response scale is a Likert-type 7-point semantic scale. The responses to the items were averaged to produce the measure of Interactive Control Systems and was reliable (! = .89).

3.4.3 Moderating Variable – Generations

The moderating variable generations is measured based on the year of birth of the respondents. Only targeted respondents who satisfy the criteria with a year of birth between 1946 and 1999 are taken into account in this research. Based on the year of birth the respondents are divided into three different groups: (1) 1946 – 1964 (Baby Boomers), (2) 1965 – 1981 (Generation X), (3) 1982 – 1999 (Millennials).

3.4.4 Control Variables – Strategy, Culture and Structure

A control variable is a variable that is held constant or whose impact is removed in order to analyse the relationship between other variables without interference. The reason to include control variables in this research is to exclude alternative explanations while testing hypotheses. Chenhall (2003) reported a relationship between various contingencies and MCS and identified six groups of variables that have evolved historically as key to the understanding of MCS: (1) the external environment, (2) technology, (3) structure, (4) size, (5) strategy, and

(22)

22

(6) culture. I investigated for all these contingencies if the literature also describes a direct effect of one of those contingencies on motivation.

Organisational Culture is an elementary part of an organisation to integrate employees within the organisation (Yusof, Said & Ali, 2016). In other words, employees must be given a chance to participate and involve in the organisation. The management has to believe that their employees are capable of to fulfil their tasks with a good result (Yusof et al., 2016). This can support the need for competence and autonomy and can suggest that it will contribute to the fulfilment of the basic needs. When the basic needs are more fulfilled, this could lead to more internalisation of extrinsic motivation, and therefore the employee could become more autonomous motivated. The studies of Yusof et al. (2016), Towers (2006), Ritchie (2000) and Noordhoorn (2010) found a direct relationship between organisational culture and motivation and argue that strong organisational culture will often result in motivated employees.

Structure and Strategy are also important elements of an organisation and related to each other (Baines & Langfield-Smith, 2003). Strategy is a plan with which the objectives are achieved (Nickols, 2016). The Strategy can be shared by a mission and vision through an organisation (Nickols, 2016). Structure is how all tasks and departments within an organisation are divided and interrelated. It shows the various departments and associated responsibilities and the functions of employees who work within those departments (Nickols, 2016). The choice of the organisational structure depends on the strategic positioning chosen by the organisation (Wood, 2001). That organisational structure should follow a strategy (Baines & Langfield-Smith, 2003; Mintzberg, Ahlstrand, & Lampel, 2010).

The organisational Structure and Strategy can be achieved by rewards or behavioural influence of the employees (Bratton, Callinan, Forshaw & Sawchuk, 2010). Rewards are examples of extrinsic motivation. While constraining the behaviour of employees can decrease their feeling of freedom and consequently, the basic need for autonomy. Also, by sharing a vision and mission by employees, this could have an impact on their basic need for relatedness. Therefore, I suggest that organisational structure and strategy might impact the basic needs of the employees, which could impact the way an employee is motivated (Frey, & Osterloh, 2002).

In this research, I use Structure (! = .74), Culture (! = .68) and Strategy (! = .80) as control variables. To measure the control variables, I rely on the surveys of respectively Bedford et al.

(23)

23

(2016), Kruis et al. (2016), and Bedford and Malmi (2015). The variable Culture has a Cronbach Alpha of .68 and is therefore below the reliable value of 0.70. When comparing this Alpha to other studies, I found in a similar study also some Alpha’s lower than .70 (Kruis et al., 2016). Because the Cronbach is close to .70 and between the minimally respectable range of 0.65-0.70, I decided to accept the level and do not exclude an underlying question to increase the level of Alpha.

3.5 Data Analysis

Data analyses were performed to determine the direction and extent of the relationships among MCS and employees’ motivation. Followed by performing data analyses to measure the influence of generations on this relationship between MCS and motivation. Before analysing the data with IBM SPSS Statistics v.26, an examination of the sample characteristics was performed. Then, the data were further examined by investigating possible correlations between the variables and a comparison of the means. Thereafter, multiple regression analyses were performed controlling for relevant variables (Strategy, Culture and Structure).

Before analysing the data based on multiple regression (linear regression models that use more than just one explanatory variable) some key assumptions were checked to make it applicable. To apply for any type of ordinary least squares (OLS) regression the model is tested on multivariate and univariate normality, multicollinearity and homoscedasticity. The OLS requires that the errors between observed and predicted values (the residuals of the regression) should be normally distributed as shown in the Q-Q-Plots and confirmed by the results of the Kolmogorov-Smirnov test as presented in Appendix C. The univariate normality is also tested for all the main constructs included in the analysis by the Kurtosis value and skewness’s. All constructs are at acceptable values. The most extreme value of Kurtosis (1.26) does not exceed the accepted values of respect -7 and 7. The most extreme value for skewness (1.18) is also acceptable between -2 and 2.

Multicollinearity is checked by performing a correlation analysis, as shown in Table 4. Besides, the Variance Inflation Factor (VIF) is an indicator of multicollinearity between variables, and it is commonly accepted that this value should be below 10 (Kline, 2011). None of the VIFs in the regression analyses approach this threshold (max. value of 4.54). Homoscedasticity is performing a scatterplot of the residuals versus predicted values. As shown in Appendix D no

(24)

24

clear pattern in the distribution is shown; the data is heteroscedastic. Hence, there is no indication of multicollinearity and homoscedasticity so OLS can be applied.

Because the dependent variable and independent variable are continuous, Hypotheses 1 and 2 were examined by using the standard type of OLS regression. Dummy variables are created to estimate the moderating effect of generations for Hypotheses 3, 4, and 5 by moderated OLS regression models (Aiken & West, 1991). Because generations are a categorical moderator, this can be done by Hierarchical Least Squares (HLS). In the HLS-analyses, the independent variables were rescaled by centering of the means (subtracting their mean from each observation) to transform them into a comparable metric. Centering of the independent variables is applied to reduce collinearity in moderated multiple regression and to interpret the influence of the moderator and independent variables more easily (the zero becomes the mean). The interaction terms are the product of the standardised values (Aiken & West, 1991). The interaction terms and the used models are tested on significance.

(25)

25

4. RESULTS

This chapter describes the findings of the data analyses. First, the sample characteristics and descriptive statistics are presented. Then, a correlation matrix demonstrates the correlations between the different variables used in this study. After that, the results of the multiple regression analyses are presented. Finally, robustness checks and additional analyses are explained.

4.1 Sample Characteristics

Table 1 presents the characteristics of the sample. The sample is almost equally divided into male and female employees (respectively, 51.1%, 48.9%). Generations are not equally divided. In the sample, there are three times more respondents of the Baby Boomers and Generation X compared to the Millennials. Millennials are presented as 13.7% of the total sample, with an n of 43 this is sufficient to perform regression analysis (Hair, Black, Babin, Anderson & Tatham, 2014). Furthermore, the majority of the respondents is Dutch, has no supervisory or managerial role and has a fixed contract. Also, the type of employment is equally distributed.

Characteristics Frequency Percentage Characteristics Frequency Percentage

Gender Main Activities

Male 161 51.1% Educational Staff 225 71.4%

Female 154 48.9% Educational Support 90 28.6%

Generation Degree supervisor or manager

Baby Boomers 124 39.4% Full-time job as supervisor/manager 54 17.1%

Generation X 148 47.0% No job as supervisor/manager 232 73.7%

Millennials 43 13.7% Part-time job as supervisor/manager 29 9.2%

User Language Type of Contract

EN 36 11.4% Temporary contract 29 9.2%

NL 279 88.6% Fixed contract 286 90.8%

Organization Type of Employment

A 12 3.8% Part-time 135 42.9% B 52 16.5% Full-time 180 57.1% C 101 32.1% D 150 47.6% n = 315 Sample characteristics TABLE 1

(26)

26

4.2 Descriptive Statistics and Correlation Analysis

Table 2 presents the descriptive statistics of the data. Table 2 shows the means (µ) and modes (M) of the control, dependent and independent variables, including all usable respondents (n = 315). The mean (µ) shows the average value of all the employees for the specific variable. The mode (M) is the value that is answered the most by the employees on that specific variable.

The average score on the different subitems of the construct MCS is between 3.59 and 4.07 and is with an average of 3.83 slightly lower than the scale midpoint (4.00). There is more variety in the mean scores of the different types of motivation. The more controlled types of motivation (External Regulation and Introjected Regulation) are below the scale midpoint (µ = 1.92 and 3.57, respectively). The mode of External Regulation and Introjected Regulation is also far below the scale midpoint (M = 1.00; M = 2.00), this indicates that most of the employees' scores low on the questions corresponding to External Regulation and Introjected Regulation. On the other hand, the average score of more autonomous types of motivation (Identified Regulation and Intrinsic Regulation) is above the scale midpoint (µ = 5.67 and 5.53, respectively) and the score that the employees answered the most is also above the scale midpoint (M = 6.00; M = 5.00). This could suggest that the majority of the respondents answered that more controlled types of motivation less drive them compared to more autonomous types of motivation.

Because Millennials are less represented in the sample (n = 43), they will less influence the average of the group total compared to Baby Boomers and Generation X (n = 272). Therefore, Table 3 shows the means and modes for all the three different generations separately. On

Variable Mean Mode SE Minimum Maximum

Structure 4.44 4.80 1.20 1.00 7.00

Culture 4.36 4.60 .68 1.80 6.00

Strategy 4.41 4.90 .78 1.80 6.40

Belief Systems 4.07 4.00 1.15 1.00 7.00

Boundary Systems 3.71 2.00 1.47 1.00 7.00

Diagnostic Control Systems 3.94 4.00 1.31 1.00 7.00 Interactive Control Systems 3.59 4.00 1.21 1.00 7.00

External Regulation 1.92 1.00 .95 1.00 5.00 Introjected Regulation 3.57 2.00 1.15 1.00 7.00 Identified Regulation 5.67 6.00 1.07 1.00 7.00 Intrinsic Regulation 5.54 5.00 0.06 1.33 7.00 n = 315 TABLE 2 Descriptives Statistics

(27)

27

average, Baby Boomers and Generation X points a higher mean on Belief Systems and Interactive Control Systems (lowest µ = 3.86 for the positive controls) compared to Boundary Systems and Diagnostic Control Systems (highest µ = 3.79 for the negative controls). The majority of the Baby Boomers valued all the MCS types on the midscale, and this could suggest that they do not have an extreme preference for one specific type of MCS (M = 4.00 for all LoC). The biggest difference in the average and modes of the values of the MCS variables are for Boundary Systems. Generation X has the lowest score on Boundary Systems (µ = 3.44; M = 2.00), especially compared to Millennials Systems (µ = 4.10; M = 5.25). Millennials have a higher mean for Boundary Systems and Interactive Control Systems (lowest µ = 4.10) compared to Diagnostic Control Systems and Belief Systems (highest µ = 4.07). For mostly all the MCS and types of motivation, Millennials have a higher mean compared to Baby Boomers and Generation X.

The biggest difference in the average and modes of the values of the different types of motivation are for Introjected Regulation. Baby Boomers have a lower score on Introjected Regulation (µ = 3.46; M = 2.00), especially compared to Millennials (µ = 3.80; M = 5.00). For all generations, Identified Regulation and Intrinsic Regulation (lowest µ = 5.43 for more autonomous types of motivation) are higher valued compared to External Regulation and Introjected Regulation (highest µ = 3.80 for more controlled types of motivation).

Table 4 presents the correlations (r) among the independent, dependent and control variables. Belief Systems, Interactive Control Systems and Boundary Systems are positive correlated to Intrinsic Regulation (r = .35, p < .001; r = .13, p < .05; r = .17, p < .001). Belief Systems are

Variable Mean Mode Mean Mode Mean Mode

Structure 4.44 4.80 4.38 4.80 4.65 4.60

Culture 4.41 4.60 4.32 4.70 4.38 4.10

Strategy 4.42 4.80 4.41 4.20 4.44 4.00

Belief Systems 4.10 4.00 4.04 4.50 4.07 4.75

Interactive Control Systems 3.86 4.00 3.91 4.00 4.28 4.25

Boundary Systems 3.59 4.00 3.44 2.00 4.10 5.25

Diagnostic Control Systems 3.79 4.00 3.60 4.00 3.88 5.00 External Regulation 1.76 1.00 1.96 1.00 2.24 1.00 Introjected Regulation 3.46 2.00 3.60 4.50 3.80 5.00 Identified Regulation 5.60 5.00 5.71 6.00 5.74 6.00 Intrinsic Regulation 5.61 6.00 5.51 5.00 5.43 5.00

n = 315

Baby Boomers Generation X Millennials Descriptives of the Means and Modes per Generation

(28)

28

negatively correlated to External Regulation (r = -.12, p < .05). In addition, Boundary Systems are negatively correlated with Identified Regulation (r = -.13, p < .05).

For the different generations, Millennials are positive correlated to Diagnostic Control Systems and Intrinsic Regulation, and negative with Identified Regulation (r = .17, p < .01; r = .14, p < .05; r = -.04, p < .01). For Generation X, there is a negative correlation with Diagnostic Control Systems (r = -.12, p < .05). Baby Boomers are negative correlated with Intrinsic Regulation (r = -.14, p < .05).

Further, Structure, Culture and Strategy show a significant positive correlation with all types of MCS (lowest r = .15, p < .01), that substantiate the contingency theory as described in the previous chapter. Structure, Culture and Strategy are only positively correlated to Intrinsic Regulation (r = .19, r = .29, r = .26, p < .01).

4.3 Regression Analysis

As presented in Table 4, a few significant correlations are found between the variables. To further substantiate the different hypotheses, multiple regressions analyses are performed to test the association between the variables. I defined three models for the four subtypes of motivation to analyse the hypotheses. Model 1 is the basic model that only includes the three control variables (Structure, Culture and Strategy). Control variables are added to the regression to exclude alternative explanations while testing hypotheses. Model 2, is presented to test the first two hypotheses, including the dependent variables (four types of MCS) and control variables. In Model 3, the interaction terms are added. This model 3 presents the moderating effect of

Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14

1. Structure 1

2. Culture .42** 1

3. Strategy .37** .61** 1

4. Belief Systems .33** .50** .46** 1 5. Interactive Control Systems .27** .45** .43** .51** 1 6. Boundary Systems .19** .26** .30** .41** .45** 1 7. Diagnostic Control Systems .15** .31** .35** .38** .77** .42** 1 8. External Regulation -.12* -.10 .05 -.12* .03 -.07 .01 1 9. Introjected Regulation -.09 -.02 .05 -.05 -.01 -.10 -.05 .50** 1 10. Identified Regulation -.01 .05 .05 .06 -.08 -.13* -.05 .04 .41** 1 11. Intrinsic Regulation .19** .29** .26** .35** .13* .17** .09 -.24** .03 .42** 1 12. Millennials .07 .01 .01 -.00 .05 .10 .17** .06 .03 -.04** .14* 1 13. Generation X -.05 -.06 -.01 -.02 -.07 -.02 -.12* -.02 .03 -.02 .04 -.38** 1 14. Baby Boomers .00 .05 .00 .02 .04 -.05 .00 -.06 -.05 -.05 -.14* -.32** -.76** 1 Correlation is significant at; ***p < 0.001 **p < 0.01 *p < 0.05

TABLE 4 Correlation Analysis

(29)

29

generations on the relation between MCS and the continuum of motivation and is the main model to answer the remaining hypotheses. Based on performed F-tests for every model, Model 1, 2 and 3 are significant for testing External Regulation and Intrinsic Regulation (p < 0.01 and p < 0.001). The most extreme types of motivation on the continuum. Table 5 shows the total results of the regression analyses for the three models performed on the continuum of motivation.

4.3.1 Model 1 - Control Variables

In Model 1, Structure, Culture and Strategy are included as control variables (this is substantiated in the methodology sector of this study). They have a significant positive correlation with MCS, as presented in Table 4. The positive correlation substantiates the contingency theory. Because I also expect that Structure, Culture and Strategy have a direct impact on motivation, those control variables are included in the model. Structure, Culture and Strategy have a significant effect on External Regulation (β = -.11, β = -.23, β = .25, p < .05). That indicates that the more open, co-operative, unregulated and free the Culture and Structure will be, it could lower more controlled types of motivation. On the other hand, a low cost, innovative and customer focus Strategy could have a positive effect on more controlled types of motivation. In addition, Culture has a significant effect on Intrinsic Regulation (β = .33, p < .05), indicating that for example, a co-operative Culture has a positive effect on more autonomous types of motivation.

4.3.2 Model 2 – Main Effect of MCS on Motivation

Model 2 tests the influence of different types of MCS on the continuum of motivation as presented in Table 5. For Interactive Control Systems, I found a significant positive relationship with External Regulation (β = .15, p < .05). This could imply that controls that are used to stimulate organisational learning and innovation have a positive impact on the more controlled types of motivation. Belief Systems have a significant negative relationship with External Regulation (β = -.13, p < .05) and a significant positive relationship with Identified Regulation and Intrinsic Regulation (β = .13, β = .27, p < .05, p < 0.001). The significant findings indicate that controls that are used to share a mission and vision are positively related to more autonomous types of motivation and negatively related to more controlled types of motivation. Hypothesis 1 suggests that Belief Systems and Interactive Control Systems are positively related to more autonomous types of motivation. Hypothesis 1 can be partly supported because

(30)

30

I only found a significant relationship between Belief Systems and more autonomous types of motivation.

I did not find any significant effect of Diagnostic Control Systems on any type of motivation. For Boundary Systems I found a significant negative relationship with Identified Regulation (β = -.11, p < .05), indicating that setting limits on opportunity-seeking behaviour lowers more autonomous types of motivation. Hypothesis 2 suggest that Boundary Systems and Diagnostic Control Systems are positively related to more controlled types of motivation. I did not found any evidence to support Hypothesis 2.

(31)
(32)

32 4.3.3 Model 3 - Moderating Effect of Generations

Model 3 tested the influence of different generations on the relationship between MCS and motivation, and this is shown by the two-way interaction terms in Table 7. The two-way interaction terms explain how much the coefficient for the relationship between MCS and motivation changes for each generation, the moderating effect. I found no significant moderating influence of one of the generations on the relationships between the different types of MCS and motivation. Therefore, I did not found any evidence to support Hypothesis 3, 4 and 5. Whereas I found no significant moderating relationship, Baby Boomers have a significant negative direct effect (p < 0.05) on External Regulation (as shown in Table 6). That could suggest that Baby Boomers are significantly less driven by more controlled types of motivation, for example, external rewards, compared to Millennials.

IV Generation compared to DV β SE β 95% CI External Regulation .02 .16 [-.29, .33] Introjected Regulation -.07 .26 [-.59, .45] Identified Regulation -.27 .18 [-.62, .90] Intrinsic Regulation -.21 .18 [-.56, .14] External Regulation .11 .17 [-.22, .43] Introjected Regulation .07 .28 [-.48, .62] Identified Regulation -.08 .19 [-.46, .30] Intrinsic Regulation -.32 .19 [-.69, .50] External Regulation -.03 .22 [-.47, .40] Introjected Regulation -.46 .37 [-1.18, .27] Identified Regulation .15 .25 [-.35, .65] Intrinsic Regulation .28 .25 [-.21, .77] External Regulation -.21 .24 [-.68, .26] Introjected Regulation -.65 .40 [-1.44, .14] Identified Regulation -.19 .28 [-.73, .36] Intrinsic Regulation .03 .27 [-.5, .56] External Regulation .16 .13 [-.90, .40] Introjected Regulation .10 .21 [-.31, .51] Identified Regulation .09 .14 [-.19, .37] Intrinsic Regulation .03 .14 [-.25, .31] External Regulation .16 .13 [-.11, .42] Introjected Regulation .21 .22 [-.23, .64] Identified Regulation .18 .15 [-.12, .48] Intrinsic Regulation -.05 .15 [-.34, .25] External Regulation -.16 .19 [-.54, .22] Introjected Regulation .18 .32 [-.46, .81] Identified Regulation -.11 .22 [-.55, .32] Intrinsic Regulation -.21 .22 [-.63, .22] External Regulation .06 .21 [-.35, .48] Introjected Regulation .34 .35 [-.34, 1.30] Identified Regulation -.12 .24 [-.59, .36] Intrinsic Regulation .13 .23 [-.34, .59] n = 315; ***p < 0.001 **p < 0.01 *p < 0.05

Diagnostic Control Systems Millenials vs.

Gen X

Baby Boomers Gen X

Baby Boomers

Interactive Control Systems Millenials vs.

Gen X

Baby Boomers

Boundary Systems Millenials vs.

Gen X

Baby Boomers

Belief Systems Millenials vs.

TABLE 7 Regression Analysis

(33)

33

4.4 Robustness Checks

4.4.1 Results separated by Type of Employment

A way to check the robustness of the used regression model is by splitting the sample into different groups using dummy coding. Dummy coding implies that one group is a reference group (Cohen, 2003). That can, for example, be done by splitting the observations on the moderator below, above or equal to the median (this represents the reference group). Because in this study, generations is a categorical moderator, it is not possible to split the sample into different groups by the median. Therefore, I split the group between part-time and full-time employment (n = 135; n = 180). The full-time employment group is the reference group. The results of this robustness test are presented in Table 1 in Appendix E. The results confirm a significant positive relation between Belief Systems and more autonomous types of motivation (β = .57, p < .01; β = .49, p < .05). Those results are very much similar to the main model and could underpin that the significance of the moderating effect of generations does not improve for this study.

4.4.2 Age as a Moderating Effect

A generation is a group of people that share similar birth years and experiences of important life events at significant developmental phases (Kupperschmidt, 2000). Three different generations are used in this research; Baby Boomers (1946-1964), Generation X (1965-1981) and Generation Y (1982-1999). Whereas the timeframes differ for the various definitions of the three generations, all generations correspond to certain ages. Age is the length of time that a person has lived. Therefore, I performed a robustness test using the continuous variable Age as a moderating effect in the regression analysis. The results of this robustness test are presented in Table 2 in Appendix E. The results confirm a significant positive relationship between Belief Systems and more autonomous types of motivation (β = .14, p < .01; β = .25, p < .001) but also a negative relationship with External Regulation (β = -.12, p < .01). Age itself has a small significant negative relationship with External Regulation (β = -.01, p < .01). For the other types of motivation, the effect of Age is almost zero for every year a person gets older and not significant. Suggesting that Age has, in line with generations, no moderating effect on the relationship between MCS and motivation. This test provides robustness to the non-significant effect of generations on the suggested relationship between MCS and motivation for this study.

Referenties

GERELATEERDE DOCUMENTEN

The nine differentiated test areas are: the total emotional intelligence quotient, empathy, self-knowledge, self-control, self-motivation, self-esteem, emotional

When a management control package is perceived as predominately negative, hence more constraining controls relatively to facilitating controls, it could negatively affect management

Therefore, I expect that boundary systems have a negative effect on autonomous motivation as long as they are more strongly present in the MCS package than the MCSs

In particular, the effects of Simons’ levers-of-control (i.e. beliefs systems, boundary systems, diagnostic control systems and interactive control systems) for two different

Moreover, dynamic tension has a positive impact on autonomous motivation under an organic structure, and a negative impact when the organizational structure is

All the devices and systems that are used by managers to ensure the consistency of the decisions and behaviour of employees with the organization’s strategies and

To test the influence that management control systems have on corporate sustainability performance and the influence that the proposed interaction effects have on these

It is important to check whether these interactive controls are experienced as a negative influence on employees intrinsic motivation Mallin &amp; Bolman Pullings (2009) , or