• No results found

The effect of customer reliance on the control tightness in the different PSF types

N/A
N/A
Protected

Academic year: 2021

Share "The effect of customer reliance on the control tightness in the different PSF types"

Copied!
80
0
0

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

Hele tekst

(1)

The effect of customer reliance on the control tightness in the

different PSF types

Name: Meilyn Huynh Student number: 11093773

Thesis supervisor: Prof. Dr. Marc Wouters Date: 25 June 2017

Word count: 19640

MSc Accountancy & Control, specialisation Accountancy and Control Faculty of Economics and Business, University of Amsterdam

(2)

Statement of Originality

This document is written by student Meilyn Huynh who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

ABSTRACT

This study aims to create an understanding of how management control systems are designed in professional service firms (PSF). More specifically, this study uses the concept of customer reliance to determine its effect on the explicit/implicit control tightness of the results/action/personnel/cultural controls. To investigate these effects, data have been collected through a survey. This study uses the data of 312 professionals, working in PSFs in various industries. Subsequently, the professionals are divided into either Reciprocal Service Design firms or Pooled Service Design firms. The two types of PSFs are drawn from the framework of Larsson and Bowen (1989). Results in this study indicate that customer reliance only has a positive and significant impact on the explicit results controls and the implicit personnel controls for PSFs. Breaking this down into two PSF types, Pooled Service Design firms tend to use explicit and implicit action controls more than Reciprocal Service Design firms in relation with the level of the customer reliance. Contrary to the beliefs that both PSF types use implicit personnel controls with equal intensity, findings in this study indicate that this is not true. In fact, Pooled Service Design firms use implicit personnel controls slightly more than Reciprocal Service Design firms. Furthermore, when splitting the dataset into the two types of PSFs, results of the analyses indicate that the customer reliance has a positive and significant effect on the explicit results controls and the explicit action controls for Reciprocal Service Design firms. For the Pooled Service Design firms, however, results indicate that customer reliance only has a positive and significant effect on the implicit personnel controls and implicit cultural controls. The contradictions between expectations and results can be mainly explained due to other factors that could have possibly affect the relations between customer reliance and the respective control tightness dimensions. Nevertheless, based on these findings, this study enables the possibilities for future research.

Keywords: Customer Reliance; Management Control System (MCS); Explicit Control Tightness; Implicit Control Tightness; Results Controls; Action Controls; Personnel Controls; Cultural Controls; Professional Service Firms

(4)

ACKNOWLEDGEMENT

I would first like to thank my thesis supervisor, Prof. Dr. Marc Wouters, for his guidance throughout the entire master thesis process. His remarks, comments, and suggestions were extremely useful as it steered me into the right direction to help me put this thesis together. The provoked discussions during the meetings were also very insightful and helpful. Furthermore, I would also like to thank Ms. Helena Kloosterman for making this research project possible. I would also like to acknowledge the respondents who were involved in answering the survey, and therefore, provided useful and valuable data for this research project.

Also, I would like to express my deepest gratitude towards my family for helping me find sufficient respondents for the survey. In addition, my close friends, who I would like to thank as well for their best supports, advices, and suggestions. Finally, I am especially grateful towards my mother and my younger brother for continuously encouraging, supporting, and motivating me throughout the process. Thank you very much.

Meilyn Huynh

(5)

TABLE OF CONTENT

1 Introduction ... 7

2 Literature Review ... 10

2.1 Management Control System ... 10

2.1.1 Dimensions in a Management Control System ... 10

2.1.2 Control Tightness in a Management Control System ... 12

2.2 Professionals and their Work ... 14

2.2.1 Professionals ... 14

2.2.2 Professional Service Firms ... 15

2.2.3 Professional Service Firm Types ... 16

2.3 Customer Reliance ... 18

3 Hypothesis Development ... 20

4 Research Method ... 26

4.1 Survey ... 26

4.1.1 Pre-tests for the Survey ... 26

4.2 Sample Criteria ... 27

4.3 Sample Selection ... 27

4.4 Constructs used from the Survey ... 29

4.4.1 Customer Reliance ... 29

4.4.2 Control Tightness Dimensions ... 29

4.4.3 Professional Service Firm Type ... 30

4.4.4 Control Variables ... 31

4.5 Validity of the Constructs ... 32

4.5.1 Principal Component Analysis ... 32

4.5.2 Reliability Analysis ... 37 4.6 Statistical Analysis ... 38 5 Results ... 39 5.1 Descriptive Statistics ... 39 5.2 Correlation Analysis ... 42 5.3 Regression Analysis ... 43 5.3.1 Hypothesis 1 ... 44 5.3.2 Hypothesis 2 ... 45 5.3.3 Hypothesis 3 ... 46 5.3.4 Hypothesis 4 ... 47

(6)

5.3.5 Summary Results Regression Analysis ... 48

5.4 Additional Analysis ... 49

6 Conclusion ... 51

6.1 Discussion ... 51

6.2 Limitations ... 54

6.3 Suggestions for Future Research ... 55

7 References ... 57

8 Appendices ... 61

Appendix A: Constructs’ Definition ... 61

Appendix B: Survey ... 62

Appendix C: PSF Type Distribution per occupation ... 74

Appendix D: Survey Constructs ... 75

Appendix E: Principal Component Analysis – All Constructs ... 78

(7)

1 Introduction

In a recent research, Stojkoski, Utkovski, and Kocarev (2016) analysed the influence of services on the countries’ productive structures. They find that Economic Complexity Indices for services are higher than those for goods. This notion is also reflected in countries’ ranking, whereby countries with a developed service sector rank higher than countries focusing on goods. In other words, the researchers believe that, in both developed worlds and developing worlds, services can create a bigger long-term economic growth in contrary to economies that focus on goods.

Other researches have also indicated that, within countries, a large part of their economy consists of services while the manufacturing sector has continued to shrink over the years (Clark, 2016; Stiglitz, 2016). Additionally, this view seemed not to be only limited on a national level. Wirtz, Tuzovic, and Ehret (2015) state that a large part of the growth of the global service economy is explained by business services, which consists of a variety of knowledge-intensive and creative professional services. According to these researchers, the fast growth of business services is directly related to the shift from domestic production to the global outsourcing of services.

Given the growing importance of professional services at an economic perspective, this has also led to an increase of interest in the scientific knowledge of this subject (Ciarli, Meliciani, & Savona, 2012). Von Nordenflycht (2010) notes that due to the distinctive characteristics of professional service firms (PSF) compared to the other types of firms, distinctive theories are needed for the management to properly manage this type of firm.

Greenwood, Li, and Rajshree (2005) define PSFs as firms with highly educated (professional) workforce as their primary asset and whose output consist of intangible services contained with complex knowledge. Based on this definition, the researchers also define two aspects that influence the strategic and the organisational choices of these firms. First, there is an information asymmetry between the firm and its client, making the client dependent on the firm. Second, due to the high importance of the firm’s human asset, the human asset is dependent on its professional workforce.

Based on the study of Greenwood et al. (2005), the relation between the firm and its client seemed to be an important key aspect in a PSF. Yet, there is still a lack of research on how PSFs design their management control systems (MCS) to manage this relationship (Homburg & Stebel, 2009). The researchers mention that the information asymmetry between a PSF and its client is also related to the double moral hazard risk as both parties can behave opportunistically due to the

(8)

interdependent and interactive characteristics of this relationship. Hence, they remark that this aspect needs to be taken into consideration when designing a MCS.

To analyse the influence of customers on the operations of PSFs, this research uses the control tightness as the determinant for designing a MCS. However, van der Stede (2001) mentions that control tightness does not have a uniform definition across the many studies. For example, Merchant and van der Stede (2012) define the control tightness as a benefit of a MCS, as they state that “the benefit of any MCS is derived from the increase in the likelihood that the organisation objectives will be achieved relative to what could be expected if the MCS were not in place” (p. 123). On the other hand, Anthony and Govindarajan (1998) believe that the control tightness is determined by looking at the budgeting process, such as how much emphasis is placed on meeting the budget or the amount of detailed revisions in budget reviews. In this research, the control tightness is defined as the degree of flexibility in the MCS.

Von Nordenflycht (2010) also mentions that prior studies have mainly been focusing on the ‘classic’ types of PSFs, such as accounting, law, and architecture firms. However, this raises the question whether the existing theories of these types of PSFs can also be applied to other types of PSFs. Consequently, he states that there is a need of empirical researches that goes beyond the ‘Classic PSFs’. Auzair and Langfield-Smith (2005) share the same opinion. They also agree that the knowledge of MCS in specific service industries are still limited. Similarly, they also argue that this limitation is caused by many prior studies focusing on single industries or single organisation (e.g., Gooneratne & Hoque, 2013; Pierce & Sweeney, 2005).

This study investigates the impact of customer reliance on the MCS design. More specifically, the control tightness will act as the determinant of a MCS design. Furthermore, this study will also address the other PSF types, other than the ‘classic PSFs’, to determine whether existing studies of this matter can also be applied to other PSFs. This leads to the following research question:

To what extent does the customer reliance affect the control tightness in a MCS design of the different PSF types?

To answer this question, this survey-based research looks at the control tightness in a MCS in two ways, the explicit and implicit control tightness. According to the framework of Merchant and van der Stede (2012), a MCS consists of three types of controls, namely results controls, action controls, and social controls. The latter can be further split into two different types of controls, i.e. personnel and cultural controls. This study identifies two types of PSFs, drawing upon the framework of Larsson and Bowen (1989). This framework categorises the different types of PSFs

(9)

by looking at the customers as their main variable. More specifically, the researchers looked at the customers at two perspectives, i.e. customers’ diversity of demand and customer disposition, to define the different PSF types.

This study starts with a literature overview of the different constructs, which will be presented in Chapter 2. Based on the literature overview, each construct will have their own established definition. Next, the development of the hypotheses is presented in Chapter 3. Chapter 3 covers the underlying relations between the different constructs, namely the customer reliance, and the explicit/implicit control tightness of the results/action/personnel/cultural controls for each PSF type and without. Chapter 4 presents the research method of this study. This chapter explains the creation of the final dataset and the constructs used in this study. Subsequently, the validity of the constructs will be analysed through a principal component analysis (PCA) and a reliability analysis. The results of both analyses will also be presented in Chapter 4. Then, in Chapter 5, the results, which this study is solely based on, will be presented. Specifically, this chapter covers the results of the independent samples t-test, correlation analysis, and the regression analyses. Finally, this study ends with the discussion of the findings, the limitations of this study, and possible future research, which will be covered in Chapter 6.

(10)

2 Literature Review

This section discusses the literature of the relevant topics, such as the MCS, the professionals and their work, and the customer reliance. The MCS will be further elaborated on the dimensions within a MCS and the control tightness of each control dimension. A short overview of the definition of each construct is presented in Appendix A.

2.1

Management Control System

In the 60s, Anthony (1965) defined a MCS as a process by which managers ensure that resources are obtained and used effectively and efficiently to accomplish the organisation’s objectives. About half a century later, Merchant and van der Stede (2012) still use the same definition. They describe MCSs as systems that managers use to ensure that the actions and the behaviours of their employees are in line with the organisation’s objectives and strategies. Elaborating on this definition, they believe that when MCSs are designed in a proper way, it will result in positively influencing the employees’ behaviour, and so, increasing the probability that the organisation will meet its goals. In other words, the main function of a MCS is to influence employees’ behaviour in desirable ways and its benefit is that the MCS increases the probability, for organisations, to achieve their objectives. The following sections discusses the dimensions within a MCS and its control tightness.

2.1.1 Dimensions in a Management Control System

The controls used in MCSs have been defined in many ways (Langfield-Smith, 1997). For instance, Ouchi (1979) sees clan, bureaucratic, and market controls as the three main dimensions in a MCS. On the other hand, Simons (1994) identifies four controls, which the author calls as ‘Levers of Control’. The Levers of Control consist of the belief system, boundary system, diagnostic system, and interactive system. Langfield-Smith and Smith (2003) even view the controls as patterns as they call them the trust pattern, bureaucratic pattern, and market pattern. Merchant and van der Stede (2012) also identifies three controls, which are social controls, action controls, and results controls. Social controls in their definition can be split into two more controls, which are personnel controls and cultural controls, making a total of four types of controls. Despite having different names, each control mentioned above share a similar definition as their respective control mentioned in the other study. This study applies the framework of Merchant and van der Stede (2012), which means that, from onwards, social controls, action controls, and results controls will be used for further elaborations.

First, Merchant and van der Stede (2012) define results controls as a control that involves rewarding employees based on their performance. This reward does not necessarily mean that only

(11)

monetary compensation or incentives are involved; other rewards, such as promotions and job security are also seen as a form of reward. There are four steps involved when implementing results controls: (a) defining the dimension(s) on which results are desired; (b) measuring performance in the chosen dimensions; (c) setting performance targets for employees to attain for each of the measures; and (d) providing rewards for target attainment to encourage the behaviour that will lead to the desired results. However, the effectiveness of using results controls can be tainted if the following conditions are not met: (a) organisation can determine what results are desired in the areas being controlled; (b) the employees whose behaviours are being controlled have significant influence on the results for which they are being held accountable; and (c) organisations can measure the results effectively.

Second, actions controls are used to ensure that employees perform or do not perform certain actions that are beneficial or harmful to the organisation. Specifically, action controls do not only direct how employees should not behave, but also direct employees how they should behave. There are four underlying controls for the actions controls: (a) behaviour constraints; (b) pre-action reviews; (c) action accountability; and (d) redundancy. The first type can be considered as a ‘negative’ control, as such, this control makes it very difficult for employees to act the way they should not act. The second type of action control involves the scrutiny of the action plans of the employees being controlled. The third type of control is to make sure that employees are held accountable for their actions. The fourth one should prevent employees from assigning too many employees or equipment to a task. Like the results controls, action controls also have conditions to be met before it works effectively, which are (a) organisations can determine what actions are (un)desirable and (b) organisations can ensure that the (un)desirable actions (do not) occur.

Third, personnel controls are designed to increase the probability of employees performing a task in a satisfactory way by themselves. This type of control serves three purposes: (a) ensuring that employees understand what the organisation wants; (b) ensuring that employees can do a good job; and (c) increasing the likelihood that employees will engage in self-monitoring. There are three steps involved for implementing personnel controls: (a) selection and placement; (b) training; and (c) job design and resourcing. In summary, finding the right people to do a certain task, training them, and giving them a good work environment will increase the likelihood that the tasks will be performed in a proper manner.

Finally, cultural controls function as controls to shape organisational behavioural norms and to encourage employees to monitor and influence each other’s behaviours. The latter is a form of group pressure on individuals who deviate from the group’s norms and values. To determine an

(12)

organisational culture, one can look at the code of conduct, group rewards, intra-organisational transfers, physical and social arrangement, and tone at the top.

The social controls, which consist of personnel and cultural controls, are hard to determine whether the controls are effective or not. This is due to the contingent characteristics every individual and/or group has.

2.1.2 Control Tightness in a Management Control System

Over the years, control tightness has not been consistently defined in the literature (van der Stede, 2001). Many researchers define only a characteristic or component of the control tightness.

Merchant and van der Stede (2012) define the control tightness as a benefit of a MCS. According to them, they believe that “the benefit of any MCS is derived from the increase in the likelihood that the organisation objectives will be achieved relative to what could be expected if the MCS were not in place” (p. 123). More specifically, they argue that a tighter MCSs should provide more assurance that employees act in the organisation’s best interest.

Anthony and Govindarajan (1998) see the properties of a budgeting process as a determinant for tight controls, such as the amount of emphasis is put on meeting the budget or the amount of details in budget reviews. Merchant (1985), on the contrary, argues that tight controls are not necessarily associated with detailed budget reviews and the frequent reviews of performance. Instead, he views tight controls as an obstruction that limits the degrees of freedom of the individual being controlled.

Kober, Ng and Paul (2007) base their judgement of the control tightness on the extent of monitoring during the process. This means that they focus on the extent of alignment to the rules, policies, plans, and activities being monitored. A similar definition is also seen in the research of Auzair and Langfield-Smith (2005) as they define tight controls as frequent monitoring of the employees’ actions and targets in a precisely and timely manner. In opposition, loose controls are viewed as less frequent monitoring of the employee’s actions and targets while also allowing some discretion in achieving these targets.

The tight and loose dimension can also be based on the strategic context. Bedford, Malmi, and Sandelin (2016) find that firms prioritising efficiency and conformance will have a more effective MCS when using tight controls. Loose controls, on the other hand, are more effective for firms focusing on innovation and flexibility.

In this study, the control tightness is defined as the degree of flexibility in a MCS. This aspect is split into two dimensions, namely the explicit and implicit control tightness. The former

(13)

is defined as the degree of extending the scope of the MCS. In other words, a tight system is defined as a system that has a lot of controls in terms of amount and scope. The latter, on the other hand, is defined as the degree of acceptance of deviations from the MCS. With this definition in place, for example, a tight system implies that deviations from standards established in a MCS is not allowed.

Summarising, control tightness, in this study, is defined as the extent of amount and scope of controls in a MCS (explicit controls) and the degree of acceptance of deviations from a MCS (implicit controls).

Results Control Tightness

In this study, explicit results control tightness indicates the extent of use of goals, targets, and performance measures as part of a MCS (Jaworski, Stathakopoulos, & Krishnan, 1993). Implicit results control is defined as the degree to which deviation from goals, targets, and performance measures is tolerated and/or encouraged (Simons, 1987).

Action Controls Tightness

Action controls are only considered as tight when it is highly likely that employees act to the operation’s success of their firm and will not engage in actions that will harm the firm (Merchant and van der Stede, 2012). As such, explicit action controls tightness is described as the extent of use of standardised processes, procedures, rules, and routines as part of a MCS. Implicit action controls tightness is described as the degree to which deviation from standardised processes, procedures, rules, and routines is tolerated and/or encouraged.

Personnel Control Tightness

This study describes explicit personnel control tightness as the extent of use of employee selection procedures as part of the MCS. Personnel controls are considered as tight when the employee selection procedure is extensive as the organisation seeks to actively select employees with certain knowledge, skills, and/or abilities. Implicit personnel control tightness is defined as the degree to which deviation from human resource standards is tolerated. Thus, personnel controls are considered as tight when deviations from the human resource standards are not tolerated.

Cultural Control Tightness

Explicit cultural control tightness in this study is described as the extent of use of employee socialisation procedures as part of a MCS. A tight system is considered when the use of socialisation procedures is extensive. In other words, firms actively seek to bring new employees into their organisation’s culture. Implicit cultural control tightness is described as the degree to which the employees’ norms, values and beliefs are tolerated to deviate from those of the organisation. It

(14)

depends on the degree to which employees’ norms, values, and beliefs are in line with the organisation’s norms, values, and beliefs to determine whether the system is tight or loose. More specifically, when there is a high congruence, the system is considered as tight. At the same time, when there is a low congruence, a system is then considered as loose.

2.2

Professionals and their Work

Professionals are deemed to be the main asset for PSFs across different industries. However, specific knowledge and skills gained through their specialised educational training makes them have special traits and preferences towards the management control. To gain an understanding why professionals behave this way and how they work in their work environment, this section focuses on the professionals, professional service firms, and the professional service firm types.

2.2.1 Professionals

For starters, the word ‘professional’ has a wide variety of definitions (Watson, 2002). Often, people indicate professionals as people who get paid for doing a certain task. In the business world, managers are sometimes referred as professional managers. Commonly, it is implied that these managers are competent in their occupation. However, Watson (2002) argues that by simply saying that being paid for doing a task or being competent in their occupation does not distinctively define a professional from others.

Goodale, Kuratko, and Hornsby (2008) claim that a PSFs’ production capacity basically consists of professionals with abstract knowledge who apply this knowledge to complex and customised tasks. This perception is also shared by Von Nordenflycht (2010) as he also states that essential characteristic of a professional is having a specific knowledge or expertise to execute the tasks.

This knowledge can be gained through formal education or through learning on the job (Hitt, Bierman, Shimizu, & Kochhar, 2001). Hitt et al. (2001) believe that through extensive education and training will lead to a high level of knowledge in a specific area. Furthermore, they also suggest that professionals must engage in a continuous learning process to gain knowledge to be able to deliver the highest quality service as possible at any time. After all, in PSFs, professionals are viewed as the most important asset.

In addition to this, Von Nordenflycht (2010) also explains that due to the high knowledge intensity that the professionals possess, their preference for a bureaucratic form of organisation is mostly unlikely. The reason for this is because professionals put a much more emphasis on the social status and autonomy than on the formal controls. In other words, they enjoy their workplace much more where they have more freedom to act.

(15)

Summarising, professionals are individuals who have a high level of knowledge, gained through education and their job, to be able to execute complex and customised tasks. Additionally, the professionals enjoy a less bureaucratic working environment much more than working in an environment where processes, procedures and rules are standardised.

2.2.2 Professional Service Firms

Based on six dimensions drawn from the framework of Silvestro, Fitzgerald, Johnston, and Voss (1992), Auzair and Langfield-Smith (2005) distinguished service firms in different types. The six dimensions are equipment/people focus, customer contact time per transaction, degree of customisation, degree of discretion, value added back-office/front-office, and product/process focus. Consequently, they identified three types of service firms: (a) professional services; (b) mass services; and (c) service shop. The last one is a hybrid between the first two.

(a) Professional services are identified as organisations with relatively few transactions, highly customised, process-oriented, with relatively long contact time, with most value added in the front-office, and where considerable judgment is applied in meeting customer needs. Examples are management consultancies and corporate banking. (b) Mass services have many customer transactions, involving limited contact time and little customisation. The offering is predominantly product-oriented with most value being added in the back-office and little judgment applied by the front-office staff. Examples are tobacco and newspaper retailers and transport. (c) Service shop is a categorisation that falls between professional and mass services. Examples are hotels and rental services.

Von Nordenflycht (2010) revised the characteristics of a PSF by categorising the characteristics defined in different studies in the same category. As for the result, he identified three distinctive characteristics of a PSF: (a) knowledge intensity; (b) low capital intensity; and (c) professionalised workforce.

Chan, Yim, and Lam (2010) and Erramilli and Rao (1993) also define similar characteristics. Chan et al. (2010) mention that professional services feature high credential qualities, high degrees of customer contact and customisation, and high independence between customers and service providers. Erramilli and Rao (1993) also define professional services as having high level of professional skills, specialised know-how, and customisation.

Based on these viewpoints, it is important to note that it seems that the extent of customer reliance is viewed as one of the central points when differentiating professional services from normal services and other type of organisations. For the former, for example, Auzair and Langfield-Smith (2005) define professional services as organisations with relatively few transaction, highly

(16)

customised outputs and relatively long contact times as opposed to the mass services that have many customer transactions, limited contact time and little customisation.

In summary, this study defines a PSF as a type of firm having main characteristics such as having relatively few transactions, high customisation, are process-oriented, long contact time with customers, and high value creation in the front-office. In addition, PSFs are also characterised as having a high knowledge intensity, low capital intensity, and a professionalised workforce. These are the key characteristics that differentiate PSFs from other types of services and organisations. 2.2.3 Professional Service Firm Types

As Von Nordenflycht (2010) mentioned, there is a need for researches going beyond the typical PSFs. Hence, this study will not only focus on the ‘Classic PSFs’, but also on other type of firms that are also classified as a PSF. To categorise the different PSFs, this study applies the framework of Larsson and Bowen (1989). This framework shows how the two contingent variables of input uncertainty creates four categories. Figure 1 illustrates this framework.

Figure 1: A typology of service interdependence patterns matching input uncertainty. Adapted from “Organization and Customer: Managing Design and Coordination of Services,” by R. Larsson and D.E. Bowen, 1989, The Academy of Management Review, 14(2), p. 221.

(17)

Larsson and Bowen (1989) state that uncertainty in this concept is defined as incomplete information about productions or tasks. The researchers explain that customer inputs are a source for this incomplete information, since service providers do not know what kind of inputs customers will have before the service process. Thus, as final definition, they state that “customer-induced input uncertainty is the organisation's incomplete information about what, where, when, and how customer input is going to be processed to produce desired outcomes” (p. 217). In addition, it is important to note that the input uncertainty must not only be seen at the perspective of a service provider. The customers may have a picture of what the service provider can offer and have certain expectations (Mills & Morris, 1986), but before the actual encounter, they cannot be fully certain about what the service provider can offer them. Hence, this can also be an input uncertainty for a production or task at the perspective of the customer as they may not precisely know what kind of resources are available before the initial encounter. This framework, however, looks at the input uncertainty of the customer, and thus, focuses on the perspective of the service provider.

Input uncertainty varies with two environmental variables: (a) diversity of customer demand and (b) the tendency of customers to participate in the service process (Larsson & Bowen, 1989). Diversity of demand refers to that each customer’s demand is unique. This includes the input, but also the output of the service process. A low diversity of demand means that the output is standardised. A high diversity of demand, on the contrary, means that the output is more unique. The second variable is referred as customer disposition. This concept involves the degree of customers willing to actively participate in supplying service providers with inputs to the service process itself. When the customers are more willing to participate in the service process, it is more likely that there is a higher level of input uncertainty, because the service provider has more incomplete information about what the customers are going to do before the encounter. This willingness to participate stems from two reasons. The customers either want to do it themselves, because they find it intrinsically attractive or they believe that being actively involved in the process guarantees the quality. The latter is also related to the moral hazard risk (Homburg & Stebel, 2009). In this case, the customer (principal) believes that by actively participating in the process reduces the moral hazard risk by reducing the information asymmetry between them and the service provider (agent) (Larsson & Bowen, 1989).

Putting these two variables against each other, Larsson and Bowen (1989) created a framework, in which they identify four types of service firms: (a) Sequential Customized Service Design; (b) Pooled Service Design; (c); Reciprocal Service Design; and (d) Sequential Standardized Service Design.

(18)

This study focuses only on the two quadrants, namely the Pooled Service Design firm and the Reciprocal Service Design firm. The other two categories, Sequential Customized Service Design and Sequential Standardized Service Design, are, in this study, disregarded as they do not conform to the definition of a PSF defined earlier. The Pooled Service Design firm has a low diversity of demand and a low customer disposition to participate. In this context, this means that the customer reliance is low. For example, we see this in banks and insurance companies. The Reciprocal Service Design firm, on the contrary, has a high diversity of demand and a high customer disposition to participate, which means that this perception is considered as having a high customer reliance. This is seen, for example, at medical care and legal organisations.

2.3

Customer Reliance

Customer reliance refers to the level of customer involvement (Homburg & Stebel, 2009). In fact, a service cannot be finalised without the integration of external factors, e.g. customers (clients). The researchers define the external factors as production factors that is related to the customers. To produce a service, the production factors of the customer and the production factors of the service provider need to be combined. They further argue that professional services are typically characterised with a high level of customer interaction. Therefore, customers from PSFs have a big influence on the process and the output of the service.

Larsson and Bowen (1989) support this notion. They state that the production and the delivery of the service require a considerably amount of contact between the customer and the organisation, which means that customers often participate in the service process. Furthermore, they mention that due to the shift from a manufacturing-oriented economy to an economy that is more dominated by services, the interaction between the organisation and the customers also becomes more important within the analysis of the organisational theory.

In this study, customer reliance is described as the extent the customer is involved in the process to come to a satisfied result. In other words, customer interaction is not only about customers providing information to the service provider, but also the other way around. More importantly, the multiple interaction between the customer and the service provider, from a customer reliance perspective, is argued to be the key factor in achieving a satisfied result (Fließ & Kleinaltenkamp, 2004).

In summary, since the input of professional services are unknown each time, the process cannot start before a service provider has been given the information to use it as the input. PSFs must rely on their customers who provide them this information. For example, a doctor cannot recommend a certain treatment before examining the health condition of a patient. A lawyer can

(19)

only give their client legal advice after knowing the details of the case. Additionally, architectures also listen to the customer’s wishes first before starting the whole process of planning, designing and constructing buildings or structures. Even in organisations where service providers can already pre-determine small pieces of information before the initial encounter, service providers can still not provide the solution before fully understanding the customer’s situation. In both cases, this means that the service providers can only start the service process when sufficient information has been provided by the customers. In these examples, it also shows that not only customers provide information for the service provider, but also the service provider provides information based on the information given by the customers.

(20)

3 Hypothesis Development

Auzair and Langfield-Smith (2005) suggest that the use of MCS varies according to the service process type. For PSFs, they find that professional services use less bureaucratic controls. A bureaucratic form of controls is associated with action, formal, tight, restricted, and impersonal controls (Auzair & Langfield-Smith, 2005). A less bureaucratic form of controls, on the other hand, is associated with results, informal, loose, flexible, and interpersonal controls. In addition to that, they also addressed the tight and loose dimension based on the nature of task. They believe that when tasks are easy to define, tight controls ensure that employees are acting as the organisation wishes. Since the task in professional services are hard to define, loose controls are more suitable in PSFs. They argue that tight controls in the situation with a high task uncertainty prohibits creativity, and thus, could lead to dysfunctional behaviour. Putting it differently, in this case, this means that loose controls are used as a supplement for the lack of task certainty. Eisenhardt (1985) also uses the task (un)certainty as a determinant for whether action controls or results controls are more effective in either of the situation. His study shows that when tasks are uncertain, results controls are more likely to be used. Consistently, Homburg and Stebel (2009) argue that a service process cannot be measured before it has finished. However, in the case of PSFs, customers are needed before the start of service the process and during the service process. Hence, this also suggests that results controls are more effective in this situation.

Given the high task uncertainty, and thus, a high customer reliance is needed, PSFs focus more on the results controls than the action controls when measuring the service process. This gives a straightforward proposition about professional services more likely adapting results controls than action controls in their MCS, leading to the following hypothesis:

H1a: Customer Reliance has a positive effect on the Explicit Results Controls

Fitzgerald, Johnston, Brignall, Silvestro, and Voss (1991) find that professional services are more flexible than mass services in terms of interacting with customers, which has an association with having a less formal system. An informal system allows professional services to have some discretion due to the specific requests from each customer. Pierce and Sweeney (2005) and Hartline, Maxham, and McKee (2002) share this viewpoint. Pierce and Sweeney (2005) imply that when the task uncertainty is high, firms are more likely to put an emphasis on the less bureaucratic forms of controls. Hartline et al. (2000) also suggest that due to the variability of service requests and the uncertainty associated with direct customer involvement, it creates an environment where formalisation is less effective.

(21)

Based on these perceptions, this means that when organisations faces a high task uncertainty, they need to rely on their customers as customers provide the necessary information to decrease the task uncertainty. However, this is only possible when flexibility is the key point and informal controls are used in the MCS. Thus, the following relation is expected:

H1b: Customer Reliance has a negative effect on the Implicit Results Controls

Professionals in Pooled Service Design firms are most likely subjected to action controls than to results controls. The interaction level between the customer and the Pooled Service Design firms is much lower than the interaction level between the customer and the Reciprocal Service Design firms. Due to the low demand of diversity and the low customer disposition to participate, professionals in Pooled Service Design firms are more active in the back-office than in the front-office. This makes it difficult to measure the results controls, because professionals in this type of firms are not directly involved with the customers. Therefore, it is expected that customer reliance in relation with Pooled Service Design firms will less likely result in explicit results controls than in relation with Reciprocal Service Design firms. Customers of Pooled Service Design firms are less likely to participate in the process due to the standardised processes as opposed to customers of Reciprocal Service Design firms, whereas the process is more unique. Thus, in this case, flexibility is less needed. Hence, it is more likely that Pooled Service Design firms use tighter implicit results controls. This leads to the following hypotheses:

H1c: Customer Reliance in relation with a Pooled Service Design Firm has a less positive effect on the Explicit Results Controls than in relation with a Reciprocal Service Design Firm

H1d: Customer Reliance in relation with a Pooled Service Design Firm has a less negative effect on the Implicit Results Controls than in relation with a Reciprocal Service Design Firm

On the other hand, Auzair and Langfield-Smith (2005) also find that mass services use more bureaucratic controls and tight controls. This comes forth from the case studies of Fitzgerald et al. (1991). In these case studies, the researchers found that quality of PSFs are measured by looking at the customer satisfaction level. In mass services, the relation is more between a customer and the whole organisation instead of an individual in an organisation. This means that with less direct responsibility for quality and less discretion on the front-office, compared to professional services, many more structured quality control systems are needed. Typically, this also signifies that service quality measurements in mass services are more structured and formal. Since the tasks in this type of service firm are more standardised than in professional services, a tight control system is more suitable. So, in this type of service firm, they have a more bureaucratic form of organisation in place.

(22)

This means that when the customer reliance is low, more tight action controls are used. Putting this in another way, action controls can be viewed as an alternative for the low customer reliance that compensates for the lack of customer’s involvement due to the lack of variety in the service. In this case, manuals, guides, and procedures are easier to drawn upon during the service process. So, as opposed to the need of results controls for firms with high customer reliance, action controls are less effective. Similar as to the need for flexibility reasoned for the implicit results controls, a negative effect is expected between the customer reliance and the implicit action controls as a less formal system is associated with high customer reliance. This leads to the following expected relations:

H2a: Customer Reliance has a negative effect on the Explicit Action Controls H2b: Customer Reliance has a negative effect on the Implicit Action Controls

Pooled Service Design firms put most of their value in their back-office, because customers are less involved in the service process, so less direct customer interaction is needed. The reasons for customers to not participate less in these processes could be that the service processes are standardised. The standardised processes in this type of firm also suggest that tasks are easier to determine. For similar reasons, reasoned for the results controls tightness constructs in both PSF types, Pooled Service Design firms focus more on both action controls tightness constructs as opposed to the Reciprocal Service Design firms. This means that it is expected that customer reliance in relation with Pooled Service Design will have a less negative effect on both action controls tightness constructs. This leads to the following propositions:

H2c: Customer Reliance in relation with a Pooled Service Design Firm has a less negative effect on the Explicit Action Controls than in relation with a Reciprocal Service Design Firm

H2d: Customer Reliance in relation with a Pooled Service Design Firm has a less negative effect on the Implicit Action Controls than in relation with a Reciprocal Service Design Firm

In firms where services are more standardised, these firms appear to be less flexible (Fitzgerald et al., 1991). In this case, flexibility refers to the range of options that the customer can choose. Professionals, who are highly skilled, are often used to maximize this flexibility. Putting it differently, firms with unique processes are more flexible regarding the options that they can offer to the customers. Thus, these firms are less likely to put a strong emphasis on the personnel controls as opposed to firms where there is less flexibility.

Since professionals are required to have a certain specific knowledge to be able to serve the customers whose demands have a high level of uniqueness, firms in this situation put more

(23)

emphasis on selecting the right people and providing them adequate training, so that they can perform their task and serve their customers adequately. This means that these kinds of firms have stricter employee selection procedures and provide much more training for employees opposed to firms where services are more standardised. Based on these perceptions, this following relation is expected:

H3a: Customer Reliance has a positive effect on the Explicit Personnel Controls

Since professionals are considered the most important resource in their main processes for PSFs, little to no deviations from human resource standards are allowed. Quality of the service is often measured based on the customer satisfaction (Fitzgerald et al., 1991). To attain a satisfied customer satisfaction level, professionals need to act in line with the standards to be able to serve the customers in a proper manner. Therefore, a positive effect between the customer reliance and the implicit personnel controls is expected, leading to the following proposition:

H3b: Customer Reliance has a positive effect on the Implicit Personnel Controls

Service firms that have standardised services are more likely to put less emphasis on the personnel controls, because there is less flexibility, and thus, highly skilled professionals are less needed. However, service firms that have more customised services have this kind of flexibility, and thus require them to employ highly skilled professionals. This means that these firms are more likely to put a strong emphasis on the personnel controls. In this regard, the first notion corresponds with the Pooled Service Design firms while the second notion corresponds with the Reciprocal Service Design firms. This means that for Pooled Service Design firms, it is expected that the relation between customer reliance and explicit personnel controls is less positive. For the implicit personnel controls, on the other hand, quality of a service is considered as an important issue in both types of PSFs. Both PSF types also consider the professionals as an important asset. Hence, in Pooled Service Design firms, implicit personnel controls will be equally tight as in Reciprocal Service Design firms. These perceptions lead to the following hypotheses:

H3c: Customer Reliance in relation with a Pooled Service Design firm has a less positive effect on the Explicit Personnel Controls than in relation with a Reciprocal Service Design firm

H3d: Customer Reliance in relation with a Pooled Service Design firm has no different effect on the Implicit Personnel Controls as when in relation with a Reciprocal Service Design firm

Next, Auzair and Langfield-Smith (2005) also addressed the impersonal and interpersonal control style of each service type of organisation by looking at the leadership style based on the research of Otley and Pierce (1995). Service firms that relatively faces a lower level of task

(24)

uncertainty put their emphasis on the impersonal controls as it is associated with the more bureaucratic style of organisation (Auzair & Langfield-Smith, 2005). In this case, controls are uniformly applied thorough the whole organisation without taking the individuals’ personalities and preferences into account.

In public accounting firms where the working environment is highly uncertain, a structured style of leadership is highly ineffective (Otley & Pierce, 1995). In this case, a structured style of leadership is associated with tight controls. A considerate style of leadership, on the other hand, is associated with mutual trust and respect, helpfulness, and friendliness. This supportive working environment effectively addresses the uncertainty in the working environment (Otley & Pierce, 1995).

Elaborating on this perceptions, this means that firms facing a higher task uncertainty have an environment where employees are considerate and helpful with each other. Thus, these firms are more likely to put a strong emphasis on their cultural controls. On the other hand, firms, where the level of the certainty of the task is higher, are more likely to put a less stronger emphasis on the cultural controls, since a supportive environment is less necessary. In this environment, it is important that employees have the same norms, values, and beliefs to create synergies between the employees. Hence, deviations from the norms, values, and beliefs are less tolerated. Taken these notions into account, this leads to the following propositions:

H4a: Customer Reliance has a positive effect on the Explicit Cultural Controls H4b: Customer Reliance has a positive effect on the Implicit Cultural Controls

Since service processes are more standardised in Pooled Service Design firms, a high task uncertainty does not apply to this type of firm. Employees face less uncertainty and depend less on their colleagues. So, creating a helpful and supporting working environment is less important for Pooled Service Design firms than for Reciprocal Service Design firms. In other words, the effect between customer reliance and the cultural controls tightness is expected to be less positive for Pooled Service Design firms. Consequently, this leads to the following hypotheses:

H4c: Customer Reliance in relation with a Pooled Service Design firm has a less positive effect on the Explicit Cultural Controls than in relation with a Reciprocal Service Design firm

H4d: Customer Reliance in relation with a Pooled Service Design firm has a less positive effect on the Implicit Cultural Controls than in relation with a Reciprocal Service Design firm

(25)

Figure 2 illustrates the model for the hypotheses. Depending on the customer reliance of both PSF types, one PSF type should lead to using a certain control in a more, less, or same intensive manner compared to the other PSF type.

(26)

4 Research Method

This section addresses the research method that have been used in this study. This study comes forth from the PSF Thesis Survey Project 2016-2017 under guidance of H. Kloosterman at the University of Amsterdam. To be part of this project, every individual had to find at least ten respondents for the survey before they were granted access to the dataset. The participants are encouraged to find possible respondents of the survey within their personal network, as this ensures that the respondents will, more likely, answer the survey. This section discusses the survey, the sample, the constructs, its validity and reliability, and ends with the statistical models for the analyses in the next chapter.

4.1

Survey

This project investigates how PSFs design their MCSs and what kind of effects these systems have on the professionals working in this type of organisation. As such, the dataset collected through the digital survey of this project will help in answering the research question in this study.

The survey is designed by H. Kloosterman, whereby many constructs related to the MCS design are included, such as human capital intensity, task complexity, environmental uncertainty, strategy, professionalised workforce, organisational reputation, organisational size, organisational type, compensation/reward structure, organisational structure, customer reliance, customer tightness, professional tension, and performance. In total, there are 42 questions in the survey with the addition of descriptive questions. The survey is presented in Appendix B.

4.1.1 Pre-tests for the Survey

For the survey, two separate pre-tests were conducted by H. Kloosterman. The pre-tests tested the survey on its quality and effectiveness.

The first pre-test was conducted to ensure the quality of the items used to measure the constructs for each type of management control. This was done by an item sort task. For the task, two sheets of paper were handed out to the participants. The first paper consisted of the construct definition for the eight control constructs (implicit/explicit – results/behaviour (action)/cultural/personnel control). The second paper consisted of 52 statements to test those constructs. The participants were asked to match each statement with each control construct. There were fourteen participants in this test, all working in a PSF. The number of correct and incorrect matches was put in a table. Then, the four statements for each type of control that had the least number of incorrect matches were put into the survey. Finally, the 32 statements that were included in the survey ranged from a low of zero incorrect matches (explicit action control) to a high of six incorrect matches (implicit action control).

(27)

The second pre-test made sure that the quality of the whole survey was sufficient. For this test, an additional of 20 participants were asked to give feedback on the digital survey. They were asked to comment on the content, clarity, the appearance of the survey, and the amount of time they needed to complete the survey. Thirteen participants answered through written answers while the remaining participants answered by phone. Based on their comments, only a few minor changes in wording were done and additional multiple choice questions were also included in the survey.

4.2

Sample Criteria

The sample population consists of professionals working in a PSF. The survey indicates that non-profit organisations, publicly owned organisations, government organisations, NGOs, and social work agencies are not considered as PSFs. Hence, these types of organisations are not included in the survey. Furthermore, professionals must be also subjected to the MCS. In other words, owners, partners and board members of an organisation do not qualify. Furthermore, the professionals must have also worked more than three years in this field before they can fill in the survey. This is because in most PSFs employees, who have worked less than three years in the field, are subjected to a tighter control system than when they have worked several years in the field. Finally, the respondents must work in a firm that has over 50 employees to be able to be classified as a medium to large firm, as such firms have a MCS present compared to smaller firms.

4.3

Sample Selection

The initial dataset has a total of 612 respondents in which 312 respondents has been selected as the final total respondents based on the sample criteria mentioned in the previous section.

First, 89 respondents from the 612 respondents, who have not completed the survey, have been removed. Since it is important in this research to specifically identify professionals, 7 respondents have also been excluded for not stating their occupation. In addition to that, 91 respondents have stated that their expertise falls into the ‘other’ category. Despite given a wide range of PSF occupation to choose from, the respondents who have chosen the ‘other’ category are difficult to determine whether they pass the PSF criteria or not. However, upon further inspection, it has been decided to not remove the audit and the controlling occupation in the ‘other’ category as both occupations obviously fall into the accounting occupation category. This means that only 80 of the 91 have been removed from the ‘other’ category. Finally, 124 respondents have either not filled in that how many years they have worked in their field or that they have worked less than three years in their field. Hence, these respondents are also ruled out from the sample.

With 312 respondents as the final dataset sample, 175 respondents are classified as working in a reciprocal service design firm and 137 respondents are classified as working in a pooled service

(28)

design firm. The distribution of the PSF types per industry can be found in Appendix C. Table 1 illustrates the sample selection.

Table 1: Sample Selection

As for the distribution of the industries, Figure 3 shows that the largest group in the final dataset consists of professionals active in the accounting field with a total of 18.27%. This group is followed by a group of professionals active in the medicine/physician practices who make up 10.58% of the final dataset. The third, fourth, and fifth largest groups are consulting management/strategic, consulting IT, and research/R&D with 8.01%, 7.69%, and 7.69% respectively.

Figure 3: Distribution of the Occupations

Further examining the final dataset sample, Table 2 shows that the ages of the respondents range from 23 to 64 with an average age of 37. Gender-wise, most of the respondents are males

Initial Respondents 612

Survey not completed 89

Occupation not mentioned 7

'Other' occupation 80

Experience criteria not met 124

Final Respondents 312

Reciprocal Service Design Firm 175

Pooled Service Design Firm 137

1 1 1 2 2 3 3 4 4 4 4 5 7 7 8 9 9 10 12 14 18 21 24 24 25 33 57 0 10 20 30 40 50 60 Graphic design Actuarial services Consulting Engineering Consulting Technology Architecture Software development Talent management/agency Investment management (hedge funds, VC,mutual funds) Financial advising Real estate Media production (film, TV, music) Recruiting - executive Investment banking Other Risk management services Law Marketing/public relations Pharmaceutical Consulting HR Biotechnology Project management Engineering Consulting IT Research/R&D Consulting Management/Strategic Medicine/Physician practices Accounting Respondents In d u st ry

(29)

(66.99%). As for the education, respondents with a bachelor degree or lower are a slightly larger group (41.76%) than the groups with respondents holding a master degree (40.06%). Lastly, the respondents have an average of 8 years of work experience in their field.

Table 2: Descriptive Statistics of Respondents

4.4

Constructs used from the Survey

The survey consists of an extensive composition between existing items and new items of the constructs. This research specifically focuses on the customer reliance, control dimensions, the control tightness constructs in the survey, and the control variables. The questions related to each of the constructs can be found in Appendix D.

4.4.1 Customer Reliance

Customer reliance is measured by six statements, whereby respondents indicate their extent of agreement on a Likert scale from 1 (strongly disagree) to 5 (strongly agree). The statements related to this construct will help in answering the question to what extent professional service firms depend on the customers’ cooperation and collaboration for a successful service provision. For this research, this variable is the independent variable for the analysis. An independent variable is a variable that is thought to be a cause of an effect (Field, 2014). This construct contains previously unused items, so it is viewed as a new item. For item 3, reverse coding was needed (1 = 5, 2 = 4, 3 = 3, 4 = 2, and 5 = 1) as it was phrased the opposite direction opposed to the other items within this construct.

4.4.2 Control Tightness Dimensions

In this study, the control dimensions are the dependent variables. A dependent variable is a variable that is thought to be affected by the changes in the independent variable (Field, 2014). Numerous

Percentage N Minimum Maximum Mean

Std. Deviation Age 310 23.00 64.00 37.40 9.00 Gendera 310 1.00 2.00 1.33 0.47 Male 66.99% Female 32.37% Educationb 309 1.00 3.00 1.75 0.73 Bachelor Degree 41.76% Master Degree 40.06% Doctorate Degree 17.31% Experience in Fieldc 312 5.00 11.00 8.47 2.41 a. Male = 1, Female = 2

b. Bachelor Degree or lower = 1, Master Degree = 2, PhD or other Professional Doctorate Degree = 3 c. 4 years = 5, 5 years = 6, [...], 10 years or more = 11

(30)

statements were also given to the respondents with the same five-point Likert scale (1 = strongly disagree and 5 = strongly agree). The statements are related to the four types of control dimensions, which are results controls, action controls, personnel controls, and cultural controls.

The control tightness per control dimension is defined in two ways. First, the degree of extending the scope of the MCS (explicit), and second, the degree of acceptance of deviations from the MCS (implicit). As of now, with the addition of a breakdown of the control tightness for each control dimension, new items have been added in the survey. Hence, the concept of explicit and implicit controls is new in this survey. Respondents who have indicated that they strongly disagree (1) with a statement imply that loose controls are in place. While respondents who strongly agree (5) with a statement imply that tight controls are in place.

In the survey, there are eight items given that regards the results controls. Items 2, 3, 4, and 5 focus on the explicit results controls while items 1, 6, 7, and 8 focus on the implicit results controls. For items 6, 7, and 8, its scale needed to be recoded in reversal. Action controls also have eight items. While items 1, 2, 3, and 5 focus on the explicit side of the control tightness, items 4, 6, 7, and 8, on the other hand, focus on the implicit action controls side. The scale of all items of the implicit action controls needed to be recoded in reversal as well. Regarding the personnel controls, eight statements were also given to the respondents to indicate their extent of agreement on each statement. Items 1, 2, 4, and 5 relate to the explicit personnel controls. Items 3, 6, 7, and 8 are related to the implicit personnel controls instead. For this control dimension, only the scale of item 3 needed to be recoded in reversal. For the cultural controls, eight statements were also given to the respondents. Items 1, 5, 6, and 7 are related to the explicit cultural controls while items 2, 3, 4, and 8 are related to the implicit cultural controls. Just like the personnel controls, only the scale of item 3 needed to be recoded in reversal.

4.4.3 Professional Service Firm Type

In the survey, the respondents indicated in which field of occupation they are actively working in. Based on their answers, the field of occupation has been divided into two types of PSFs, Reciprocal Service Design firm and Pooled Service Design firm, according to the framework of Larsson and Bowen (1989). This also led to a creation of a dummy variable, using Reciprocal Service Design firms (code = 0) as the reference group and Pooled Service Design firms (code = 1) as the other group. This research seeks to find whether the PSF type influence the relation between the customer reliance and the MCS design. This means that the PSF type will act as the moderating variable between this relation. A moderating variable is a variable that affects the relation between the independent variable and the dependent variable.

(31)

4.4.4 Control Variables

Control variables are variables that are used to rule out alternative explanations of the relation between the independent variable and the dependent variables (Spector & Brannick, 2011). For the regression analysis, it has been decided to include the environmental uncertainty, strategy, and organisation size as control variables as literature has proven that these factors do affect the design of a MCS (Gordon & Narayanan, 1984; Auzair & Langfield-Smith, 2005; Chenhall, 2003).

Environmental uncertainty refers to the (un)predictability of the business environment the organisation faces (Merchant & van der Stede, 2012). Gordon and Narayanan (1984) argue that when organisations are surrounded with a higher level of environmental uncertainty, these organisations tend to seek external, non-financial, and ex ante information while also moving toward and organic form of organisation. An organic form of organisation is perceived as an organisation not having highly specified job description and formal procedures (Gordon & Narayanan, 1984). This means that when an organisation faces a high level of environmental uncertainty, a less bureaucratic form of organisation is more effective. On the other hand, Merchant and van der Stede (2012) believe that PSFs facing a high environmental uncertainty should use a carefully chosen combination of controls, which include extensive planning and budgeting processes (to perform pre-action reviews and to ensure coordination), individual and team measures of performance and associated incentives (to ensure that the individual and team are provided with the right compensation and incentives), personnel controls (to ensure that the employees have the required knowledge and that this knowledge is shared), and cultural controls (to ensure that there is teamwork and that knowledge is shared amongst the employees). The respondents were asked to indicate the level of intensity in their industry regarding the price competition, competition for manpower, and bidding for new contracts/clients on a five-point Likert scale (1 = Of negligible intensity and 5 = Extremely intense).

The control variable strategy is based on the research of Auzair and Langfield-Smith (2005), in which they explain that regarding the strategy focus of the organisation, whether it has a cost leadership or differentiation strategy, it affects the design of a MCS. They find that organisations using a cost leadership strategy place a stronger emphasis on the bureaucratic form of MCS than organisations pursuing a differentiation strategy. However, Merchant and van der Stede (2012) argue that regardless of the strategy focus of the organisation, the requirement still stands to have an intensive coordination and collaboration across employees and organisation entities. For this construct, respondents were asked to indicate the extent of emphasis their organisation places on 11 items on a five-point Likert scale (1 = Very Little Emphasis and 5 = A Great Deal of Emphasis).

(32)

Items 2, 4, 6, and 11 are related to the cost leadership strategy while items 1, 3, 5, 7, 8, 9, and 10 are related to the differentiation strategy.

The last control variable is the organisation size. Chenhall (2003) believes that the organisation size does influence the design of a MCS. He finds that as the organisation becomes larger, the need of information quantity increases. This leads to that the need for controls for rules, documentation, specialisation of roles and functions increases (Chenhall, 2003). This perception is also shared by Woods (2009) and Sharma (2002). Both researchers also find that large organisations tend to have a formal and tighter MCS. For this construct, respondents were asked to indicate how many employees were employed in their whole organisation (1 = Less than 100, 2 = More than 100 but less than 500, 3 = More than 500 but less than 5000, and 4 = More than 5000).

4.5

Validity of the Constructs

First, it is important to make the statements face one direction to be able to determine whether the scale is reliable using the Cronbach’s a. Questions that have been phrased oppositely can mess up the calculation of the Cronbach’s a. Hence, the scales of the questions that were phrased the opposite direction were first recoded in reversal before running the factor analysis and the reliability analysis. Since the construct organisation size is not measured on an ordinal scale, it would not make sense to analyse this construct with the factor analysis and/or the reliability analysis.

4.5.1 Principal Component Analysis

A PCA is one of the factor analysis techniques1 to identify clusters of variables (Field, 2014). A

PCA calculates the maximum amount of the total variances, and put it in a correlation matrix using the smallest number of explanatory constructs by transforming the original variables into linear components. These explanatory constructs are called factors or components. These factors represent clusters of variables that highly correlate with each other. In a nutshell, a PCA looks for variables that highly correlate with each other in a group and does not correlate with other variables outside the group. After conducting the PCA on each construct, an additional PCA will also be conducted by putting all constructs into the PCA.

First, this analysis computes the Eigenvalues (EV). This EV shows the importance of each factor. Based on this EV, a decision is then made about how many factors should be extracted. Kaiser (1960) recommends extracting all factors with EVs higher than 1. Once the factors have been extracted, the PCA also calculates the loading for each item on each factor. Stevens (2002)

1 The other one is the standard factor analysis. In this case, the PCA has been used as this analysis reduces the data

Referenties

GERELATEERDE DOCUMENTEN

Some of the factors that could play a role in the prevalence of burnout and work engagement are secondary traumatic stress, the demands of counselling, lack of

Enquêtes zijn echter notoir onbetrouwbaar bij het achterhalen van de opkomst, omdat mensen vaak worden gestimuleerd om te gaan stemmen door de afspraak die ze voor de

Usually, as universities are not-for-profit organizations, they are well placed to create social value by promoting and/or supporting social innovation initiatives addressing

network organisational structure (Vicari, 2014); ‘project identity’ (Castells, 2011) is associated with activists who seek to build strong solidarity in their networked communi-

Abstract: Extending the results of Bellec, Lecu´ e and Tsybakov [ 1 ] to the setting of sparse high-dimensional linear regression with unknown vari- ance, we show that two

This exploratory study shows an increase in knee kinematics in the swing phase after functional electri- cal stimulation of the hamstrings in stroke survivors walking with

The physical modelling of tire-road interaction phenomena and the employment of advanced simulation tools developed by UniNa Vehicle Dynamics research group and engineered by its

Our analysis of the cases consists of the following steps: (1) we identify driving forces in demand articulation in the domain, including technologies and regu- latory