Amsterdam Business School
Control tightness, task complexity and individual performance
in professional service firms
Name: Tarik R. Samandar Student number: 10164626
Thesis supervisor: dr. ir. S.P. (Sander) van Triest
Date: June 20th, 2016 Word count: 16577
MSc Accountancy & Control, specialization Accountancy Faculty of Economics and Business, University of Amsterdam
!
Statement of Originality
This document is written by Tarik Samandar 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 & Business is responsible solely for the supervision of completion of the work, not for the contents.
! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !
Abstract
An extensive amount of research has been devoted to examining management control systems in “traditional” organizations (Abernathy & Brownell, 1997). The conventional view has been that tight controls are beneficial. Recently, there has been a great interest in professional service firms (Greenwood, Li and Deephouse, 2005). This paper investigates if there is a relationship between control tightness and individual performance of professionals employed in professional service firms. In particular, the focus of this paper is on the so-called formal controls, which comprise results and behavior controls. It is argued that control tightness is negatively
associated with performance. In addition, this paper investigates whether task complexity has an impact on this relationship. Furthermore, it is being examined whether the relationship between control tightness and individual performance is mediated by professional tension/stress. In order to conduct this research, surveys have been collected from 372 professionals employed by professional service firms in over 30 fields. The findings of this research support two of the hypotheses
presented throughout this paper. Contrary to what is being argued, overall, behavior control tightness is positively related to individual performance (in most models). Furthermore, it is shown that task complexity negatively influences the relationship between control tightness and individual performance.
! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !
Table of Contents
1. INTRODUCTION!...!5! 2. LITERATURE REVIEW!...!7! 2.1!PROFESSIONAL!SERVICE!FIRMS!...!7! 2.1.1!Task!complexity!in!professional!service!firms!...!8! 2.2!MANAGEMENT!CONTROL!SYSTEMS!...!10! 2.3!CONTROL!SYSTEM!TIGHTNESS!...!12! 2.3.1!Inconsistencies!across!research!...!12! 2.3.2!The!construct!of!tight!(budgetary)!control!and!its!attributes!as!described!by!Van!der! Stede!(2011)!...!13! 2.3.3!Control!tightness!in!the!present!study!...!14! 2.4!CONTINGENCY!THEORY!...!15!2.5!PROFESSIONALS!&!THE!NOTION!OF!TASK!AUTONOMY!...!16!
2.6!ROLE!CONFLICT,!ROLE!AMBIGUITY!AND!PROFESSIONAL!TENSION/STRESS!...!18!
3. HYPOTHESES DEVELOPMENT!...!20!
4. DATA & RESEARCH METHODOLOGY!...!24!
4.1!SURVEY!...!24! 4.2!RESPONDENTS!...!25! 4.3!CONSTRUCTS!...!26! 4.3.1!Results!control!tightness!...!29! 4.3.2!Behavior!control!tightness!...!29! 4.3.3!Professional!tension/Stress!...!30! 4.3.4!Task!complexity!...!30! 4.3.5!Individual!performance!–!In!role!...!31! 4.3.6!Individual!performance!–!General!...!31! 4.3.7!Control!variables!...!32! 5. PRELIMINARY ANALYSES!...!33! 5.1!EQUATIONS!...!34! 6. RESULTS!...!37! 6.1!ADDITIONAL!TESTING!...!43!
7. CONCLUSION & DISCUSSION!...!44!
7.1!IMPLICATIONS!...!44! 7.2!LIMITATIONS!...!46! 7.3!FUTURE!RESEARCH!...!47! 8. REFERENCES!...!49! 9. APPENDICES!...!54! 9.1!APPENDIX!A!...!54! 9.2!APPENDIX!B!...!54! ! !
1. Introduction
!It has been stated by several authors (Abernathy & Brownwell, 1997; Von
Nordenflycht, 2010) that there have been considerable management accounting studies that explored the design of management control systems and the suitability of the controls being implemented. However, this stream of research has primarily focused on manufacturing firms where results and behavior controls have been well suited.
Substantial attention is now being directed towards examining professional service firms (PSFs) for several reasons (Von Nordenflycht, 2010). First of all, they are assumed to be different from “traditional” manufacturing firms. Their particular environment demands other theories of management (Von Nordenflycht, 2010), and thus another approach to the conventional approach of the design of management control systems in these firms. It is also noted that the work done so far is primarily based on accounting and law firms (Von Nordenflycht, 2010). Secondly, according to Greenwood et al. (2005), professional service firms are worthwhile studying because there is a growing importance and interest in knowledge as a competitive advantage. Third, PSFs comprise a large sector of the economy if measured by size, numbers or influence (Greenwood et al., 2005)
Despite the importance of professional service firms, they have not been researched sufficiently. According to Von Nordenflycht (2010), a noteworthy hindrance in studying and understanding professional service firms is the fact that the term PSF remains ambiguous. In order to address this ambiguity problem, Von Nordenflycht (2010) distinguishes 3 distinctive attributes of PSFs, which are
knowledge intensity, low capital intensity and professionalized workforce.
Furthermore he discusses the organizational and managerial implications of these three distinctive attributes.
Professionals regard themselves as employees of special status that demand autonomy and other special rights (Raelin, 1985). However, according to Raelin (1985), organizations that employ these professionals have been increasingly unwilling to grant these special status employees unusual rights that as a matter of fact conflict with the objectives and policies of management. Accordingly, there is an unavoidable struggle between management and the organizational professionals and Raelin (1985) states that there are several attributes of professional life that
make professionals oppose managerial control.
This paper examines the implications of control tightness on employees working in professional service firms. In particular, it is being examined whether there is a negative relationship between (results and behavior) control tightness and individual performance. Moreover, this study investigates if task complexity impacts this relationship, as Shields, Deng & Kato (2000) argue that the design of
management control systems is dependent on the nature of the job or task that is being controlled. Furthermore, it is being examined whether professional
tension/stress mediates the relationship between control tightness and individual performance. Therefore, the main question is the following: is there a negative relationship between control tightness and individual performance of employees working professional service firms? This main question can be divided into two sub questions: 1) does task complexity influence this relationship? and 2) does
professional tension/stress mediate the relationship between control tightness and individual performance?
This paper has several contributions. First of all, different types of professional service firms are examined, ranging from accounting firms, investment banks and law firms to advertising firms, thereby extending prior knowledge by looking at a wider concept of professional service firms. Second, in order to respond to the call of Von Nordenflycht (2010), to look at other dimensions of professional service firms, it is being examined whether task complexity influences the relationship between control tightness and performance. Third, this study also investigates the indirect effects of control tightness, as compared to many studies (Birnberg, Shields & Young, 1989; Shields & Shields, 1998; Kren & Liao, 1988) that only consider the direct effect of control tightness (Shields et al., 2000).
The remainder of this paper is structured as follows. Section two presents the literature review followed by the hypotheses development in part three. In section four, the data and research methodology are explained. This is followed by the preliminary analyses, presented in section five. In chapter six the results are discussed. Section seven ends with a conclusion/discussion and in sections eight and nine, the references and appendices (respectively) are presented.
2. Literature Review
In this section an overview is given of the current state of knowledge of professional service firms, management control systems, control system tightness, contingency theory, professionals & the notion of task autonomy and it ends with the last section that describes role conflict, role ambiguity and professional tension. Based on this theoretical background the hypotheses are formulated and presented in the section three.
2.1 Professional service firms
Professional service firms can be found in various industries and include
accountants, lawyers, investment bankers, architects, consultants, engineers, and advertising professionals, among others. According to Greenwood et al. (2005), professional service firms are worthwhile studying because there is a growing importance and interest in knowledge as a competitive advantage. Furthermore, PSFs comprise a large sector of the economy if measured by size, numbers or influence. There are several definitions provided for professional service firms. In the simplest sense, PSFs are organizations that employ a large amount of professionals. According to Greenwood et al. (2005) professional service firms are organizations whose fundamental assets are a highly educated (professional) workforce.
Furthermore, the outputs of these firms are services, which are intangible and contain complicated knowledge.
Von Nordenflycht (2010) describes the distinctive attributes that characterize professional service firms and the managerial implications as well as the
organizational responses to these distinctive attributes. After reviewing many articles the author focused on three attributes, which are knowledge intensity, low capital intensity and last but not least a professionalized workforce. The emphasis has been placed on these attributes for several reasons. First of all, the above-mentioned characteristics can be defined properly. Second, in other papers they are also commonly referred to as distinctive attributes of PSFs. Third, these attributes have been associated with distinctive managerial challenges and organizational outcomes (Von Nordenflycht, 2010). In the following paragraph these attributes are described in more detail.
Knowledge intensity, according to Starbuck (1992), points out that an organization’s output depends on a considerable body of complex knowledge and that it is possibly the most crucial, distinctive attribute of a professional service firm. However, the latter statement depends on how knowledge intensity is interpreted as it can refer to knowledge incorporated in individuals (Alvesson, 2000) or to
knowledge incorporated in physical assets, merchandise and organizational routines (Starbuck, 1992). In the context of PSFs, it is more appropriate to apply the former interpretation (Von Nordenflycht, 2010). Low capital intensity indicates that
production in an organization excludes large sums of non-human assets, like inventory, machinery, manufactories and also intangible non-human resources like copyrights and patents. Professionalized workforce is the third attribute worth
mentioning and this can be defined as the degree to which the profession is formally organized (Von Nordenflycht, 2010). In particular, Toress (1991) mentions that there are three crucial features of a professionalized workforce. These include a specific knowledge base, regulation and control related to that specific knowledge base and its implementation. The last feature is ideology and this relates to the “professional codes of ethics (Goode, 1957; Nanda, 2002), as well as less explicit norms that define appropriate behavior for professionals” as stated by Von Nordenflycht (2010, p. 169).
In addition to these 3 characteristics described by Von Nordenflycht (2010), this paper identifies customer reliance and task complexity as distinctive
characteristics of professional service firms. Customer reliance is the degree to which the successful provision of services is contingent on the client’s joint effort with the professional service firm (Homburg & Stebel, 2009). Task complexity, according to Jaworski, Stathakopoulos & Krishnan (1993, p. 60), is “the extent of predictability and variety in the activities to be performed for a given position”.
!
2.1.1 Task complexity in professional service firms
!
According to Jaworski et al. (1993) there are four dimensions of task complexity. These are interdependence, routineness, “learn-job time” and last but not least completeness of the tasks (Jaworski et al., 1993). Task interdependence is the extent to which team members’ performance depends on the performance of other team members (Ouchi & Maguire, 1975).According to indications of a number of
researchers like Dalton (1971) and Otley (1980) the problem of control is aggravated in the case that duties necessitate the collective effort of group members (Jaworski et al., 1993). According to Baumler’s (1971) research, formal control systems
(systems dominated by behavior and results controls) are adequate in the case that task interdependence is limited, but inadequate in situations where task
interdependence is high. For that reason, Lawrence & Lorsch (1967) argue that a clan system (focusing on personnel and cultural controls) is more suitable in situations characterized by high task interdependence (Jaworski et al., 1993). Another dimension of task complexity is task routineness and it is described as the extent of variety in the task (Hage & Aiken, 1969). According to Dalton (1971), task routineness should be viewed across a continuum. On the one hand, there are tasks that are subject to general operating procedures, prescribed rules, and explicit performance standards. On the other hand, there are certain tasks that cannot be specified, articulated or arranged beforehand. Therefore, Jaworski et al. (1993) argue that if tasks are more routine, formal controls are more suitable, whereas when tasks are at the other end of the spectrum they may not be adequate.
The third dimension of task complexity comprises “learn-job time”. This relates to the amount of time it is required to learn a job. In situations where it takes a
considerable amount of time to learn and understand certain tasks, Jaworski et al. (1993) argue that it is very reasonable to depend on co-workers for peer advice and recommendations. This directly relates to the present study where professionals (working in professional service firms) are being faced with complex decisions that need to be taken. Furthermore, Jaworski et al. (1993) state that in these situations employees will be continuously influenced by the values and beliefs of the
organization and thus a clan system will be more adequate.
The fourth dimension of task complexity relates to the degree to which the appraisal system appropriately captures the complete range of job duties and activities (Anderson & Chambers, 1985). In cases where the completeness of the tasks is high, Jaworski et al. (1993) argue that formal controls are likely to be dominating in organizations. In professional service firms, however, the successful completion of a task is dependent upon much more than the effort of only one
professional (Davila, 2000) and therefore it might not be appropriate to rely on formal controls. It is thus clear that professional service firms face a different environment compared to the more “traditional firms” and therefore require different management
theories and in turn the management control system (MCS) should also be designed in a different way to obtain the desirable outcomes (Von Nordenflycht, 2010). This brings us to the following section, which discusses management control systems (MCSs) in more detail.
!
2.2 Management control systems
!
When an organization has set out which strategies it is going to pursue and plans are made, it is management’s principal duty to make sure that these plans are executed properly, and if needed, that the plans are adapted (Merchant, 1981). The latter is a crucial function of management control in large companies and failures with respect to management control can lead to tremendous financial losses, damage in reputation and in worse cases also to firm failure (Merchant, 1981). Therefore, it is very important to have good management control systems in place. According to Merchant & Van der Stede (2012) management control systems include all the devices/systems that managers can use to make sure that the behaviors and decisions of the employees are aligned with objectives and strategies of the firm. If the management control systems are designed properly, the behavior of employees can be influenced in a desirable way such that the probability that the firm will achieve its targets will increase (Merchant & Van der Stede, 2012).
If all employees always acted in the best interest of the organizations, there would be no need for management and hence no need for a management control system (Merchant, 1981). Unfortunately, this is not the case and employees are at times unable or not willing to act in the best interest of the firm. Therefore it is necessary that suitable controls be implemented to avoid undesirable behavior and to persuade desirable actions of the employees.
Particularly, the cause of the need for management control systems can be grouped in three main categories. These categories are lack of direction,
motivational problems and personal limitations (Merchant & Van der Stede, 2012). Certain employees do not perform adequate, because of the simple fact that they do not know what the firm wants from them. According to Merchant and Van der Stede (2012), this lack of direction causes the desired behavior to occur irregularly. For the latter reason it is a crucial function of control systems to inform employees in what manner they can direct their contributions in order to fulfill the organization’s
objectives. In the case that employees know what is expected of them, certain employees still perform inadequately due to motivational problems. These problems arise because employee and firm objectives do not coincide perfectly (individuals are considered to be self-centered). Put differently, there is a lack of goal congruence and action should be taken to, on the one hand increase this congruency, or to make sure that employees do not act in their own interest when this goal incongruence exists (Merchant, 1981). In situations where employees know what the organization wants from them and are highly motivated as well, it is still possible for them to perform inadequately due to other (personal limitations) and management control systems should also address this behavioral problem (Merchant & Van der Stede, 2012).
There are four types of controls that can be implemented, to make sure that the problems described in the previous paragraph are addressed. These are behavior (action) controls, results controls, personnel controls and cultural controls (Merchant, 1981; Merchant & Van der Stede, 2012; Ouchi, 1979). Behavior controls are seen as the most blunt form of control as it entails taking steps to make sure that employees act in the best interest of the firm, by making their behavior themselves the center of control (Merchant and Van der Stede, 2012). Behavior control is defined as the use of processes, procedures, rules and routines (that are
standardized) to monitor, evaluate and reward the behavior of personnel with the aim of achieving organizational goals. Results controls entail the use of accounting
performance measures, such as goals or targets to evaluate and reward the actions taken by the employees. This is done with the goal of achieving organizational objectives (Merchant & Van der Stede, 2012). The two forms of control described above (behavior and results controls) are commonly referred to in the literature and practice as “hard” or “formal” controls (Merchant & Van der Stede, 2012). Personnel controls entail the use of employee selection policies. The aim of these controls is to maximize goal congruence between employee and organizational objectives
(Merchant, 1981). The last type of control is cultural control, which can be defined as the use of rituals and ceremonies to oversee, evaluate and reward employees who adhere to values, attitudes and norms of the organization. The aim of this last form of control is to have employees internalize the objectives of the organization. These last two forms of controls (personnel and cultural controls) are referred to as “soft” or “informal” controls (Merchant & Van der Stede, 2012).
2.3 Control system tightness
!
2.3.1 Inconsistencies across research !
According to van der Stede (2001) there is no uniform definition of control system tightness and furthermore, it is not operationalized in the same way across studies. One definition is that of Merchant and Van der Stede (2012) who describe control tightness as the degree of certainty that personnel will behave as the company desires. Anthony & Govindarajan (1998) state that a tight management control system is one in which an employee’s performance is assessed predominantly on his/her ability to meet budgetary objectives throughout each reporting period. Put differently, in Anthony & Govindarajan’s (1998) case, control tightness is dependent on the properties of the budgeting process, like the amount of emphasis on attaining the target and the extent of detail in budget reviews. Merchant (1985b), nonetheless, recognizes that control tightness can have great limits to the amount of freedom of the employee being controlled. However, he argues that it isn’t necessarily the case that detailed monthly budgets and frequent performance reviews are involved (as Anthony and Govindarajan (1998) describe it)).
Furthermore, there are some authors like Anthony and Govindarajan (1998) that elaborate on the components of control tightness and the means by which the latter can be achieved (Merchant, 1985b). On the other hand, authors like Merchant (1985b, 1990) deduce the tightness of a management control system from its
consequences. Relating to the consequences of tight control, Merchant (1985a) for example, regards a control system as more tight if it has a more significant effect on decision-making or if it is able to give a substantial amount of certainty that
employees act in the best interest of the organization (Merchant, 1998). Concluding, it is thus clear that the notion of control tightness is not
consistently used and operationalized across different studies. Lack thereof makes it difficult to interpret and replicate research findings (Fisher, 1995). This fact has troubled theorist for a lot of years and thus identifying what this definition entails has been called a successful area for future academic research. Therefore, the objective of the study conducted by Van der Stede (2001) is to clarify the notion of tight
relating to management accounting. !
2.3.2 The construct of tight (budgetary) control and its attributes as described by Van der Stede (2011)
!
Van der stede (2001) starts by defining the broad notion of control tightness and afterwards he narrows it down to tight budgetary control as is done by Anthony & Govindarajan (1998). Although Van der Stede (2001) acknowledges the fact that tight budgetary control is not the only way that tight control can be exercised, he mentions that accounting-based budgetary controls are a vital part of the overall management control system in the majority of for-profit organizations. Furthermore he mentions that these “formal” controls have been substantially studied in the field of management accounting. Therefore, Van der Stede (2001) argues that the tight budgetary control construct is crucial for future studies in the field of management accounting. As stated by Van der Stede (2001), the literature on budgetary control tightness and whether it is bad or good for organizational and/or individual
performance is not conclusive. In addition, issues relating to the whether it
encourages or discourages certain employee behavior and the effects it has on job tension are also unclear form prior literature (Van der Stede, 2001).
In order to describe the general notion of tight control, the author takes the view of Merchant (1985b), who states that a management control system that is tight provides a high level of certainty that individuals act in the way the organization deems appropriate. Merchant (1985b) states that management control systems can be tightened in four distinct ways. First, tightness can be increased by defining goals in such a way that they are more specific and complete, and/or by making the goals more congruent with objectives of the organization. Second, communicating more often, convincingly and effectively in the hope that employees will understand and accept the objectives of the organization better. Third, more detailed, frequent and prompt monitoring of the results and actions of the employees. Lastly, tightening the control system can be achieved by increasing the worth of rewards, which is given to the employees that are evaluated.
The above-mentioned ways in which controls could be tightened indicate that overall control tightness might exist in the presence or absence of tight budgetary
control, which mainly deals with constant, detailed and timely monitoring of
accounting performance measures (as pointed out in the above paragraph). Anthony & Govindarajan (1998), though, take a narrower view as denoted by tight budgetary control. Therefore, according to Anthony & Govindarajan (1998), a tight management control system is one in which an employee’s performance is assessed
predominantly on his/her ability to meet budgetary objectives throughout each reporting period.
In turn, after having described what is understood by a tight control system and how control tightness can be increased, Van der Stede (2001) identifies the following fives features of tight control. The first attribute is the emphasis placed on meeting budget targets. A second attribute is the extent of budget commitment (i.e. If revisions of the budget are allowed throughout the year). Van der Stede (2001, p. 122) states “amount of detail of interim budget reviews” is a third attribute.
Furthermore, an attribute of tight budgetary control relates to the amount of tolerance for deviations of interim budgets. The fifth attribute as stated by Van der Stede
(2001) relates to the extent of top management’s involvement in the businesses of the subordinates.
! !
2.3.3 Control tightness in the present study !
Not only is it important to decide which controls an organization should use, but also the tightness of management control systems is a crucial management decision (Merchant & Van der Stede, 2012) and, as mentioned in the above-paragraph, despite its importance it has not received the proper attention it deserves in the literature. The few times that it has been researched, the focus was predominantly on results control. Though the idea of tightness can be applied to controls in the results-context, there are several other ways in which tight management control can be achieved. This can be done by using other forms of control and with reinforcing compositions of control types (Merchant & Van der Stede, 2012). Theoretically, for tight control to be implemented effectively, it is required that management has a good understanding of how one or more of the controls (behavior (action), results, personnel & cultural)) relate and contribute to the main objectives of the firm.
In the present study control tightness is defined as the amount of flexibility in the management control system and tight control can be achieved in two ways. On the one hand, this can be achieved by putting in place more controls and having more rules and procedures to which the professionals must adhere (explicit tightness). On the other hand, it can be achieved by decreasing the amount of tolerance for divergence form the control system (implicit tightness).
It is argued that management control systems (MCSs) that are tighter are supposed to give more assurance that personnel will act in the best interest of the firm
(Merchant & Van der Stede, 2012). Control tightness, however, has its downside and this is especially relevant for this research that examines professionals (Raelin, 1985). Van der Stede (2001), for example, recognizes that control tightness can have great limits to the amount of freedom of the employee being controlled. Not granting individuals enough autonomy can have negative effects on individuals and in particular professionals who tend to resist formal forms controls (Raelin, 1985). The latter will be elaborated upon in the paragraph “professionals & the notion of task autonomy”. In addition, Ouchi (1979) mentions that tight formal controls will most probably offend professionals’ sense of autonomy and self-control, resulting in an unenthusiastic response that is purely compliant. Professionals require even closer supervision in the latter state, because of the fact that they have been alienated from the company, as a consequence of its tight control mechanisms (Ouchi, 1979). Generally, if the measurement becomes more obvious and explicit, it will also be more harmful to its employees. The latter results in higher costs to the firms that use such methods. Furthermore, Van der Stede (2001) mentions that certain (tight) control alternatives may not always be suitable and in some cases they can also have potentially detrimental side effects. The latter stems from contingency theory and will be discussed in more detail in the following paragraph.
!
2.4 Contingency theory
!
According to contingency theory the use and design of management control systems should be contextually dependent on the organizational setting. In other words, there is no universal management control system that is suitable for all organizations. Rather, contingent factors such as strategy should be taken into consideration when designing management control systems (Fisher, 1998). Contingency theory actually
emerged to respond to the universalistic view, which states that there is a one-size-fits-all formula to be used with respect to the design and use of management control systems (Fisher, 1998). The universalistic view is an extreme one and at the other end of the spectrum lies the situation-specific view, which states that the
circumstances that affect each management control system are distinctive and therefore widespread rules and standard models are not applicable (Fisher, 1998). Somewhere in between these two extremes lies the contingency theory and it argues that the suitability of the several control systems depends on the setting of the
organization. Nevertheless, contrasting the situation-specific, management control systems can be generalized for main classes of business settings (Fisher, 1998). Of particular interest to this study and relating to contingency theory is the paper by Abernathy & Brownell (1997). The authors focus on the suitability of controls in particular settings (management control systems in a research and development setting). Specifically, they examine the role of accounting and non-accounting controls and how task characteristics influence the appropriateness of the several forms of controls (Perrow, 1970). Their main finding is that
non-accounting controls, specifically personnel controls, contribute to the effectiveness of the organization (in particular in the case where task characteristics are not suitable to the use of accounting controls) (Abernathy & Brownell, 1997). Furthermore, the results indicate that “programmed” forms of control (accounting and/or behavior controls) are not appropriate where the number of exceptions in the task being performed is high.
!
2.5 Professionals & the notion of task autonomy
!
Raelin (1985) examines professionals, the education that they have followed, and the transition from graduates to employees working in organizations. Moreover, he reveals the cause of the difficulty that these professionals have when integrating into organizations. These so called “organizational professionals” think of themselves as employees that have a special status. As Raelin (1985) states, this special status results in a demand for autonomy and other particular rights that are usually awarded to these organizational professionals. These particular rights entail, discretion over numerous task of the professional’s work, the freedom to call into question management with regards to decisions that affect the professional’s work
and the dependence on professional norms of evaluation (Raelin, 1985). These rights probably depend on the professional’s great level of expertise, which is considered crucial to the operation of the organization (Raelin, 1985). However, according to Raelin (1985), organizations employing these professionals have been increasingly unwilling to grant these special status employees special rights that as a matter of fact conflict with the objectives and policies of management. Accordingly, there is an unavoidable struggle between management and the organizational professionals. The latter is the focus of Raelin’s (1985) study and put more
specifically, he concludes (after having investigated the several characteristics of the education and early socialization experiences of the professionals) that the attributes of professional life make professionals oppose managerial control.
In order to further elaborate on the notion of task autonomy and its relevance for this study, the paper by Langfred & Moye (2004) (who study the effects of task autonomy on performance of individuals) will be drawn upon. According to the authors, task autonomy is described as giving a person who performs certain tasks substantial freedom and control in deciding how these tasks should be carried out (translated to this study, the latter can be interpreted as loose behavior control). They further state that the concept of task autonomy is just as old as the
organization itself. However, the widespread use of the latter concept in
organizations is a fairly new phenomenon. The authors mention that previous studies have found empirical support for the (positive) association between task autonomy and performance, but that the magnitude of the association is very small. In addition, the positive effects resulting from task autonomy as proposed by existing theoretical models have several drawbacks (Langfred & Moye, 2004). According to Wall, Kemp, Jackson & Clegg (1986), very few of the research findings allow for causal inference and furthermore, the time-span of the research is often limited to a few months. However, when considering granting employees task autonomy, the intention is to accomplish enduring change in the attitudes and behaviors of the employees.
Therefore a time-span of years (and not months) is needed. Previous studies on the topic, in the majority of the cases, only support short-term effects and thus can be transitory (Wall et al., 1986). The reason for the latter is not particularly surprising, due to the fact that conducting research of this nature is very costly and not to mention time consuming (Wall et al., 1896). Langfred & Moye (2004) conclude that there is not one theoretical model that practitioners and/or researchers can use to
establish and understand the effects of according task autonomy to individuals in organizations. For the latter reason the authors are interested in factors that explain why and in what manner task autonomy affects the performance of individuals working in an organization. The authors use a mediational model and hypothesize and find that task autonomy (independent variable) influences the mediator variable (motivational, informational and structural mechanisms), which in turn influence individual performance (dependent variable).
In particular, what the authors mention and that is of interest for this study is the notion of task variability. Langfred & Moye (2004) argue that individuals that are performing tasks, which are highly variable, need to be able to amend their actions in order to successfully finalize those tasks. As the authors explain, the latter can be done appropriately in case the individual is granted task autonomy, because he/she has discretion over the realization of tasks. On the other hand, when individuals are not granted task autonomy, they either cannot finalize tasks that are highly variable or they need to wait for new instructions from their superiors before being able to proceed. In other words, cases where tasks are characterized by high variability, in combination with low autonomy can be described as extremely inefficient at best and dysfunctional at worst (Langfred & Moye, 2004). Consequently, when task variability is high, there is a need for autonomy to a certain extent in order to successfully carry out the task. In the case of low task variability and when tasks are highly formalized however (especially the latter), granting task autonomy might be problematic in the sense that it limits the level to which the employee can factually exercise discretion when performing a task. As Raelin (1985) and Bailyn (1985) demonstrate, the more a certain task is formalized, the less discretion is probable in its implementation (Langred & Moye, 2004).
2.6 Role conflict, role ambiguity and professional tension/stress
!
Rizzo, House & Lirtzman (1970) describe role conflict and role ambiguity (both part of role theory) in complex organizations. The authors mention that according to the chain-of-command principle, it is deemed more satisfying to organizational members if there is a single flow of authority from the top to the bottom. In theory, the latter provides top management with effective control and furthermore it is consistent with the unity of command principle. According to the latter principle an employee should
receive orders from only one superior and the reasoning behind this, is that the organizational structure should prevent a member from being caught in the middle of conflicting orders and or expectations from more than one superior. According to role theory, in the presence of role conflict an employee will experience stress, become dissatisfied and as a consequence perform less well than in the case that the expectations imposed on him did not conflict (Rizzo et al., 1970). In addition, role theory also states that the lack of required information available to a certain job position (role ambiguity) is positively correlated to the probability that an individual will be dissatisfied with his role, face anxiety and also as a consequence the
individual will perform less effectively. Rizzo et al. (1970) mention that professional (service) organizations, which are at the core of this study, often violate the chain-of-command principle, due to the existence of multiple-authority (based on position power at the one hand and professional expertise on the other hand, which is supported by collegial authority). Etzioni (1959) and Evan (1965) find that
professionals often experience stress, because they find themselves caught in the middle.
Aranya & Ferris (1984), who study the organizational-professional conflict (henceforth OPC), state that this conflict stems partly from the use of management control systems. The behavior of the professional is probably dictated by a code of ethics that is monitored externally by a collegial peer group. However, their behavior can also be controlled by orders from the organization that employs the
professionals. If the two are consistent, problems should not arise. However, in the case that organizational and professional goals are not compatible (which is a traditionally held belief, because of the presumed incompatibility of the
organizational-bureaucratic and professional value system), the OPC is likely to occur. The authors study the nature of the OPC and the outcome on organizational professionals. With respect to the latter (outcome), Aranya & Ferris (1984)
hypothesize and find that there is a negative relation between OPC and job
satisfaction of accountants. In addition they also hypothesize and find that OPC is positively associated with turnover intentions of accountants. In the current study, professional tension/stress is defined as the extent to which the professional feels s/he cannot perform the tasks in the way s/he thinks it should be done and
resembles the concepts of role conflict and role ambiguity discussed by Rizzo et al. (1970) and Aranya & Ferris (1984).
3. Hypotheses development
Based on prior literature it is safe to say that professionals demand a certain amount of autonomy (Raelin, 1985; Langred & Moye, 2004). Professionals can be regarded as “special” employees who have received sufficient training to conduct complex tasks in an independent manner (Abernathy & Stoelwinder, 1995). Furthermore, they are able to resolve difficulties arising when conducting the complex tasks by using their experience and expertise (Derber & Schwartz, 1991). Therefore, constant supervision/monitoring and adherence to rules is perceived as irritating by professionals and offends their sense of autonomy (Ouchi, 1979). In addition scholars such as Ouchi (1979) and Thompson (1967) mention that the suitability of behavior controls is contingent on the capability of management to determine the cause/effect relations of the tasks performed (Thompson, 1967). Specifically, in the case of professional service firms, tasks are often non-routine and completion depends on other team members. Behavior controls are usually particular rules that are very clear and specifically formulated to make sure that tasks will be completed. Therefore it is argued that they are not appropriate in professional service firms where tasks are relatively complex (Zucker, 1991).
Another form of formal control is results controls and its suitability is contingent on management’s capability to determine the desired results, the capability of the professionals to influence the desired results and the capability of management to measure the desired results in an effective manner (Merchant & Van der Stede, 2012). In the case of professional service firms, the use of results control can be problematic because the successful completion of a task is contingent on other team members.
As described earlier, Langfred & Moye (2004) provide evidence on the positive effects of task autonomy on individual performance. Translated to the current setting, it can be said that loose behavior control is beneficial with respect to the individual performance of employees. Furthermore, Davila (2000) argues that by implementing a lot of rules (thereby constraining the behavior of engineers), the amount of creativity needed by the experts will be reduced. This will have a negative effect on performance (Amabile, 1998). In addition, Jaworski et al. (1993) examine how the characteristics of strategic business units (SBUs) and task complexity determine the appropriate controls to be implemented. In turn they also investigate
what the implications of these controls are on SBU managers. They argue that job performance of the SBU managers is higher in the presence of clan control systems (high reliance on “informal forms” of control such as personnel and cultural controls and a low reliance on “formal” forms of controls such as behavior and results
control).
This leads to the first two hypotheses being formally tested:
H1: There is negative relationship between results control tightness and individual performance.
H2: There is negative relationship between behavior control tightness and individual performance.
Hirst (1983) examines the effects of accounting performance measures (APMs) (results controls) (in evaluating employee performance) on employee behavior. Particularly, Hirst (1983) investigates what this reliance on APMs has on employee tension and social withdrawal (kind of dysfunctional behavior which is a sort of resistance like Raelin (1985) describes it) and does this in high as well as low task uncertainty situations. It should be noted that task uncertainty, as Hirst (1983) describes it, has some resemblance with the notion of task complexity as described by Jaworski et al. (1993) (and what is adopted in this study). Results indicate that there is a positive association between reliance on accounting performance
measures and tension (in the presence of task uncertainty). This is relevant for this study as task complexity is one of the two features that characterize a professional service firm (in addition to the three features that are mentioned by Von
Nordenflychy (2010)). Furthermore, in the particular setting of high task uncertainty there was increase in tension reported as reliance on accounting performance measures (results controls) increased (Hirst, 1983). Jamal (1985) provides evidence that an increase in job-related tension will negatively affect individual performance.
This leads to next hypothesis being formally tested:
H3: The relationship between results control tightness and individual performance is more negative in the case of high task complexity situations.
In professional service firms, where situations are characterized by highly complex tasks, the professional might experience feelings of loss of control (Shields et al., 2000). According to Davila (2000), only in cases where management control systems are able to provide useful information for coordination and learning in these
situations, it might be the case that the use of behavior controls is beneficial to steer the employees in the right direction (Davila, 2000). However, as Langfred & Moye (2004) point out, when tasks are very complex professionals can perform at there best when they are given a certain amount of freedom. This is required so that they can amend their actions in order to successfully finalize the tasks. In the situation where a professional is not granted task autonomy (which can be translated to tight behavior control for this study), they either cannot finalize tasks that are very
complex or they need to wait for new instructions from their superiors before being able to proceed. The latter situation can lead to professional tension and thus negatively affect individual performance (Langfred & Moye, 2004; Jamal, 1985)
This leads to the following hypothesis being formally tested:
H4: The relationship between behavior control tightness and individual performance is more negative in the case of high task complexity situations.
Shields et al. (2000) examine the design and effects of control systems on engineers in a Japanese automotive industry. The authors do this by developing a direct and an indirect model. The former posits that management control systems directly influence the performance of the professionals, while the latter posits that
consequences of management control systems are indirect through the mediating variable job-related stress. In this study we further theorize from the indirect model and following this model it is presumed that professional tension is a consequence of
the use of management controls (Shields et al., 2000). In the presence of tight results controls, professionals might encounter feelings of role ambiguity and/or loss of control, due to the fact that they encounter uncertainty with respect to the outcome of their job and/or how their input affects job outcome. In professional service firms this is likely to be seen, as the successful completion of a task depends on several other members (Abernathy & Stoelwinder, 1995). As mentioned before,
professionals enjoy their autonomy and in the case of tight behavior control, professionals will tend to resist managerial control (Raelin, 1985). In particular, Abernathy & Stoelwinder (1995) mention that professionals will put in great amounts of effort to avoid these bureaucratic controls. Furthermore, the authors state that this will have adverse effects (e.g job-dissatisfaction and job-related stress) on the
professionals employed by the organization (Copurl, 1990). Consequently it might be the case that professionals will exert less effort and hence individual performance decreases (Shields et al., 2000). As mentioned earlier and in order to further substantiate that professional tension can have adverse effects on individual
performance of employees, Jamal’s (1985) (who examines the relationship between job stress and job performance for managers and blue-collar workers in a Canadian organization) results generally support the negative relationship between job stress (which comprises role ambiguity, role overload, role conflict and resource
inadequacy) and job performance.
This leads to last hypotheses being formally tested:
H5: The relationship between results control tightness and individual performance is mediated by professional tension.
H6: The relationship between behavior control tightness and individual performance is mediated by professional tension.
4. Data & Research methodology
4.1 Survey
!
In order to gather empirical data a survey was conducted. This survey was part of the PSF thesis survey project 2015/2016 of the Faculty of Economics & Business (FEB) of the University of Amsterdam (UvA). After the survey instrument was
constructed, two distinct pre-tests were carried out. The goal of the first pre-test was to determine the quality of the survey. In particular, the objective was to evaluate the quality of the items that were used to measure the variables relating to the several types of controls. These controls are explicit and implicit - behavior, results,
personnel and cultural controls (thus totaling eight control constructs). In order to accomplish this, an item-sort task was used. Throughout the scale-creation process, item-sort task is a frequently used procedure that is used with the objective of
eliminating over-representative item lists (Howard & Melloy, 2016). Each participant was given two pieces of paper in order to carry out the task at hand. On the one piece of paper the eight control constructs were stated and on the other there were fifty-two statements that were designed to test the respective constructs. Participants had to match each statement to the construct they best thought it belonged to.
In total twenty professionals were asked to take part in this pre-test and thereof fourteen completed the task at hand. These fourteen professionals
comprised 1 graphic designer, 4 lawyers, 1 accountant, 2 management consultants, 1 security consultant, 1 IT consultant, 1 psychologist, 1 architect, 1 marketing
professional, 1 dentist. After completion of the task the amount of accurate and inaccurate matches were accounted for. Afterwards, for every type of control the four statements that had the least inaccurate matches were chosen and included in the survey. The thirty-two statements that were chosen for inclusion in the survey ranged from a high of six inaccurate matches (implicit behavior control) to a low of no
inaccurate matches (explicit behavior control).
The objective of the second pre-test was to determine the overall quality of the survey and for that task a further twenty participants had been asked to
participate. The participants had to take a look at the survey online and answer several questions with respect to the clarity, content and appearance of the survey in addition to the total time needed to finish the survey (2 marketing professionals, 3
medical doctors, 1 advertising executive, 1 dentist, 3 chemists, 3 lawyers, 2 management consultants, 1 architect, 1 graphic designer, 1 artist, 1 electrical engineer, 1 mining engineer). Thirteen participants gave written answers and the other ones gave the answers by telephone. This led to some minor changes in wording and the inclusion of extra answers pertaining to some multiple-choice questions.
The PSF thesis survey project 2015/2016 was a research project
administered by a PhD student at the Amsterdam Business School (ABS). Students could participate in this project to jointly collect a meaningful amount of surveys. In order to be granted access to the ultimate dataset, a minimum of 7 responses of suitable professionals had to be collected. As a consequence of this collective effort a total amount of 372 surveys were gathered.
4.2 Respondents
!
The majority of the respondents were based in The Netherlands. Some of the
respondents, however, were also employed in professional service firms outside The Netherlands. Respondents considered for this study were mid-level employees of professional service firms, the latter being of medium to large size. Professional service firms include, among others, the following occupational fields: accounting, investment banking, risk management services, law and advertising. The sample population excludes occupations in non-profit, publicly owned organizations
(universities), government organizations, non-governmental organizations and social work agencies. However, it does include medical doctors that are employed at public or not-for-profit hospitals. Specifically, qualified respondents had to meet some criteria. First of all, the respondent has more than 3 years of work experience in the field (and probably less than 10 years of experience). This is an important criterion that needs to be fulfilled, because management control systems for professionals that are still in the learning phase are quite different. Secondly, the respondent is not an owner/partner or board member of the organization. This criterion is of essence, because employees need to be subject to the management accounting and control system rather than design it. Third, the respondent works in a medium to large size firm that employs more than 50 employees. Lastly, the respondent must speak and
understand English at a business level, in order to successfully fill-out the questionnaire.
Out of the 372 surveys gathered, 320 were useful due to the fact that the other respondents did not fill out all of the answers relating to the constructs being used. In order to get a better idea of the sample being analyzed, demographics (data about the characteristics of the participants) are presented in table 4.1. Of the
respondents, the average age is 34.46 years, the youngest respondent being 19 years and the oldest 63 years. The average work experience in the (current) field was about 7 years (6.94), while the average job tenure was approximately 6 years (5.78). About two thirds (61%) of the participants were male (39% female). The participants had a Bachelor degree or lower (36.9%), Masters degree (53.1%) or PhD or other professional doctorate degree (10%). Professionals from over thirty fields participated in this research; accountancy being the most represented one (23.7%). Lastly, the vast majority was Dutch (69.2%).
4.3 Constructs
The survey employed for the present research was a pre-developed survey, designed by the PhD student. It was designed in a very broad manner in order to
Table&4.1&Characteristics&of&the&respondents Age Mean%(St.%Deviation)/Min/Max 34.46%(7.81)/19/63 Work(Experience Mean%(St.%Deviation)/Min/Max 6.94%(3.07)1/11 Job(Tenure Mean%(St.%Deviation)/Min/Max 5.78%(3.14)/1/11 Gender %%Male 61 %%Female 39 Educational(Level %%Bachelor%degree%or%lower 36.9 %%Masters%Degree 53.1 %PhD%or%other%professional%doctorate%degree%(i.e.%J.D./M.D./D.D.S) 10 Occupation %%Accounting 23.7 %%Consulting%Management/Strategic 9.5 %%Medicine/Physician%practices 7.5 %%Consulting%IT 5.6 %%Marketing/Public%relations 4.2 %%Law 3.6 %%Financial%Advising 3.1 %%Other 42.8 Nationality %%Dutch 69.2 %%Other 30.8
investigate several aspects related to the topic of management control systems in professional service firms. Therefore, the survey consisted of quite a lot of
constructs. For this research, six constructs were used (in addition to a few control variables) in order to (try and) answer the research question. For the dependent variable, individual performance – in role & individual performance – general were used as constructs and for the mediating variable (in addition to being a control variable) professional tension was used. For the moderating variable, task complexity was used. Results and behavior control tightness were used as
constructs for the independent variables. The purpose of this survey was two folded. First of all, it looked at how management control systems in professional service firms needed to be designed. Secondly, it was being examined what the
consequences (with respect to professional tension and individual performance) were for the professionals working in these firms. For this particular study, the emphasis was on the latter and the items used in order to measure the constructs can be found in the appendix. For all the constructs (except for individual
performance – general) used in this study, participants had to rate the statements on a five-point likert scale with the following interpretation: (1) strongly disagree, (2) disagree, (3) neutral, (4) agree, (5) strongly agree. For the construct individual performance – general, participants had to rate the statements on a five-point likert scale as well. However, the interpretation was as follows: (1) needs much
improvement, (2) needs some improvement, (3) satisfactory, (4) good, (5) excellent. For all the relevant constructs a varimax principal components analysis (PCA) was executed. First of all, a PCA was conducted per construct (not tabulated).
Secondly, a joint factor analysis was conducted (table 4.2) where the factor loadings under .50 were suppressed. In order to determine the reliability of the constructs, Cronbach’s alpha was computed for each construct. In the next paragraphs the specific constructs are further elaborated upon in addition to the results of both the validity and reliability analysis.
Table&4.2&Joint&Factor&Analysis&of&all&latent&variables&(factor&loadings&<.50&suppressed) Components Construct Items 1 2 3 4 5 6 7 8 9 10 Results8Control8Tightness Q5_18In8my8job,8there8is8a8performance8measure8for8everything .719 Q5_48My8organization8sets8a8large8number8of8performance8goals/targets8that8I8am8expected8to8meet .682 Q5_68Employee8attainment8of8goals/targets8is8checked8constantly .844 Q5_78My8supervisor8frequently8checks8to8make8sure8that8I8am8meeting8my8performance8targets. .832 Behavior8Control8Tightness Q4_188Whatever8situation8arises,8we8have8existing8processes,8procedures8or88rules8to8follow8in8dealing8with8it .748 Q4_28Established8processes,8procedures8and8rules8cover8all8of8my8job8tasks.8 .701 Q4_48In8my8organization,8we8have8rules8for8everything. .702 Q4_88My8supervisor8frequently8monitors8the8extent8to8which8I8follow8established8process,8procedures8and8 rules Professional8Tension/Stress Q33_18I8do8not8have8enough8time8to8complete8my8work8the8way8that8I8think8it8should8be8done Q33_28Due8to8a8lack8of8adequate8resources8and8materials,8I8cannot8execute8my8assignments8properly. .664 Q33_38The8type8and8structure8of8my8employment8gives8me8the8opportunity8to8fully8express8myself8as8a8 professional. Q33_48In8my8organization,8there8is8a8conflict8between8the8work8standards8and8procedures8of8the8organization8 and8my8own8ability8to8act8according8to8my8professional8judgment. .769 Q33_58My8organization8hinders8me8from8doing8my8work8properly .797 Q33_68I8have8to8alter8my8professional8behavior8in8order8to8perform8my8job8the8way8my8organization8wants8 me8to .580 Q33_78In8this8organization,8I8can’t8perform8my8job8the8way8that8I8think8I8should .754 Q33_88I8could8do8my8job8much8better8without8the8conditions8imposed8by8my8organization. .787 Task8Complexity Q12_18I8can8easily8determine8whether8I8have8performed8my8work88correctly .712 Q12_28Most8of8the8time,8I8know8what8to8outcome8of8my8work8efforts8will8be .608 Q12_38I8often8encounter8problems8in8my8work8for8which8there8are8no88immediate8or8apparent8solutions .800 Q12_48In8my8work,8I8spend8a8lot8of8time8solving8difficult8problems8with8no8immediate8solutions. .830 Q12_58The8situations,8problems8and8issues8that8I8encounter8in8performing8my88major8tasks8are8usually8the8 same. .708 Q12_68I8would8describe8my8work8as8routine. .712 Q12_78My8job8depends8on8the8work8of8many8different8people8for8its8completion. Individual8Performance8In8Role Q35_18This8employee8always8performs8all8tasks8that8are8expected8of8him/her .823 Q35_28This8employee8always8performs8all8essential8duties .848 Q35_38This8employee8always8fulfills8all8responsibilities8required8by8his/her88job .837 Q35_48This8employee8always8engages8in8all8activities8that8will8directly8affect8his/her8performance8evaluation .710 Q35_58This8employee8always8meets8all8formal8performance8requirements8of8the8job .836 Q35_68This8employee8always8completes8all8duties8specified8in8his/her8job8description .818 Q35_78This8employee8never8neglects8aspects8of8the8job8that8he/she8is8obligated8to8perform .708 Individual8Perofrmance8General Q36_18Quantity8of8work8output Q36_28Quality8of8work8output .539 Q36_38Accuracy8of8work .699 Q36_48Customer8service8provided8(internal8and8external) Q36_58Obtaining8personal8career8goals .784 Q36_68Developing8skills8needed8for8your8future8career .590 Q36_78Making8progress8in8your8future8career .770 Q36_88Seeking8out8career8opportunities .766 Q36_98Coming8up8with8new8ideas .776 Q36_108Working8to8implement8new8ideas .779 Q36_118Finding8improved8ways8to8do8things .784
4.3.1 Results control tightness !
The construct for results control tightness is adapted from Aiken & Hage (1968), Simons (1987), Van der Stede (2001) and Jaworski et al. (1993). In addition to the existing items pertaining to this construct, other items were added to the come about the results control tightness construct for this study. Items Q5_9, Q5_10 and Q5_11 were reverse coded into new variables to measure tightness (instead of looseness of the control system). Based on factor analysis, it can be concluded that the items loaded on 2 factors. This is understandable since the concept of results control tightness can be divided into items that measure implicit tightness (achieved by decreasing the amount of tolerance for divergence form the control system) and explicit tightness (achieved by putting in place more controls and having more rules and procedures to which the professionals must adhere). The first four items (Q5_1, Q5_4, Q5_6 and Q5_7) measured explicit tightness, while the last four (Q5_9, Q5_10, Q5_11 and Q5_12) measured implicit tightness. The items were separated and factor analysis was once again performed. The items loaded on two components for implicit tightness, while for the items of explicit tightness, however, the items loaded on one component and therefore the latter items are taken into consideration for the joint factor analysis (JFA). Based on JFA, all the items loaded on one single component with Chronbach’s Alpha of .82.
!
4.3.2 Behavior control tightness !
The construct for behavior control tightness is adapted from the authors Aiken & Hage (1968), Van den Ven & Ferry (1980), Cunningham and Rivera (2001), Van der Stede (2001) and Bodewes (2000). First of all, the items Q4_10, Q4_11, Q4_12 and Q4_13 were reverse coded. Similar to the construct for results control tightness, behavior control tightness loads on two components for the same reason. The first four items (Q4_1, Q4_2, Q4_4 and Q4_8) measured explicit tightness, while the last four (Q4_10, Q4_11, Q4_12 and Q4_13) measured implicit tightness. The items were divided and factor analysis was once again conducted. The items loaded on two components for implicit tightness, while the items loaded on one component for explicit tightness. Therefore the items measuring explicit tightness were taken into
consideration for the JFA and based on these results all items loaded on a single component except for item Q4_8 with Chronbach’s Alpha of .76.
4.3.3 Professional tension/Stress !
In order to measure the concept of professional tension/stress, the constructs
measuring role conflict and role ambiguity by Rizzo et al. (1970) and Aranya & Ferris (1984) are drawn upon. In addition to the items pertaining to these constructs, other items were added in order to create the construct professional tension/stress as defined in the current study. Item Q33_3 was reverse coded due to the fact that it measured the opposite of “professional tension/stress”. Having performed a PCA on the construct, all items load on a single component. When performing a JFA on this construct, items Q33_1 and Q33_3 did not load. However, these items were not deleted because it would not lead to an (significant) improvement of Chronbach’s Alpha, which was .83 for all eight items.
!
4.3.4 Task complexity !
The construct for task complexity is taken as is from Van den Ven & Ferry (1980). The first step was to reverse code four items (Q12_1, Q12_2, Q12_5 and Q12_6). When conducting a PCA on this construct, it seems that it loads on two different components. Items Q12_1, Q12_2, Q12_5 and Q12_6 loaded on one component, while items Q12_3, Q12_4 and Q12_7 loaded one another component. Based on JFA, it became apparent that the items loaded on three components. This seems plausible because the item: I can easily determine whether I have performed my work correctly (Q12_1) and the item: Most of the time, I know what to outcome of my work efforts will be (Q12_2), measure predictability, which is an aspect of task
complexity (as used in this study). The Item: I often encounter problems in my work for which there are no immediate or apparent solutions (Q12_3) and the item: In my work, I spend a lot of time solving difficult problems with no immediate solutions (Q12_4) focus on the variety in task. The item: the situations, problems and issues that I encounter in performing my major tasks are usually the same (Q12_5) and the item: I would describe my work as routine (Q12_6), measure routineness. Due to the fact that this construct was designed to take into consideration different (but related) aspects of task complexity, all these items were taken into consideration for the
construct. The item: My job depends on the work of many different people for its completion (Q12_7) did not load. Furthermore, when excluding this item,
Chronbach’s Alpha improved by .03 from .59 to .62. Thus, item Q12_7 was deleted. !
4.3.5 Individual performance – In role !
The construct for individual performance – in role is adapted from Podsakoff & Mackenzie (1989) and Welbourne, Johnson & Erez (1998). All items loaded on one component after having conducted a PCA on this construct as well as a JFA.
Chronbach’s Alpha for the seven items was .91 and no items were deleted, as they would not increase the reliability of the construct.
!
4.3.6 Individual performance – General !
The construct used for individual performance – general (just like the previous construct) is adapted from existing literature by Podsakoff & Mackenzie (1989) and Welbourne et al. (1998). Based on the results of the PCA conducted on this
construct, it seems to be the case that the items load on three components. In addition, item: Customer service provided (internal and external) (Q36_4), did not load. Furthermore, when conducting a joint factor analysis (JFA), it loads on three components as well. Item: Q36_4 was deleted and Chronbach’s alpha was .89. When taking a closer look at the items measuring this construct, it seems reasonable that the items load on three components. As the name of the construct already
points out, it is a general measure of individual performance that takes into consideration different aspects of performance. The item: Quantity of output
(Q36_1), the item: Quality of output (Q36_2) and the item: Accuracy of work (Q36_3) are very direct statements addressing “hard performance” measures. The item: Obtaining personal career goals (Q36_5), item: developing skills needed for your future career (Q36_6), item: Making progress in your future career (Q36_7), item: Seeking out career opportunities (Q36_8) address aspects of personal improvement. Lastly, item: Coming up with new ideas (Q36_9), item: Working to implement new ideas (Q36_10), item: Finding improved ways to do things (Q36_11), item: Creating better processes and routines (Q36_12) focus on innovative performance. In order to determine whether the results differ when conducting analyses, 3 separate
constructs were created based on the three components and Chronbach’s Alpha was .85, .84, and .83 (not tabulated) respectively for Individual performance – Hard, Individual performance – Self-improvement and Individual performance –
Innovativeness.
4.3.7 Control variables !
In order to control for other determinants of individual performance, professional tension/stress, job tenure and educational level are added to the regression model(s) (discussed in the next section) as control variables. Professional tension/stress is thus used as both mediating and control variable. As Jamal (1985) points out, job-related stress is negatively associated with individual job performance. Due to the fact that the latter construct is already extensively explained in one of the previous paragraphs, it is not discussed again as one of the control variables.
4.3.7.1 Job tenure !
In order to measure job tenure, the participants were asked how many years they had been working in their organization. It could be argued that individuals that have been employed by their organization for a longer time (in comparison with those that have been employed by their organization for a relatively short period of time) have acquired the necessary skills, are better acquainted with the work environment and are more adaptable, which means that they can better react to uncertain situations (Caldwell & O’Reilly, 1982). The latter is in line with human capital theory.
Furthermore, McDaniel & Schmidt (1988) report a positive relationship between job experience and job performance. In other words, it is expected that job tenure is positively associated with individual performance.
4.3.7.2 Educational level !
The participants were asked what the highest level of education they had completed was, where 1 was Bachelor degree or lower, 2 Masters degree and 3 PhD or other professional doctorate degree (i.e. J.D./M.D./D.D.S.). This variable was recoded into a different variable (dummy), where 0 stood for Bachelor Degree or lower and 1 for
