The effect of the mismatch between personal traits and the role of management accountants on stress.

(1)

0

MASTER THESIS S3535274

Faculty Economy and Business administration March 1st, 2019

MSc A&C Accountancy EBM869B20.2018-2019.1B

The effect of the mismatch between personal

traits and the role of management accountants

on stress.

Sveva Stummel

University of Groningen

Abstract

This study examines to what extent the match between the different types of roles of a controller (e.g. bean counter role or business partner role) and their personal traits consistent with that role affect their level of stress. The data for this research is collected from 146 Dutch controllers who filled in an online questionnaire. The results suggest a negative association between the match consisting of the role a controller performs and the personal traits he possesses, and the level of stress. This means the more a controller performs the role of a bean counter, possessing the personal traits accurateness and formality, the less stress he experiences. The negative association between the match and the level of stress also means the more a controller performs the role of a business partner possessing the personal traits consistent with that role, the less stress he experiences. Thereby, the results show that most respondents performing a business partner role possess the personal traits communicative, linguistic proficiency, teamworker, flexibility, proactiveness and creativeness. Most respondents performing a bean counter role possess the personal traits accurateness and formality. The findings of this research create awareness for the HR-department about which controller will fit best in a certain role with certain personal traits and will experience as little stress as possible

Keywords: Level of stress, management accountants, controller’s roles, personal traits, accurateness, reservedness, conservativeness, flexibility, communicativeness, match, mismatch

Supervisor UG: W.G. De Munnik

(2)

1

he controller’s social skills become more important for controllers, because the role of a controller is changing from a financial director, with hierarchical power (bean counter role), to a management advisor (business partner role), according to a recent article in Dutch media (PRIMO Nederland, 2018). Since 1980, the role of the controller is changing from a bean counter role to a business partner role. The main responsibilities of controllers are the firm’s internal control environment and the financial report (Simons, 1990; Bouwens & Abernethy, 2000; Malmi & Brown, 2008; Macintosh & Quattrone, 2010). The reason for the changing role is Hopper’s criticism in 1980 on the controller’s passivity in management, in what is now called a bean counter role (Hopper, 1980; Grandlund & Lukka, 1997; Byrne & Pierce, 2007; Järvenpää, 2007; Baldvinsdottir, Burns, Nørreklit & Scapens, 2009;.Goretzki, Straus & Weber, 2013). Hopper argued that the controller’s “inability to meet service expectations were perhaps related to ambiguity and stress emanating from the accounting workflows” (p.401). This is the beginning of a new role of the controller, in which the controller has a more active interpersonal relationship with the management, which is called a business partner role. So, a bean counter focuses on the past (e.g. differences between actual and expected results) (Goretzki et al., 2013), while the business is more forward-looking and involved in assisting management in their decision-making process (Sorensen, 2009; Goretzki et al., 2013). Goretzki et al. (2013) define four roles a controller can perform, namely a bean counter, a reporter, a navigator and a business partner. Since the bean counter and the business partner are most apart from each other and their differences are highly studied, this research focuses on these two controller roles (Granlund & Lukka, 1997; Byrne & Pierce, 2007; Järvenpää, 2007; Baldvinsdottir et al., 2009;.Goretzki et al., 2013).

The profession of the management accountant (after this: the controller) is one of the most stressful positions in America (Fogarty, Singh, Rhoads & Moore, 2000. p. 32). Also Dutch controllers experience a higher level of stress than most other positions, according to Control Magazine (2017). According to Collings and Murray (1996) a prolonged period of stress could lead to a burnout. Worldwide, the number of burnouts significantly increased in recent years (Financial Times, 2018). In the Netherlands, 15 per cent of the working population experienced a burnout, while this was 12.8 percent in 2013 (Financieel Dagblad, 2017; CM, 2017). Stress is an emotional and physical response to internal and external influences that disrupt an individual’s body and state of well-being (Selye, 1975; Lloyd, King & Chenoweth, 2002; Middlebroeks & Audage, 2008). Emotional responses are, for example, variances in mood, having difficulty to relax and caring less about things/others (Bolger, DeLongis, Kessler & Schilling, 1989). Examples of physical responses are increased heart rate and

(5)

4 blood pressure, low energy and frequent colds and infections (McEwen, 2008). An increase in the stress level of the controller can lead to misreporting, e.g. biasing of forecasts and earnings smoothing, thereby lowering their job performance (Hammer & Tosi, 1974; Schuler, 1975; Cohen, 1980; Beehr, 1981; Maas & Matějka, 2009; Smith, Rasmussen, Mills, Wefald & Downey, 2012).

Fogarty et al. (2000) indirectly suggest that the controllers’ stress level is influenced by the type of role he performs. Furthermore, the level of stress is influenced by the employee’s environment, which includes the work environment (Chenhall & Langfield-Smith, 1998; Byrne & Pierce, 2007; De Loo, Verstegen & Swagerman, 2011; Ebstrup, Eplov, Pisinger & Jørgensen, 2011). So, since the type of role a controller performs is part of his work environment, and the work environment affects stress, the type of role a controller performs affects the controller’s level of stress. Besides the controller’s role, personal traits are also associated with the level of stress (Tett & Burnett, 2003; Miron, Erez & Naveh, 2004; Byrne & Pierce, 2007; De Loo et al., 2011; Ebstrup et al., 2011). Kobasa, Maddi & Puccetti (1982); Cooper & Payne (1991); Smith & Mackenzie (2006); Conner-Smith & Flachsbart (2007) conclude that a relation between personal traits and stress exists. This research focuses on the match between the role a controller performs and his personal traits, because both influence the level of stress and are relevant to examine. A match occurs when a controller possesses the personal traits belonging to the role he performs. A mismatch occurs when a controller does not possess the personal traits belonging to the role he performs.

Thus, earlier research found a causal relationship between the type of role a controller has, as well as the controller’s personal traits, and the level of stress separately (Hopwood, 1994; Granlund & Lukka, 1997; Granlund & Lukka, 1998; Friedman & Lyne, 2001; Byrne & Pierce, 2007; Järvenpää, 2007; Baldvinsdottir et al., 2009;.Goretzki et al., 2013). Though, no earlier research examines the match between the role a controller performs and the personal traits he possesses as an influencing factor on his level of stress. This study aims to fill this research gap. Therefore, the main contribution of this study is the examination to what extent a match or a mismatch between a controller’s role and his/her personal traits influences the level of stress. This leads to the following research question: To

what extent does the match between the role of controllers and their personal traits affect their level of stress? To answer this question, the following sub questions are formulated:

1. Which controller’s roles exist?

2. What are the personal traits consistent with a certain role?

3. When does a match between a controller’s role and his personal traits exist? 4. How does the match affect the level of stress?

(6)

5 This study has two main contributions which add to the scientific relevance and the practical relevance. Firstly the findings of this study contribute to the literature on the role performed by management accountants, more specific on the match between the role a management accountant performs and the personal traits inherent to this role in relation to his level of stress. The results might be used to distinguish more roles or a combination of roles in further research. Secondly, this research creates awareness for companies about the match between the controller’s role and their personal traits and their level of stress, and about what personal traits the ideal candidate for a particular job as controller must possess. This gives their human research department the possibility to select the candidate who will best fit the role he has to perform leading to a lower level of stress. This research will be conducted by filling in a questionnaire online by Dutch controllers working for different companies in different sectors.

In addition to the introduction this study consist of a chapter with theoretical framework which is based to formulate the hypotheses (chapter two). In chapter three the research method will be discussed. In chapter four the results of the research will be presented. Lastly, chapter five provides the conclusion, the discussion, implications and limitations of the research.

2. THEORETICAL BACKROUND

In this section the theoretical framework based on the research question will be set out. This research combines the three concepts roles, personal traits and the level of stress. This theoretical framework addresses the definitions, assumptions and implications to these concepts. Based on the definitions, assumptions and implications of the three concepts, the hypotheses and a conceptual model will be developed.

2.1. The role of a controller

“A role is a useful tool for examining issues as it links three central concepts: expectations – what the controller and others believe he should do; behaviour – what he actually does; structure – how his position is linked to others, thus shaping his role set” (Hopper, 1980, p. 401-402). All roles in an organisation are explained by their tasks and responsibilities, which will be used to define the role of a controller as well. A controller can be defined as an individual who reports financial information about the organization, controls the internal control practices, and provides information and recommendations that can be used by decision makers during decision-making (Simons, 1990; Bouwens & Abernethy, 2000; Malmi & Brown, 2008; Macintosh and Quattrone, 2010). One of the factors influencing the controller’s stress level is the type of role a controller performs, which is part of the working environment of a controller (Chenhall & Langfield-Smith, 1998; Byrne & Pierce,

(7)

6 2007; De Loo et al., 2011; Ebstrup et al., 2011). The role a controller can perform will be described next.

Four different types of roles a controller can perform, exist: bean counter, reporter, navigator and business partner (Goretzki et al., 2013). The roles in between are mostly a combination of the bean counter and business partner role, and are called the hybrid roles (De Loo et al., 2011). This research focuses on the role of a bean counter and the role of a business partner, which comprise responsibilities, tasks and personal traits which are most distinct considering all the roles, as can be seen in Table 1 and Table 2 (which will be discussed in section 2.2). In addition, much literature is available since the differences between these roles are a highly studied subject compared to the hybrid roles (Granlund & Lukka, 1997; Byrne & Pierce, 2007; Järvenpää, 2007; Baldvinsdottir et al., 2009; Goretzki et al., 2013).

As can be seen in Table 1, different aspects, comprising responsibilities and tasks, are distinguished to define the role of a bean counter and a business partner. Controllers have two types of responsibilities within an organization: functional responsibility and local responsibility (Sathe, 1983; Maas & Matějka, 2009). Functional responsibility, which accords to the controller’s role as a bean counter, is the responsibility of a controller to report to the management with accurate financial information and to ensure that the internal control of the organization is in line with corporate policy and procedures. Local responsibility, which accords to the controller’s role as a business partner, is the responsibility to help management make decisions about the continuity of the organization (e.g. investment decisions or production decisions) (Maas & Matějka, 2009).

The tasks of a controller leaning to a bean counter role are tasks that focus on the past. Composing reliable financial statements, reports with financial information and the comparison of the actual numbers with proposed numbers including the reasoning behind it are tasks that focus on the past, because the activities the controller reports have already taken place (Siegel, 2003; Järvenpää, 2007; Lambert & Sponem, 2012; Goretzki et al., 2013). The tasks of a controller leaning to a business partner role are tasks that focus on the future, the continuity of the company by helping management with decision making on the policy of the company. So, a business partner is more business-oriented, is mostly looking forward, is a member of the management team and is a change agent (Granlund & lukka, 1997; Siegel, 2003; Järvenpää, 2007; Lambert & Sponem, 2012; Goretzki et al., 2013).

(8)

7

Table 1

Characteristics MA-roles (Granlund & Lukka, 1997; Grandlund & Lukka 1998, p. 202)

Aspect Bean counters Business partners

Temporal orientation Mostly looking to the past Mostly looking to the present and the future

Knowledge of the business in which the firm operates

Not expected Expected

The primary aim of the communication

Fulfilling of formal information requirements

Active attraction in order to get the message through

Communication channel used Mostly via written reports Written reports and also a lot of personal

(face-to-face) communication

Working method Mostly on his/her own Team work

General operating style Information collector and

processor

Member of the management team and a change agent

Oriented Around number crushing Business

Attitude towards management Passive Interactive

Cross-functional appreciation Limited; based often on fear High for an active and capable person

Responsibility Functional Local

Felt scope of responsibility Narrow Wide

Mainly activities Report financial information Help with decision making of management

team and report financial information

2.2. Personal traits of bean counters and business partners

According to Tokar, Fischer and Mezydlo Subich (1998) a relationship exists between the role a controller performs and the personal traits he possesses. When the task and responsibilities are clear, the controller that best fits the job is a controller with the personal traits who best fits to that certain role (Granlund & Lukka, 1997; Friedman & Lyne 2001).

Different research examined personal traits of controllers (Hopwood, 1994; Granlund and Lukka, 1997; Friedman & Lyne, 2001; Byrne and Pierce, 2007; De Loo et al., 2011). Every individual, in this case a controller, has his own unique personal traits (Allport, 1961; Weinberg & Gould, 1999; Feist & Feist, 2006) and performs a certain type of role as controller. According to Conner-Smith and Flachsbart (2007) and John & Srivastava (1981) personal traits are part of someone’s personality, leading to personality defined as “characteristic patterns of thoughts, feelings, and behaviors over time and across situations” which are influenced by environmental factors.

Hopwood (1994), Granlund and Lukka (1997), Friedman & Lyne (2001), Byrne and Pierce (2007) and De Loo et al. (2011) describe the most common personal traits of a controller and personal traits of controllers in different roles, like bean counters and business partners. The personal traits almost all controllers possess according to Granlund and Lukka (1997) are reliability, honesty,

(9)

8 purposefulness, prudence, thrusting and pragmatism. Most controllers in all type of roles possess these personal traits, because the role implies that management shares confidential information with them. For now, the research focuses on the different personal traits between the two roles. The most common personal traits of bean counters according to Hopwood (1994), Granlund and Lukka (1997), Friedman & Lyne (2001), Byrne and Pierce (2007) and De Loo et al. (2011) are accurateness, withdrawal, formality, modesty, conservativeness and reservedness. These personal traits are most common for a bean counter, because people possessing these traits are mostly working flawlessly and discreetly, focusing on one thing at the time, wanting to hold on to the past, and obeying rules and legislation. For business partners the most common personal traits according to Hopwood (1994), Granlund and Lukka (1997), Friedman & Lyne (2001), Byrne and Pierce (2007) and De Loo et al. (2011) are linguistic proficiency, proactiveness, creativeness, flexibility, teamworker and communicative. These personal traits are most common for business partner instead of bean counter because they are more business-oriented, which means taking into account the future when providing management with creative ideas and working together with management which implies more communication than in the bean counter role. An overview of the personal traits of a controller in general and a controller as bean counter and business partner is presented in Table 2. This does not imply that each type of controller (bean counter or business partner) possesses all the personal traits presented in the Table 2, because each controller is unique in his own variety of personal traits and could possess other personal traits than presented in Table 2 (Allport, 1961; Weinberg & Gould, 1999; Feist & Feist, 2006).

Table 2

Personal traits of controllers

Bean counters Business partners Controllers

Accurate Linguistic Reliable

Withdrawal Proactive Honest

Quiet/Reserved Creative Purposeful

Formal Flexible Prudent

Modest Teamworker Thrusting

Conservative Communicative Pragmatic

Considering the aspects presented in Table 1, the type of role a controller performs, might influence the desired personal traits consistent with this role, as shown in Table 2. This suggests the following hypotheses:

(10)

9 Hypothesis 1: The more the controller operates as a bean counter, the more he possesses the

personal traits of accurateness, withdrawal, reservedness, formality, modesty, and conservativeness.

Hypothesis 1a: The more the controller operates as a bean counter, the more he possesses the

personal traits of accurateness and formality.

Hypothesis 1b: The more the controller operates as a bean counter, the more he possesses the

personal traits of reservedness, modesty and withdrawal.

Hypothesis 1c: The more the controller operates as a bean counter, the more he possesses the

personal trait of conservativeness.

Hypothesis 2: The more the controller operates as a business partner, the more he possesses the

personal traits of linguistic proficiency, proactiveness, creativeness, flexibility, teamworker, and communicative.

Hypothesis 2a: The more the controller operates as a business partner, the more he possesses

the personal traits of flexibility, proactiveness and creativeness.

Hypothesis 2b: The more the controller operates as a business partner, the more he possesses

the personal traits of communicative, linguistic proficiency and teamworker.

Hypotheses 1a, 1b, 1c, 2a and 2b are formulated to test if the controller in a certain role possesses a part of the personal traits consists with his role. If data supports the hypotheses, a match will exist between the role a controller perform a the personal traits consistent with that role. When a controller (bean counter or business partner) does not possess the expected personal traits, a mismatch can occur between the personal traits of a controller and the role the controller performs. In the following section 2.3 this match and mismatch will be explained.

2.3. Match between the role a controller performs and his personal traits

In this research it will be examined to what extent a match exists between a controller’s role and his personal traits consistent with this role. Baird (1970) mentions this match in slightly different words. Namely, he concluded that a vocational fit involves a match between an employee’s personality and the occupation requirement of the job. In this research, personality is defined by among other things the personal traits of an individual and occupation requirements is defined as the tasks and responsibilities of a controller in a certain role. It can be stated that a vocational fit involves a match between a controller’s personal traits and the role this controller performs. So, a match exists when the controller in the role of a bean counter possesses all or partly the expected personal traits of a bean counter (e.g. accurateness, withdrawal, reservedness, formality, modesty, and conservativeness) to a certain extent. A match also exists if a controller in the role of a business partner possesses all or

(11)

10 partly the expected personal traits of a business partner (e.g. linguistic proficiency, proactiveness, creativeness, flexibility, teamworker, and communicative) to a certain extent. So, a controller in a certain role does not possess all expected personal traits to the same extent.

On the contrary, from a bean counter perspective, a mismatch occurs when a controller who performs his role as a bean counter while being more linguistic, proactive, creative, flexible, a teamworker and communicative and who is less, not or just slightly accurate, withdrawal, quiet/reserved, formal, modest, and/or conservative. From a business partner perspective, a mismatch is a controller who performs his role as a business partner while being more accurate, withdrawn, reserved, formal, modest and/or conservative and who is less, not or just slightly linguistic, proactive, creative, flexible, a teamworker and/or communicative.

2.4. Stress

Stress is defined differently by different scholars (Fink, 2010). Hans Selye (1956, p. 1) came up with the first definition of stress: “Stress is the nonspecific response of the body to any demand”. He also argued that stress is not the same as emotional arousal or nervous tension, therefore he used in his definition of stress: nonspecific response (Selye, 1975). According to Lloyd et al. (2002, p. 256) stress can be defined as: “the emotional and physical reactions to stressors. This is a demand, a situation or a circumstance that disrupts a person’s equilibrium that initiates the stress response of increased autonomic arousal.” In other words, a stressor is a stimulus that causes stress (e.g. activity or event). Stress could also be defined as “internal and external influences that disrupt an individual’s normal state of wellbeing”, according to Middlebrooks and Audage (2008). These internal and external influences could affect a person’s health and could induce emotional stress. Thereby these influences could lead to several physical changes, e.g. a dramatic rise of hormone levels and/or elevated blood pressure (Middlebrooks & Audage, 2008). In this research the following definition of stress will be used: an individual’s response on external influences that disrupt his body and normal state of wellbeing. The symptoms signaling an employee experiences stress could be physical (e.g. chronic headaches, fatigue) and mental/emotional (e.g. memory problems) and could be seen in changes in personal habits (e.g. nervous nail biting) (Hollien, 1970).

Wright, Cropanzano & Bonett (2007) and Adaramola (2012) state that stress influences job satisfaction and job performance. Job stress is the emotional and physical response on “a conflict between job demands on the employee and the amount of control an employee has over meeting these demands” (Adaramola, 2012). This conflict could also cause disruption in an individual’s body and normal state of wellbeing. In other words, job stress is negatively related to personal wellbeing, which is together with personal satisfaction positively linked to job performance (Wright et al., 2007). This

(12)

11 means that job performance will be higher when personal wellbeing and job satisfaction is higher and the level of stress is lower. So, the experience of stress is positive to employees and organizations, because stress motivates, inspires to be better and helps to develop skills (Adaramola, 2012; Wardwell et al., 2015). Other research shows the experience of stress negatively affects the employee and the organization, because a high level of stress could lead to burnouts and lower job performance (Collings & Murray, 1996; Wright et al., 2007). So, stress can be useful when its level is not too high. When the stress level of an employee is too high, several symptoms show that an employee experiences stress negatively (Draskovic, Stanic, Uram-Benka & Fabri, 2014).

Stress is among other things caused by the employee’s environment (e.g. private and work environment) (Chenhall & Langfield-Smith, 1998; Byrne & Pierce, 2007; De Loo et al., 2011; Ebstrup et al., 2011; Lazarus, 1991). In the 1970s Richard S. Lazarus came with the transactional stress theory (a new focus on stress) that stress results from the ongoing relationship between the environment of a person and the person himself (Vollrath, 2001). Therefore, the match between the role a controller performs and the personal traits he possesses also cause stress. This relation will be explained in the following section.

2.5. The effect of the match/mismatch on stress

Cluskey Jr. and Vaux (1997) found out vocational misfit influences stress, in which vocational misfit was defined as a mismatch between an employee’s personality and the occupational requirements of the job. Byrne & Pierce (2007) and De Loo et al. (2011) state the level of stress is influenced by their environment. Chronic stress is also influenced by personality (Ebstrup et al., 2011). Thereby, Hopper (1980), Kobasa et al., (1982), Cooper & Payne (1991), Smith & Mackenzie (2006), and Conner-Smith & Flachsbart (2007) suggested that a relationship between personal traits and the level of stress exists and a relationship between the role a controller performs and the level of stress exists.

So, earlier research shows that a relation exists between the role a controller performs and the personal traits he possesses, as well as a relation between personal traits a controller possesses and the level of stress he experiences, as well as a relation between the role a controller performs and the level of stress he experiences. Based on this earlier research on personality, occupational requirements and the environment as influencing variables on stress, another closely-related variable influencing stress, has been taken into account in this current research, namely the match between a controller’s role, and his personal traits consistent with this role.

This leads to the expectation that when a match exists the level of stress is lower than when a mismatch exists, suggesting the following hypotheses:

(13)

12 Hypothesis 3: The higher the match between a bean counter with the personal trait(s) consistent with

his role, the lower the level of stress this controller experiences.

Hypothesis 3a: The higher the match between a bean counter with the personal traits

accurateness and formality, the lower the level of stress this controller experiences.

Hypothesis 3b: The higher the match between a bean counter with the personal traits

reservedness, modesty and withdrawal, the lower the level of stress this controller experiences.

Hypothesis 3c: The higher the match between a bean counter with the personal trait conservativeness, the lower the level of stress this controller experiences.

Hypothesis 4: The higher match between a business partner with the personal trait(s) consistent with

his role, the lower the level of stress this controller experiences.

Hypothesis 4a: The higher match between a business partner with the personal traits

flexibility, proactiveness and creativeness, the lower the level of stress this controller experiences.

Hypothesis 4b: The higher match between a business partner with the personal traits

communicativeness, linguistic proficiency and teamworker, the lower the level of stress this controller experiences.

To conclude Chapter 2, Table 3 shows insight into which theory/which most import literature will be used to which hypothesis.

Table 3

Link hypotheses to theory

Hypothesis Theory

1 and 2 (relation between roles and personal traits) Byrne & Pierce (2007) and Granlund & Lukka (1998)

3 and 4 (relation between match and stress) Transactional stress theory

2.6. Conceptualmodel

To visualize the expectations of this research, the four hypotheses are formulated in the conceptual framework below (Figure 1). The first two hypotheses show the positive relationship between the role a controller performs and the personal traits a controller possesses. With this relationship the match between the role a controller performs and the personal traits consistent with that certain role can be made, resulting in the third and fourth hypotheses. The third hypothesis shows the expected negative

(14)

13 relation between the match of a controller performing his role as a bean counter and the personal traits consistent with this role and the level of stress this controller experiences. The fourth hypothesis shows the expected negative relation between the match of a controller performing as a business partner and the personal traits consistent with this role and the level of stress this controller experiences. In other words, this conceptual framework shows that a match from a bean counter perspective and a match from a business partner perspective is negatively associated with the level stress. Figure 1: Conceptual framework H1 (+) H2(+) H3 (-) H4 (-) 3. RESEARCH DESIGN

In this chapter the research method will be described. Firstly, the research method will be explained, this includes the description of the sample and variables. Secondly, the way the variables were measured are described and lastly the empirical test of the hypotheses is mapped.

The match between a

beancounter and the personal trait(s)consistent with his role

The match between a business partner and the personal trait(s) consistent with his role

Level of Stress The controller operates more as

bean counter than a business partner

The controller operates more as a business partner than a bean counter

The controller possesses mostly the personal traits of accurateness, withdrawal, reservedness, formality, modesty, and conservativeness

The controller possesses mostly the personal traits of linguistic

proficiency, proactiveness,

creativeness, flexibility, teamworker, and communicative.

(15)

14 3.1. Research Design

To address the research question and test the predicted hypotheses, a deductive quantitative research method is applied. The research is conducted with online questionnaires. The collected data is processed and is statistically tested to accept or reject the hypotheses. The unit of analysis is the controller of company, so an individual.

3.2. Sample

The sample is drawn from Dutch controllers who fulfil roles that tend to be the roles of bean counters and business partners. With an e-mail to befriended controllers, a post via LinkedIn and personal messages to controllers, Dutch controllers were approached to fill in the online questionnaire. Between December 10 and December 20 the approach to Dutch controller took place. It is difficult to estimate the total number of Dutch controllers reached by this approach, because it is unknown how many of the 590 million LinkedIn member world-wide (2018) work as a Dutch controller. Thereby, the total number of Dutch controllers in general are unknown.

3.3. Measures

According to the conceptual model in Figure 1, section 2.6, this research tests four hypotheses. For H1 and H2 the dependent variable in this research is the personal traits of a controller, and the independent variable is the role a controller performs. These variables are measured by indicators with different validated scales. For H3 and H4 the dependent variable in this research is the level of stress, because the mismatch from bean counter perspective and the mismatch from business partner perspective could affect the level of stress. The personal traits (e.g. accurateness, withdrawal, reservedness, formality, modesty, conservativeness, linguistic proficiency, proactiveness, creativeness, flexibility, teamworker, and communicative) of bean counters and business partners are the independent variables as well as the controller’s roles: bean counter and business partner. Earlier studies mention that differences in the controller’s personal traits are needed to perform the role of a controller as a bean counter or as a business partner. These variables are measured by indicators with different validated scales. Thereby, stress is caused by the controller’s background (e.g. education, experiences, age, gender etc.) and the controller’s work environment (e.g. firm size, sector etc.) (Chenhall & Langfield-Smith, 1998; Byrne & Pierce, 2007; De Loo et al., 2011; Ebstrup et al., 2011). These examples are the control variables in this research. The independent, dependent and control variables will be described below.

(16)

15 3.3.1. Independent and dependent variables

In this section the measures of the independent variable will be explained. First the measure of the role of a controller and the measure of the personal traits of a controller will be described. These are necessary to measure the independent variable match between a controller’s role and his personal traits (bean counter and business partner).

The role of a controller

In order to answer the research question, the role of the controller needs to be measured. In this research, the validated questionnaire of Hartmann and Maas (2011) will be used, which measures the importance of tasks to a controller. The respondents were asked (134 business unit controllers) with what frequency they perform a specific task. The tasks used, can be seen in Appendix 1, question 10. The respondent uses a Likert scale from 1 (never) to 5 (very often) indicating how frequent he performs a certain task. The tasks that indicate whether a respondent performs a bean counter role, are 1, 3, 5, 7 and 9 of question 10 Appendix 1. The tasks that indicate whether a respondent performs a business partner role, are 2, 4, 6, 8, 10 and 11 of question 10 Appendix 1.

Next, the mean of the questions concerning tasks 1, 3, 5, 7 and 9 (a bean counter role) and the mean of the questions concerning tasks 2, 4, 6, 8, 10 and 11 (a business partner role) are calculated, indicating to what extent the respondent performs one of the two roles. This mean is scored on a scale of 1 (Never) to 5 (Very often) which is linked to a certain percentage, as can be seen in Figure 2. Figure 2 shows at the x-axis the mean of the questions concerning tasks of both roles (Appendix 1, question 10). The y-axis shows the percentages which are linked to the mean and shows the extent to which a controller performs a certain role. For example, when the mean of all the tasks of a bean counter role is equal to or higher than 3, this indicates the respondent performs 50% or more the role of a bean counter. If the respondent scored higher on the tasks indicating a bean counter role compared to the tasks indicating a business partner role, this respondent will qualify as a bean counter. The respondent will qualify as a business partner when the respondent shows a higher score on the tasks indicating a business partner role compared to the tasks indicating a bean counter role. When the score on tasks indicating a bean counter equals the score on tasks indicating a business partner, a hybrid role exists. When a respondent performs a hybrid role, this respondent will be removed from the dataset.

(17)

16

Figure 2:

Measuring the extent to what the controller performs the role as a bean counter or as a business partner.

Personal traits of bean counters

To measure the personal traits of bean counters and business partners, indicators from earlier research are used. These indicators consist of validated questions, which are constructs of the expected personal traits, as described next.

The validated instrument of John and Srivastava (1981), comprising extraversion, conscientiousness and openness as three of the five factors of the “Big Five” personal traits, will be used to indicate whether the respondent possesses the expected personal traits of bean counters (e.g. accurate, withdrawal, quiet/reserved, formal, modest, conservative). The John and Srivastava’s (1981) instrument measures the extraversion, conscientiousness and openness with a number of characteristics, which are shown in Appendix 1, question 12. These characteristics can be linked with the expected personal traits of a bean counter, as shown in Appendix 2. These characteristics are measured by a Likert Scale between 1 (Strongly disagree) and 5 (Strongly agree). The characteristics indicating conscientiousness are used as a construct for the personal traits accurateness and formality, since conscientiousness is equal to accurateness and precise, and being organized implies a formal person, one who is precise and follows the rules, as can be seen in Appendix 1 (question 12, characteristic 2, 3, 5, 8, 11, 14, 16, 18, 20). The characteristics indicating extraversion are identical to the following expected personal traits of bean counters: reservedness, withdrawn and modesty (shy), as can be seen in Appendix 1 (question 12, characteristic 1, 4, 7, 10, 13, 15, 17, 19). In essence, introversion instead of extraversion will be used as in an indication for the personal traits reservedness, withdrawal and modesty. That is why the characteristics about extraversion will be used reversely (reversed characteristics are indicated with an (R) in Appendix 1 question 12), which means that the score will be used reversely. For example, when a respondent scored a 5 on a reversed characteristic, a score of 1 will be used to measure the expected personal traits of a bean counter (e.g. reservedness, withdrawal and modesty). The characteristics about openness will be used as an

(18)

17 indication for the expected personal trait conservativeness, as can be seen in Appendix 1 (question 12, characteristic 2, 6, 9, 12). Openness is contrary to conservativeness, which implies not being curious and sticking to the past. Therefore, similar to the characteristics about extraversion, the characteristics indicating openness will be used reversely and are also indicated with an (R) in Appendix 1 question 12.

The mean of all the characteristics (after calculated the reversed characteristics) in question 12 Appendix 1 per respondent will be measured and linked to a percentage to measure to what extent the respondent possess the personal traits of a bean counter. Figure 3 shows at the x-axis the mean of the questions concerning characteristics of the personal traits of a bean counter (Appendix 1, question 12). The y-axis shows the percentages which are linked to the mean and shows the extent to what a controller possesses the expected personal traits of bean counter. For example, when the mean of all the characteristics of the personal traits of a bean counter is equal to or higher than 3, this indicates the respondent possesses 50% or more of the characteristics of a bean counter (personal traits of a bean counter). When the mean is equal to 2.8 this means that the respondent possesses 45% of the characteristics of a bean counter. To calculate the match between the respondent’s role of a bean counter and his personal traits, this percentage will be matched with the percentage indicating the role of the respondent, which was defined previously in this section.

Figure 3:

Measuring the extent to what the controller possess the expected personal traits of a bean counter.

Personal traits of business partner

Also, the expected personal traits of business partners (linguistic proficiency, pro-activeness, creativeness, flexibility, teamworker, and communicative) will be measured. The twelve aspects of cognitive flexibility of Martin and Rubin (1995) are used as constructs of the expected personal traits of business partners, because these twelve aspects have similarities with the expected personal traits of

(19)

18 a business partner. Aspects 1, 10 and 11 can be linked to the personal traits communicative, linguistic proficiency and teamworker, as can be seen Appendix 3. The personal traits communicative and linguistic proficiency can be linked to ‘communicate’ (Aspect 1), ‘difficulty using my knowledge’ (Aspect 10) and ‘listen’ (Aspect 11), since both personal traits imply receiving and processing information, which can be compared to listening, and giving a response, which may be difficult in some situations. The personal trait teamworker can be linked to ‘willing to listen’ and ‘consider’ (both Aspect 11), since listening and considering interests are important for successful teamwork.

Aspects 2, 3, 4, 5, 6, 7, 8 and 12 can be linked to the personal traits creativeness, proactiveness and flexibility. Aspect 2, 3, 8 and 12 can be linked to the personal trait proactiveness, because these aspects imply ‘behaviour/decisions’ that react immediately on ‘new and different situations’. The aspects 2, 4, 5, 6, 7 and 9 can be linked to the personal trait flexibility, because these aspects imply adjustments in behaviour for example to new, unusual, different and/or difficult situations/problems. The aspects 5, 6, 9 and 12 can be linked to the personal trait creativeness, because the need of creative solutions in work, decisions or behaviour is mentioned. Appendix 1, question 11 shows the twelve aspects of which the respondents need to indicate to what extent they agree with the statements. These twelve aspects are measured by a Likert Scale between 1 (Strongly disagree) and 6 (Strongly agree). As well as question 12 in Appendix 1 about the expected personal traits of a bean counter, some of the twelve aspects will be used reversely and are indicated with an (R) in Appendix 1 question 11 (aspects 2, 3, 5 and 10) and Appendix 3. This means that, for example, the reversed questions which were rated 2 (Disagree) by the respondent will, in this research, be adopted as 5 (Agree).

The mean of the twelve aspects (after calculated the reversed aspects) in question 11 Appendix 1 per respondent will be measured and linked to a percentage to measure to what extent the respondent possess the personal traits of a business partner. Figure 4 shows at the x-axis the mean of the questions concerning the aspects of the personal traits of a business partner (Appendix 1, question 11). The y-axis shows the percentages which are linked to the mean and shows the extent to what a controller possesses the expected personal traits of business partner. For example, when the mean of all the aspects of the personal traits of a business partner is equal to or higher than 3.5, this indicates the respondent possesses 50% or more of the aspects of a business partner (personal traits of a business partner). This percentage will be taken into account when calculating the match between the role a controller performs and the personal traits he possesses. When one of the percentages is below 50%, no match will exist. How the match will be measured is explained below.

(20)

19

Figure 4:

Measuring the extent to what the controller possess the expected personal traits of a business partner

As a final step in determining the extent to which a respondent possesses a certain personal trait, the percentages which are linked to the personal traits of a bean counter and the percentages which are linked to the personal traits of a business partner will be compared to each other. This comparison indicates to what extent a respondent possesses the personal traits of a bean counter or a business partner. The highest percentage determines whether a respondent possesses the personal traits of a bean counter or a business partner. When the percentages are equal to each other for one respondent, these respondent will be removed from the database.

Match between the role of a controller and his personal traits

The independent variable in this research is the match between the role of a controller and his personal traits which consists of two parts, namely personal traits of bean counters and business partners and the role a controller performs: a bean counter role or a business partner role. Two measures are used to calculate whether a respondent qualifies as a match. First, the match is calculated by taking into account all expected personal traits at once (Measure 1). Secondly, the match is calculated by taking into account the expected personal traits separately (Measure 2). Using the first measure shows which role has the highest influence on the level of stress. Using the additional measure makes it possible to identify which group of personal traits per role has the highest influence on the level of stress. The groups will be identified based on the constructs used to identify the personal traits, as described previously in section 3.3.1.

Measure 1

To measure the extent to which a combination of personal traits and a role of a controller is a match or a mismatch, the percentages which are linked to the means of questions 10, 11 and 12 in Appendix 1 (mentioned above in Figure 2, 3 and 4) will be compared as follows. The extent to which

(21)

20 a match between the expected personal traits of a bean counter and the bean counter role exists is measured by multiplying the extent to which a controller performs a bean counter role and the extent to which a controller possesses (partly) the personal traits of a bean counter. For example, when the respondent rated himself on the characteristics in question 12, leading to a mean of 5 (Strongly agree) which is linked to 100%, and also rated himself on the tasks (1, 3, 5, 7 and 9) in question 10, leading to a mean of 4 (Often), which is linked to 75%, these percentages are multiplied with each other (100%*75%), leading to a percentage of 75%. To qualify as a match from a bean counter perspective, a percentage of 50% or more is required. Since 75% is higher than 50%, this respondent qualifies as a match on the bean counter perspective.

The extent to which a match between the expected personal traits of a business partner and the business partner role exists, is measured by multiplying the extent to which a controller performs a business partner role and the extent to which a controller possesses (partly) the personal traits of a business partner. For example, when the respondent rated himself on the aspects in question 11, leading to a mean of 5 (Agree) which is linked to 80%, and also rated himself on the tasks (2, 4, 6, 8, 10 and 11) in question 10, leading to a mean of 4 (Often) which is linked to 75%, these means are equated to percentages and multiplied with each other (80%*75*), leading in this case to a percentage of 60%. To qualify as a match from a business partner perspective, a percentage of 50% or more is required. Since 60% is higher than 50%, this respondent qualifies as a match on the business partner perspective.

Measure 2

To measure the extent to which a group of personal traits and a role of a controller is a match or a mismatch, the percentages which are linked to the means of questions 10, 11 and 12 in Appendix 1 (mentioned above in Figure 2, 3 and 4) will be compared, similar as they are compared in Measure 1. The difference between Measure 1 and Measure 2 is the use of groups of personal traits. The expected personal traits of a bean counter will be separated into three groups. Group 1 contains accurateness and formality, group 2 contains reservedness, modesty and withdrawn and group 3 conservativeness. For example, once a respondent qualifies as a bean counter and has the highest score on accurateness and formality compared to the other personal traits, he belongs to group 1.

The expected personal traits of a business partner will be separated into two groups. Group 1 contains flexibility, proactiveness and creativeness; group 2 contains communicative, linguistic proficiency and teamworker. Once a respondent qualifies as a business partner and has the highest score on communicative, linguistic proficiency and teamworker compared to the other personal traits,

(22)

21 he belongs to group 2. For example, when the respondent rated himself on the aspects 1, 10 and 11 (group 2) in question 11, leading to a mean of 6 (Strongly agree) which is linked to 100%, and also rated himself on the tasks (2, 4, 6, 8, 10 and 11) in question 10, leading to a mean of 4 (Often) which is linked to 75%, these means are equated to percentages and multiplied with each other (100%*75%), leading in this case to a percentage of 75%. To qualify as a match from a business partner perspective, a percentage of 50% or more is required. Since 75% is higher than 50%, this respondent qualifies as a match on the business partner perspective.

It can be concluded that respondents which do not qualify as a match using Measure 1 do qualify as a match using Measure 2, leading to a higher number of matches when using Measure 2.

The level of stress

The dependent variable the level of stress is measured by a validated survey, namely the twenty job-related stressor events of Vagg and Spielberger (1998). Vagg and Spielberger (1998) collected data in 1994 and received answers from 1.791 university and corporate employees, who were separated in analyses for male and female management-professionals and clerical-maintenance groups, in their job stress survey. Appendix 1, question 13 shows the stress events submitted to the respondents indicating to what extent they experience stress based on the stressor events resulted from the research of Vagg and Spielberger (1998). By using a Likert scale between 1 (never) and 7 (always) the respondents need to indicate to what extent they experience stress in different circumstances. Every respondent is allowed to define stress according to his own principles. This measure does not distinguish between emotional or physical responses on the stress events. Since it is frequently used in other research, by using it here as well, this study’s results are comparable with other studies.

3.3.2. Control variables

Tenure, former function, age, gender, education, number of employees, and sector

Other variables than personal traits and the role a controller performs can influence the level of stress a controller experiences. To eliminate this variable bias, different control variables were included in the empirical tests. Chendall and Langfield-Smith (1998), Byrne and Pierce (2007), De Loo et al. (2011), and Ebstrup et al. (2011) found that the role a controller performs and the level of stress the controller experiences is influenced by his personality, his personal traits, a controller’s working experience and the environment of the organization and the controller itself. With the independent variables personal traits of bean counters and personal traits of business partners the personality and personal traits of the controllers will be measured. With the control variables age, tenure, former function and education the controller’s working experience will be measured. With the control

(23)

22 variables ‘number of employees of the organization’ and the ‘sector’ the organization operates, the environment of the organization will be measured. These control variables will be measured by the questions 1 to 9, which can be found in Appendix 1. A detailed overview of these variables can be found in the next section.

Table 4 shows an overview of all the variables, labels which are linked to the variables, the explanations of the variables and the abbreviations. Appendix 1 shows all the questions used in the questionnaire to measure all variables.

Table 4

Variable descriptions

Variable Label Explanation Measurement

(questions refer to the survey in Appendix 1)

Dependent variable and independent variables

Level of stress STRESS The extent into controllers experience stress. µ of question 13

Bean counter role BC The extent into a controller fulfills a bean counter role according

to the tasks he perform.

µ (linked to %) of tasks 1, 3, 5, 7 & 9 of question 10

Business partner role BP The extent into a controller fulfills a business partner role

according to the tasks he perform.

µ (linked to %) of tasks 2, 4, 6, 8, 10 & 11 of question 10

Personal traits of a bean counter

PTBC The extent into a controller possess the expected personal traits

of the controller’s role as a bean counter.

µ (linked to %) of question 12

Group 1 personal traits of a bean counter

ACC The extent into a controller possess the personal traits

accurateness and formality which are expected personal traits of the controller’s role as a bean counter

µ (linked to %) of characteristics 2, 3, 5, 8, 11, 14, 16, 18 & 20 of question 12 Group 2 personal traits of a bean counter

RES The extent into a controller possess the personal traits

reservedness, withdrawal and modesty which are expected personal traits of the controller’s role as a bean counter

µ (linked to %) of characteristics 1, 4, 7, 10, 13, 15, 17 & 19 of question 12 Group 3 personal traits of a bean counter

CONS The extent into a controller possess the personal trait

conservativeness which is an expected personal trait of the controller’s role as a bean counter

µ (linked to %) of characteristics 2, 6, 9 & 12 of question 12 Personal traits of a

business partner

PTBP The extent into a controller possess the expected personal traits

of the controller’s role as a business partner.

µ (linked to %) of question 11

(24)

23

Group 1 personal traits of a business partner

FLEX The extent into a controller possess the personal traits flexibility,

proactiveness and creativeness which are expected personal traits of the controller’s role as a business partner

µ (linked to %) of aspects 2, 3, 4, 7, 8, 9 & 12 of question 11 Group 2 personal traits of a business partner

COMM The extent into a controller possess the personal traits

communicative, linguistic proficiency and teamworker which are expected personal traits of the controller’s role as a business partner µ (linked to %) of aspects 1, 5, 6, 10 & 11 of question 11 Match between personal traits of a bean counter and the bean counter role

MATCH PTBC

The extent into a the match between a controller who performs a bean counter role and possesses the personal traits consistent with the bean counter role: the match from a bean counter perspective. Multiplied % of BC tasks and % of PTBC characteristics Match between group 1 personal traits of bean

counters and the bean counter role

MATCH ACC

The extent into a the match between a controller who performs a bean counter role and possesses the personal traits in group 1 (e,g. accurateness and formality) consistent with the bean counter role: the match from a bean counter perspective.

Multiplied % of BC tasks and % of ACC characteristics

Match between group 2 personal traits of bean

MATCH RES

The extent into a the match between a controller who performs a bean counter role and possesses the personal traits in group 2 (e.g. reservedness, modesty and withdrawal) consistent with the bean counter role: the match from a bean counter perspective.

Multiplied % of BC tasks and % of RES characteristics

Match between group 3 personal traits of bean

MATCH CONS

The extent into a the match between a controller who performs a bean counter role and possesses the personal trait in group 3 (e.g. conservativeness) consistent with the bean counter role: the match from a bean counter perspective.

Multiplied % of BC tasks and % of CONS characteristics

Match between personal traits of a business partner and the business partner role

MATCH PTBP

The extent into a the match between a controller who performs a business partner role and possesses the personal traits consistent with the business partner role: the match from a bean counter perspective. Multiplied % of BP tasks and % of PTBP aspects Match between group 1 personal traits of a business partner and the business partner role

MATCH FLEX

The extent into a the match between a controller who performs a business partner role and possesses the personal traits in group 1 (e.g. flexibility, proactiveness and creativeness) consistent with the business partner role: the match from a bean counter perspective.

Multiplied % of BP tasks and % of FLEX aspects

(25)

24

Match between group 2 personal traits of a business partner and the business partner role

MATCH COMM

The extent into a the match between a controller who performs a business partner role and possesses the personal traits in group 2 (e.g. communicative, linguistic proficiency and teamworker) consistent with the business partner role: the match from a bean counter perspective.

Multiplied % of BP tasks and % of COMM aspects

Control variables

Tenure TEN The time a controller fulfill his function as controller (in years) µ of question 3

Former function FFUN The former function of a controller who did not starts his career

with working in the function as a controller.

Frequencies of question 4

Age AGE The age of controllers who fulfill the questionnaire in whole

years.

µ of question 5

Gender GEN The gender of controllers who fulfill the questionnaire. 1 = male,

2 = female

Educational background

EDU The educational background of controllers who fulfill the

questionnaire

Number of employees

EMPL Estimate of number of employees working in the company. The

company where the controller who fulfils the questionnaire works.

Sector of the organization

SEC The sector the organization of the controller operates in. The

survey gives eleven options of sectors inclusive the option “Other”.

3.4. Method

The hypotheses will be tested with linear regression analyses, because regression analyses test the relation between variables (Field, 2013). This research estimates the relation between the variables about the role a controller performs and the personal traits a controller possesses. This research also estimates the relation between the variables about the match between the role a controller performs and the personal traits he possesses and the level of stress this controller experiences. This research use a simple linear regression because the independent variables will be tested one at a time. The formulas of the two simple linear regressions for this research are described below:

PTBC, ACC, RES, CONS, PTBP, FLEX or COMM = β0 + β1 * BC or BP + εi

(26)

25 To test the level of stress on the basis of the independent variable and the control variables the following formulas are used:

PTBC, ACC, RES, CONS, PTBP, FLEX or COMM = β0 + β1 * BC or BP + β2 * TENi + β3 * FFUNi + β4 * AGEi + β5 * GENi + β6 * EDUi + β7 * EMPLi + β8 * SECi + εi

STRESSi = β0 + β1 * MATCHPTBCi or MATCHPTBPi + β2 * TENi + β3 * FFUNi + β4 * AGEi + β5 * GENi + β6 * EDUi + β7 * EMPLi + β8 * SECi + εi

β0 = Constant coefficient

βi = Slope, directional coefficient

εi = Error variable

Also a robustness test for the relation between the match and the level of stress is made. The match will be measured by whether a match exists instead of to what extent a match exists. This means that a match is equal to one and a mismatch is equal to zero.

3.5. Validity and reliability

In order to answer the research question it is important the results are valid and reliable. To make sure this research is valid, certain measures are taken. The questionnaire consists of validated questionnaires and scales from previous research of the dependent and independent variables. Before the questionnaire was published, colleague students, one professor and one teacher filled in the questionnaire and gave a critical review. The reviews are used to adjust the questionnaire which is distributed to the sample of controllers. Thereby, controllers from different organizations in different sectors are contacted to fill in the questionnaire to diminish the selection bias. When using a questionnaire, a bias always exists, which could cause in spurious correlations. In this research a distorted view of the respondent’s role, personal traits and level of stress could increase this bias.

To increase the reliability of this research, the number of respondents is in line with the statistical requirement, as explained in section 4.1.1. In addition, cross- and back-to-back translation of the questionnaire was applied to ensure the translation did not reduce the validity and the reliability of the instruments. The respondents could choose between English or Dutch as a language to answer the questions in the questionnaire.

Table 5 shows the Cronbach’s alpha’s of the dependent and independent variables. The minimum requirement of the Cronbach’s alpha’s is equal to or higher than 0.7. Most Cronbach’s

(27)

26 alpha’s are 0.7 or higher, indicating the questions about these variables are a reliable measure. When Cronbach’s alpha is lower than 0.7, questions with the lowest inter-item correlation will be removed until the threshold of 0.7 will be reached. Furthermore, the Cronbach’s alpha of the match between the controller’s role and his personal traits is measured by the mean of the Cronbach’s alpha’s of the role of a bean counter or a business partner, and the personal traits of bean counter or business partner.

Table 5 Cronbach’s alpha’s Variable α in this research Question in survey (Appendix 1) α after removing question(s) with the

lowest inter-item correlation Removed questions STRESS .89 13 (1 till 20) .89 BC .43 10 (1, 3, 5, 7, 9) .70 1, 5 BP .69 10 (2, 4, 6, 8, 10, 11) .72 11 PTBC .64 12 (1 till 12) .70 8, 16, 11 ACC .70 12 (3, 5, 8, 11, 14, 16, 18, 20) .70 RES .83 12 (1, 4, 7, 10, 13, 15, 17, 19) .83 CONS .39 12 (2, 6, 9, 12) .70 2, 9 PTBP .75 11 (1 till 20) .75 FLEX .61 11 (2, 3, 4, 5, 6, 7, 8, 9, 12) .70 3, 7 COMM .70 11 (1, 10, 11) .70

Matches between roles and personal traits

MATCHPTBC .52 12 (1 till 12) * 10 (1, 3, 5, 7, 9) .70 MATCHACC .57 12 (3, 5, 8, 11, 14, 16, 18, 20) * 10 (1, 3, 5, 7, 9) .70 MATCHRES .63 12 (1, 4, 7, 10, 13, 15, 17, 19) * 10 (1, 3, 5, 7, 9) .77 MATCHCONS .41 12 (2, 6, 9, 12) * 10 (1, 3, 5, 7, 9) .70 MATCHPTBP .72 11 (1 till 20) * 10 (2, 4, 6, 8, 10, 11) .74 MATCHFLEX .65 11 (2, 3, 4, 5, 6, 7, 8, 9, 12) * 10 (2, 4, 6, 8, 10, 11) .71 MATCHCOMM .70 11 (1, 10, 11) * 10 (2, 4, 6, 8, 10, 11) .71

(28)

27 4. RESULTS

In this chapter the results of the survey are presented. First, the descriptive statistics and the internal consistency of the different factors are presented to describe the variables that have been tested in the survey. Secondly, this chapter describes if the hypotheses are supported by the collected data or not.

4.1. Descriptive statistics

This paragraph consists of a description about the number of respondents which are taken into account in this research. Next to that, a description is presented about the dependent variable, the independent variables and the control variables. For most variables the descriptive statistics are mapped (e.g. means, medians, maximums, maximums etc.) or else the frequencies of the variables are described.

4.1.1. Response rate

Because controllers operate in different branches it is difficult to estimate how many Dutch controllers exist, therefore the population of Dutch controller is unknown. Nevertheless, Cochrant (1997) states that when the size of the population is unclear, 20.000 needs to be used, because with a population of more than 20.000, the sample will increase asymptotically (very slow). This population is used by using the formula of Cochrant (1977) with a confidence level of 95% and a margin of error of 5% the sample needed to come to reliable results, is calculated on 377. On the basis of available data, Dutch controllers were approached between December 10 and December 20 to fill in the online questionnaire. Most controllers are approached via a post or a (personal) message on LinkedIn and mails to befriended Dutch controllers. Befriended controllers were asked to share the questionnaire via social media. Therefore, it is difficult to estimate how many Dutch controllers have been approached, so no response rate can be calculated. One reminder has been sent after five days to the controllers who have been approached, but did not respond yet. As a result, 184 controllers filled in the questionnaire. Twenty of these respondents did not completely finish the survey, ten have never worked as a controller, two do not work as a controller anymore for the last five years, four are performing a hybrid role and two possess the personal traits of both roles. Therefore these 38 respondents were deleted from the sample. So the final sample consisted of 146 respondents. This means, using the formula of Cochrant (1977), with a confidence level of 95%, a scatter of 50% and 146 respondents, the error margin is 8.09% instead of the most commonly used 5%.

The effect of the mismatch between personal traits and the role of management accountants on stress.

The effect of the mismatch between personal

traits and the role of management accountants

on stress.

TABLE OF CONTENTS