The Impact of Management Control on Autonomous
Motivation and Performance:
The Use of Control and the Role of Job Types
Niels Löbach S4149491
N.loebach@student.rug.nl
Supervisor: Prof. dr. ir. P.M.G. van Veen-Dirks Word count: 12.888
22-06-2020
Master Thesis
MSc Business Administration Management Accounting and Control
2
ABSTRACT
3
CONTENT
I. INTRODUCTION ... 5
II. LITERATURE REVIEW ... 7
2.1 Self-Determination Theory ... 7
2.1.1 Autonomous and controlled motivation ... 7
2.1.2 Enhancing and undermining autonomous motivation ... 8
2.2 Management Control... 9
2.2.1 Management Control System ... 9
2.2.2 MCS as a package ... 11
2.2.3 MC and motivation hypotheses ... 11
2.3 The moderating effect of job types ... 13
2.3.1 Job Characteristic Theory ... 13
2.3.2 Hypothesis development ... 14
2.4 Motivation and performance ... 17
2.5 Conceptual model ... 17
III. METHODS ... 18
3.1 Research method and sample...18
3.2 Measures ...18
3.2.1 Independent and dependent variables ...18
3.2.2 Control variables ... 19
3.3 Data analysis ... 21
3.3.1 Exploratory Factor Analysis and Reliability Analysis ... 21
3.3.2 Hypothesis testing ... 21
IV. FINDINGS ... 23
4.1 Descriptive statistics ... 23
4.2 Early and late respondents... 24
4.3 Pearson correlation ... 25
4.4 Hypothesis testing ... 25
V. DISCUSSION AND CONCLUSION ... 27
REFERENCES ... 31
4
TABLES AND FIGURES
Figure 1. Self-Determination continuum ... 8
Figure 2. The job characteristic model ... 13
Figure 3. Conceptual model ... 17
Table 1. Construct descriptions ... 20
Table 2. Descriptive statistics sample ... 23
Table 3. Early and late respondents ... 24
Table 4. Pearson correlation matrix ... 25
Table 5. Results of multiple linear regression analysis and moderation analysis ... 26
APPENDICES Appendix A. Results of exploratory factor analysis for Levers of Control ... 36
Appendix B. Results of exploratory factor analysis for autonomous motivation ... 36
5
I. INTRODUCTION
The importance of autonomous motivation for individuals’ performance and overall well-being has been advocated by psychology for decades (e.g. Ryan & Deci, 2000; Koestner & Losier, 2002). In the last years management accounting researchers have begun to apply theories from the field of psychology in the organizational context. Particularly, the public sector has received increased attention (e.g. Van der Kolk et al., 2019) as employees’ motivation seem to differ from employees in private organizations. In addition, scholars have recognized the importance of employee motivation for organizational performance in universities (e.g. Sutton & Brown, 2016) and other higher educational institutions (HEIs) (e.g. Zlate & Cucui, 2015; Hanaysha & Majid, 2018; Kuchava & Buchashvili, 2016). Since the performance of both employees and managers directly influences students’ learning and their overall experience at the HEI, it is people that propel the organization’s success. This vital role of staff performance becomes even more evident when understanding that students are still the main source of income in HEIs (Kuchava & Buchashvili, 2016).
6
including autonomous motivation (e.g. Groen et al., 2017, Van der Kolk et al., 2019). Within this stream, several scholars have studied the effects of different control elements and MCSs (e.g. Sutton & Brown, 2016, Van der Kolk et al., 2019). However, the puzzle of how the MCS can promote autonomous motivation and thus, drive the organization’s performance in HEIs is still not yet solved. Many researchers studied Simon’s ‘levers-of-control’ framework (e.g. Granlund & Taipaleenmäki, 2005; Marginson, 2002; Mundy, 2010; Tuomela, 2005; Widener, 2007). Within this framework, there have been studies on the use of the MCS represented by diagnostic and interactive control systems (Bobe & Taylor, 2010; Henri, 2006). The framework presents two additional formal control elements - beliefs systems and boundary systems – that have been widely neglected (Tessier & Otley, 2012). All levers of control represent positive or negative forces that balance the amount of autonomy the MCS provides, which in turn could also impact autonomous motivation. The dual role of those opposing forces (i.e. positive and negative controls) in the MCS has received too little attention (Tessier & Otley, 2012). These opposing forces reflect trade-offs “between freedom and constraint, between empowerment and accountability, between top-down direction and bottom-up creativity, between experimentation and efficiency” (Simons, 1994, p.4). Beliefs systems and boundary systems also generate positive and negative forces and thus, could strengthen the effect of interactive and diagnostic controls by creating more leverage. The study of the joint use of controls that generate the same force might reveal new insights on combined effects of the same force and of the MCS. Therefore, this study will build on past research to examine the effect of the use of positive versus negative controls in the MCS on autonomous motivation and in turn on performance.
7
RQ: What is the impact of the use of control and the job type on autonomous motivation and in turn on performance?
The remainder of this thesis is structured as follows. First, in the literature review I will discuss past research from the field of psychology and management accounting. This thesis will draw on Self-Determination Theory and Job-Characteristic Theory to develop the hypotheses that are presented at the end of Chapter II. This section ends with the illustration of the developed hypotheses in the conceptual model. Subsequently, in Chapter III, the sample, the research design and the data analysis will be presented. Chapter IV is to summarize the findings from the data analyses, which are then discussed in the last chapter. In the fifth chapter, Discussion and Conclusion, the findings are interpreted and discussed as well as the implications and contributions to prior management accounting research as well as limitations of this thesis are presented.
II. LITERATURE REVIEW
2.1 Self-Determination Theory
Self-Determination Theory (SDT) is a theory in the field of psychology that is concerned with human motivation and behaviour. Central in SDT is to understand, what psychological factors cause different types of motivation and how social structures can facilitate or hinder these factors.
2.1.1 Autonomous and controlled motivation
People can be motivated in two different ways. Motivation can either derive from inner goals or is somehow regulated externally (Ryan & Deci, 2000). Autonomous motivation means “…engaging in a behaviour because it is perceived to be consistent with intrinsic goals or outcomes and derives from the self, i.e. the behaviour is self-determined” (Hagger et al., 2014) and can be both intrinsic and extrinsic. Intrinsic motivation as the most natural form of autonomous motivation happens when people act, because they are interested in an activity (Gagné & Deci, 2005, Ryan & Deci, 2000). Extrinsic motivation is more complex as it has both autonomous and controlled forms. If a task is not interesting, people can be externally incentivised or pressured to act, for instance through the use of rewards or punishments. In those cases, their motivation is externally controlled.
8
that describes the embodiment of values and external behavioural regulations and the subsequent transformation into personal values. This process leads to self-determined behaviour. Figure 1 illustrates SDT’s various forms of motivation, ranging from amotivation to external motivation to introjected motivation to identified motivation to intrinsic motivation. Internalization has different stages that reflect the degree of autonomous (versus controlled) motivation. Identification refers to the integration of external regulations and accepting them as their own personal values, i.e. it makes them part of who they are. In this, people have a greater feeling of autonomy as their behaviour is consistent with their personal values (Gagné & Deci, 2005). If this process is not fully successful because not all values are internalized the motivation is still moderately controlled and called introjection. Introjection involves a feeling of pressure that is caused by the regulation itself (Ryan, 1995). For instance, people perform a task to avoid shame or guilt, or because it makes them feel worthy (i.e. they involve their ego) (Deci & Ryan, 1985).
Figure 1. Self-Determination continuum. (Sheldon et al. 2003)
2.1.2 Enhancing and undermining autonomous motivation
9
Autonomy is crucial for autonomous motivation. Task choice can enhance the feeling of autonomy. Contrarily, tangible rewards, surveillance, performance evaluations, imposed goals, deadlines, threats and competitive pressure tend to reduce autonomy (Deci & Ryan, 1985; Ryan, 2000). This is because these factors can cause a shift in the perceived locus of causality (PLOC) from internal to external (Ryan & Connell, 1989), i.e. when people feel controlled, they stop being motivated by their interest, but rather by those external factors. This undermining effect is also known as the crowding-out effect. Crowding-in, in contrast, happens when external factors are perceived as supportive (or positively informative) and enhance intrinsic motivation. (Frey, 2012). Moreover, people need to feel competent in their behaviour to be autonomously motivated. Positive feedback and optimal challenging activities have shown to support this feeling of competence. (Ryan, 2000).
The third need, relatedness, is important for extrinsic autonomous motivation. As tasks are not interesting, people’s main reason for doing them, is that they mean something to people they feel (or would like to feel) connected to. Within this, competence also plays a major role. People tend to adopt activities that others value when they feel effective with regard to those activities. Most importantly, internalization can only happen, when people feel autonomous rather than controlled by rewards or punishment (Deci & Ryan, 2000). Consequentially, in order to enhance autonomous motivation those three needs must be facilitated.
2.2 Management Control
2.2.1 Management Control System
There is still no uniform definition for the term management control system. Some scholars focus narrowly on employee behaviour (e.g. Merchant & Van der Stede, 2007), others leave out the control aspect or the systematic use of management control elements (e.g. Chenhall, 2003). Malmi & Brown (2008) conclude that the most accurate definition is the following: “management controls include all the devices and systems managers use to ensure that the behaviours and decisions of their employees are consistent with the organization’s objectives and strategies[…]If these are complete systems[…]then they should be called MCSs.” (Malmi & Brown, 2008).
10
goal achievement need to be controlled as both have a great importance for the organization. For this purpose, the levers-of-control framework provides four levers - beliefs systems, boundary systems, diagnostic control systems and interactive control systems – that support innovation and creativity while constraining employee behaviour to ensure predictable goal achievement. These positive and negative forces allow managers to generate balance between creativity and control (Simons, 1995).
Beliefs systems and boundary systems are both formal control systems. Beliefs systems as positive controls are explicit sets of organizational definitions that communicate the organization’s core values and mission (Simons, 1994). The primary purpose of beliefs systems is to secure goal commitment and inspire and guide employees in their search for opportunities and solutions (Simons, 1994).
Boundary systems form the counterpart to beliefs systems and play a negative role in the MCS as they restrict the opportunity-seeking behaviour. This explicit set of organizational definitions is typically expressed in negative terms or minimum standards, such as through the code of conduct (Simons, 1995; Simons, 1994). Similar to beliefs systems, they are set to guide employees by communicating activities that are off-limits. However, they also serve as strategic boundaries for the search for innovative ideas and thereby prevent employees from wasting resources (Simons, 1994). Simons (1995) compares beliefs systems to a racing car, of which boundaries symbolize the brakes.
Diagnostic control systems represent the traditional use of the MCS as a feedback system that compares actual performance to pre-set targets. Like a cockpit of an airplane, managers use these information systems to monitor and analyse critical performance variables and correct for deviations. These systems ensure the control for predictable goal achievement. Diagnostic control systems are naturally negative as they restrict employees’ opportunity-search and experimentation to ensure the intended strategy is pursued (Simons, 1994, Simons, 1995).
11 2.2.2 MCS as a package
The notion of studying the MCS as a collective (or package), rather than each element in isolation, is a popular stream in the literature (see e.g. Widener, 2007; Malmi & Brown, 2008; Mundy, 2010). MC elements do not operate as stand-alone functions but interrelate with each other in a bigger system, the MCS (Malmi & Brown, 2008). Consequentially, if one element of the system is changed, the nature of the entire system is changed, too. Research provided strong evidence that systematic relationships between MC elements do exist in practice (e.g. Widener, 2007). Therefore, it is necessary to study the impact on autonomous motivation considering the MCS as a package.
2.2.3 MC and motivation hypotheses
Before hypothesizing the effects of the two forces on autonomous motivation, it is important to discuss two issues first. There still seems to be ambiguity in the literature what good and bad controls, enabling and coercive controls, and positive and negative controls are. Initially, controls are neutral and set to enable a human behaviour that is congruent with the organizations’ goals. However, these controls do not necessarily work as intended by the management, which makes them bad controls due to their coercive effect on employees. In other words, there is a gap between what managers want to achieve with controls and what they actually achieve (Tessier & Otley, 2012). This leads to the second issue: employee perception. Drawing back on SDT the environment created by the MCS needs to fulfil three needs – autonomy, competence, relatedness. However, whether or not these needs are fulfilled depends on how the individual perceives the used MC elements.
12
performance measures can cause people to lose sight of the strategy these measures try to represent and cause behaviour that is oriented to the (imperfect) measures rather than to the strategy (Choi et al., 2013). The risk that the used measures do not fully represent the strategy is even more likely as goals become more complex. Third, HEI staff have particularly high levels of intrinsic motivation and are less respondent to external rewards (e.g. Sutton & Brown, 2016). Consequentially, negative feedback created by e.g. performance evaluations would have a comparably much greater impact on employee motivation than, for instance, monetary rewards that reward positive behaviour. In fact, performance management systems have shown to undermine intrinsic motivation in HEIs (Ter Bogt & Scapens, 2012). An extensive use of diagnostic controls in combination with boundaries could strengthen the feeling of authority and control, create more distance to top management and pressure employees. As a result, employees would feel less autonomous and less related to the organization. Further, the increased pressure might cause employees to focus solely on performance targets and lose sight of the “big picture”, which would negatively affect their actual performance. Poor performance could also damage the feeling of competence.
Positive controls provide freedom, guidance, and inspiration. Employees in HEIs are primarily motivated intrinsically (Georgellis et al., 2011). Beliefs systems provide core values and purpose of the organization. Those are communicated through mission statements, credos and statements of purpose. An extensive use of beliefs systems can strengthen the employees’ feeling of being connected to their colleagues and students, being part of the organization and thus, satisfy the need for relatedness. Interactive controls promote dialogue throughout the organization to stimulate creative innovation. Subject to interactive control systems are face-to-face meetings, in which data generated by the system is discussed by managers, colleagues and subordinates to make action plans. This happens in a positive environment that encourages people to share information. Another feature of interactive controls is the use of positive feedback (Simons, 1994). Regular meetings and information sharing can give employees the feeling of being involved and of their effectiveness in the organization in a supportive and positive manner without restricting their autonomy. Positive feedback has shown to satisfy the need for competence (Gagné & Deci, 2005). Consequentially, based on SDT, interactive controls and an extensive use of beliefs systems are likely to be perceived as need-supportive and in turn, foster autonomous motivation. Hence, I hypothesize:
13 2.3 The moderating effect of job types
2.3.1 Job Characteristic Theory
Hackman & Oldham argue that the key for employee motivation arises from the work itself. The researchers found that tasks that are less stimulating can diminish motivation and productivity, whereas diversified activities have an enhancing effect. Job characteristics including task variety, task identity, task significance as well as autonomy and feedback from the job determine whether the individual reaches the three critical psychological states - experienced meaningfulness, experienced responsibility for outcomes, knowledge of the results of the activities - that are necessary for high levels of internal work motivation, performance and job satisfaction (e.g. Hackman & Oldham, 1975).
Figure 2. The job characteristic model adopted from Hackman and Oldham (1975)
14
Autonomy refers here to the amount of freedom and independence employees have in carrying out their work and is necessary to experience responsibility for the outcomes of the work. A job with high autonomy strengthens employees’ feeling of being responsible for their own efforts and decisions more than when being instructed by superiors (Hackman & Oldham, 1975).
Feedback is defined as “the degree to which carrying out the work activities required by the job results in the individual obtaining direct and clear information about the effectiveness of his or her performance” (Hackman & Oldham, 1975) and is crucial for the knowledge of actual results of work activities. Consequentially, these three psychological states determine intrinsic motivation and individual performance (Hackman & Oldham, 1975).
Recalling SDT, Hackman & Oldham seem to define the same necessary psychological needs for motivation and performance with the important addition of the influence of job characteristics. Internal work motivation can also be translated into autonomous motivation as it arises from within the individual, i.e. it is self-determined. The experienced meaningfulness is determined by the impact on others and refers to the same need as relatedness. In Hackman & Oldham’s model the job defines the level of autonomy, which then determine the experienced responsibility for outcomes. This is also enabled when the need for autonomy is satisfied. Lastly, the knowledge of results of the work serve the same function as the need for competence since individuals need to know what the result of their work to feel competent and motivated to then alter their behaviour accordingly. Overall, SDT agrees that job characteristics will enhance autonomous motivation (Gagné et al., 1997).
2.3.2 Hypothesis development
Job types are naturally different from each other in the way their job characteristics satisfy work-related basic needs and in turn cause employees to have different levels of autonomous motivation. Further, the MCS directly impacts individuals’ work environment as it sets the level of behavioural freedom and the degree and systematic way in which competence-relevant feedback and organizational definitions are provided. This environment can be either need-supportive or need-constraining. Consequentially, both the job characteristics and the MCS determine the level of autonomous motivation.
15
support the educational process including administrative jobs, finance and accounting jobs and management jobs and is generally characterized by more interdependence and less task uncertainty. In contrast, the second job type involves also uncertain tasks and copes with relatively new problems. Further, individuals perform tasks more independently. Relatively new problems require creativity to solve them (Perrow, 1986), for which more experimentation and flexibility are needed (Simons, 1994). In HEIs this job type is not only represented by research work, but also by high-quality teaching (Smith et al., 2014).
Both educational staff and educational support staff perform complex tasks that require various skills and talents. However, the educational job type could have relatively higher levels of task significance and task identity. Creative tasks are often performed by single employees or in small teams rather in a larger group. In addition, the outcome of those tasks can be easier attributed to individuals, whereas tasks that support the educational process are often performed to complete a larger and more complex task. Both enables educational employees to better experience tasks as a “whole” and identifiable piece of work and more directly understand their personal impact on others which determines the experienced meaningfulness of their work and is more likely to satisfy the need for relatedness. Moreover, the educational job type could get relatively more feedback from the job. Direct contact to students and peers, students’ grades, and the reputation inside and outside the institution provide direct and clear information about the effectiveness of their performance, which enables employees to feel competent. Lastly, the educational job type provides naturally more freedom and flexibility in how to carry out the work as employees often work independently and tasks are less routinized. This could cause employees to generally feel more autonomous. In sum, I argue that the educational job type is more likely to experience satisfaction of the all three basic needs – autonomy, relatedness and competence – and in turn has higher levels of autonomous motivation. Hence, I hypothesize:
H2: Individuals working in an educational job type are more autonomously motivated than individuals working in an educational support job type.
16
systems could strengthen the feeling of being connected to the organization in addition to the tasks itself and in turn enable the feeling of relatedness. Interactive controls stimulate the search for new ideas and opportunities and encourage experimentation and information sharing. Educational staff that cope with relatively more uncertain tasks are more likely to appreciate this system as it can help them to find solutions to relatively new problems. On the other hand, too much freedom can also erode predictability and cause role ambiguity, for instance, by not fully committing to budgets (Marginson & Ogden, 2005). Particularly, educational support staff that generally performs relatively more certain tasks that are predictable and often routinized could feel less guided by the system and less effective in their work, which would mitigate their feeling of competence.
17
H4 H1
H3 H2
H3: Autonomous motivation of individuals who work in an educational job type is stronger affected by the use of control.
2.4 Motivation and performance
“Motivation produces.” (Ryan & Deci, 2000). A
meta-analysis by Cerasoli et al. (2014)
provides strong evidence that all types of motivation enhance performance
. Particularly,autonomous motivation could lead to greater performance in HEI settings for several reasons. First, autonomous motivation is essential for tasks that are interesting. Employees who show great interest in an activity set themselves more challenging goals (Cerasoli & Ford, 2014), are willing to put more effort in their work and are eager to learn new skills (Simons et al., 2004). Second, autonomous motivation is beneficial when tasks are more complex (Grolnick & Ryan, 1987) or need discipline to complete them (Koestner & Losier, 2002). When goals and tasks are internalized, and basic needs are satisfied, the individual reaches its full cognitive and motivational potential which leads to better performance (Sheldon et al., 2003). Since employees in HEIs often perform complex task and are generally more interested in their work, increased autonomous motivation could lead to better performance. Hence,
I
hypothesize:
H4: Autonomous motivation is positively associated with performance.
2.5 Conceptual model
Figure 3. Conceptual model
18
III. METHODS
3.1 Research method and sample
In order to examine the effects of the MCS on autonomous motivation and the effects on performance this thesis uses a quantitative research approach. More precisely, the survey method is used as it allows to investigate the multi-faceted and complex phenomena between management control and human motivation that exist in nature, while keeping the necessary level of standardization for quantitative research and theory testing. For collecting data about individuals’ attitudes, beliefs and perceptions that motivate their behaviour the survey method is an appropriate tool (Speklé & Widener, 2018).
The data used for this study was collected by the University of Groningen in 2017. The university granted access to the database for the purpose of this thesis. The sample originates from employees and managers of two higher professional educational organizations in the Dutch public sector. Tessier & Otley (2012) argue that there is a gap between what management wants to achieve and what employees perceive. This thesis is concerned with employees’ perception of management control. Since managers are involved in designing the MCS, their perception differs from the one of employees. Hence, data from respondents with managerial functions was excluded from the analysis. The sample size of employee data has a total of 215 respondents and is presented in the descriptive statistics in Table 2.
3.2 Measures
The survey items that measure management control, motivation and performance were adopted from past studies.
3.2.1 Independent and dependent variables
19
the amount of negative controls in the package I compute a new variable that measures the positive-negative-control ratio (PNR) of the MC package. The PNR is calculated by dividing the sum of both positive controls (BELIEFS+INTER) by the sum of the two-negative control (BOUND+DIAGN). The PNR allows to measure the amount of positive controls versus negative controls on a continuous scale.
Autonomous motivation. Motivation was measured based on the different types of controlled and autonomous motivation proposed by SDT in Gagné & Deci (2005). The survey items were adopted from Gagné et al. (2014) who use the Multidimensional at Work Scale to measure intrinsic motivation as well as controlled and autonomous types of extrinsic motivation. Autonomous motivation was measured with 6 items, three items for intrinsic and three items for autonomous extrinsic motivation (Identification). All types of motivation were measured on a 7-point Likert scale.
Performance. Performance was measured on a unit level. An organizational unit is “…a more or less unified administrative entity within the larger organization in which the unit’s manager has at least some degree of authority over the set of tasks and processes of the unit.” (Speklé & Verbeeten, 2014). The five survey items were adopted from Speklé & Verbeeten (2014) and were measured on a 7-point Likert scale. Accordingly, performance is represented by the amount and accuracy of work, the number of innovations, improvements and new ideas, the reputation for work excellence, the attainment of goals, efficiency, and morale of the personnel.
Job type. The job type was determined by the main activities of the respondent. The item offered two choices (educational work or educational support work). The variable was then transformed into a dummy variable (EDUC_JOB) that indicates whether the employee mainly performs educational work or educational support work.
3.2.2 Control variables
Control variables are included since they could strongly influence the dependent variable. They must therefore be held constant to achieve reliable results. This thesis follows Spector and Brannick’s (2011) notion of explicitly control for variables based on theory or evidence rather than randomly including control variables. For this study, the following five control variables were selected.
Age. Motivation can be significantly influenced the age of the respondent. For instance, Inceoglu et al. (2012) found that older employees were more intrinsically motivated than younger ones.
20
contracts can increase stress, which would negatively affect autonomous motivation. Hence, I introduce type of contract as a dummy variable (Contract_type_dummy).
Employee agreement. Flexible working hours can positively influence employee motivation and performance significantly (Ahmad et al., 2013). Therefore, employee agreement was added as another dummy variable. The variable refers to the number of working hours of the employee and has two characteristics: fulltime agreement and part-time agreement (Agreement_type_dummy).
Educational background. The level of education could influence the level of employee motivation and performance. For instance, Kahya (2007) found differences in task performance depending on the respondent’s educational level.
Table 1. Construct descriptions
Construct α Description Measurement
Control variables
Age - Age Age of the respondent in years.
Tenure - Organizational tenure Tenure of the respondent in the organization.
Contract_type_dummy - Fixed term contract Contract type of the respondent {1=fixed term;
0= short-term}
Agreement_type_dummy - Fulltime agreement Employee agreement of the respondent
{1=fulltime; 0=part-time}
Education_dummy - Educational background Educational background of the respondent
{1=higher education (Bachelor; Master); 0=lower education (Secondary education; Secondary vocational education}
Independent and dependent variables
PNR -
Positive-Negative-Controls-Ratio of the MCS
Extent of positive controls relative to negative controls expressed by the quotient of (BELIEFS+INTER) and (BOUND+DIAGN).
BELIEFS .871 Beliefs systems Communication of organization's core values
based in Kruis et al. (2016).
BOUND .846 Boundary systems Communication of code of business conduct and
risks to be avoided based in Kruis et al. (2016).
DIAGN .965 Diagnostic control systems Use of accounting as part of a cybernetic control
cycle based on Bedford & Malmi (2015).
INTER .932 Interactive control systems Use of accounting information interactively
based on Bedford & Malmi (2015).
AUTON_MOT .823 Autonomous motivation Mean score of intrinsic motivation and extrinsic
autonomous motivation (identification) items based on Self- Determination Theory and adopted from Motivation-at-Work-Scale by Gagné et al. (2014)
PERF .721 Performance Perceived performance of the respondent
measured on a unit level.
EDUC_JOB - Educational job type Job type {1=Educational jon; 0=Educational
21
Tenure. A meta-analysis by Ng and Feldman (2010) found that organizational tenure can have a negative influence on the job performance. Organizational tenure was therefore introduced as another control variable.
3.3 Data analysis
3.3.1 Exploratory Factor Analysis and Reliability Analysis
The software IBM SPSS AMOS 26 is used to analyse the data. Before testing the hypotheses, I check the construct validity. First, I perform exploratory factor analyses on multi-item scales and retained measures following Yong & Pearce (2013). Accordingly, each measure must load into the correct factor. Items that load into multiple factors (cross-loadings) or load into the wrong factor are removed. When performing a factor analysis for all items including performance items, interactive controls and diagnostic controls items load into the same factor and are impossible to separate. When removing performance items from the analysis, they load into two different constructs. Hence, I perform three individual factor analyses: one for the four levers of control, one for autonomous motivation and one for performance. The final factor analysis is shown in the Appendix. Subsequently, multi-item constructs are computed using the mean score of the items. Next, I perform a reliability analysis for each construct to ensure the measurement scales are reliable. Cronbach alphas are calculated using the mean of all items of the construct. All multiple-item constructs have good reliabilities with Cronbach alphas ranging from .721 to .965 and are shown in Table 1.
3.3.2 Hypothesis testing
22
Model 1 is a multiple regression (1) to test for direct effects of each single lever of control (BELIEFS, BOUND, DAIGN, INTER) on autonomous motivation (AUTON_MOT). In all models I control for age (Age), tenure (Tenure), type of contract (Contract_dummy), employee agreement (Agreement_dummy) and educational background (Education_dummy).
(1) AUTON_MOT= β0 + β1 BELIEFS + β2 BOUND + β3 DIAGN + β4 INTER + β5 Age + β6 Tenure + β7 Contract_dummy + β8 Agreement_dummy + β9 Education_dummy + ε
In Model 2 a multiple regression (1) is performed test the effect of the use of positive relative to negative controls used in the MCS reflected by the PNR and the job type (EDUC_JOB) on autonomous motivation (AUTON_MOT) as hypothesized in H1 and H2.
(2) AUTON_MOT= β0 + β1 PNR + β2 EDUC_JOB + β3 Age + β4 Tenure + β5 Contract_dummy + β6 Agreement_dummy + β7 Education_dummy + ε
Model 3 is a moderated multiple regression model (3) to test the moderating effect of job type on the relationship between the PNR of the MCS and autonomous motivation. Moderated multiple regression measures the relationship between a dependent variable and an independent variable depending on the level of another independent variable (the moderator). The moderated relationship (interaction effect) is modelled by including a product term (PNR_x_EDUC_JOB) as an additional independent variable (e.g. Hartmann & Moers, 1999). Since H3 postulates a moderating effect of the job type on the relationship between the PNR of the MCS and autonomous motivation as measured by the multiplicative term, H3 is tested by examining the significance of the coefficient of the interaction term. As a result, if tests on the moderated regression reject the null hypothesis that the interaction coefficient is zero or negative, then the impact of PNR of the MCS on autonomous motivation is more positive for the educational job type and the hypothesis H3 would be supported.
(3) AUTON_MOT= β0 + β1 PNR + +β2 EDUC_JOB + β3 PNR_x_EDUC_JOB + β4 Age + β5 Tenure + β6 Contract_dummy + β7 Agreement_dummy + β8 Education_dummy + ε
23
(4) PERF= β0 + β1 AUTON_MOT + β2 Age + β3 Tenure + β4 Contract_ dummy + β5 Agreement_dummy + β6 Education_dummy + ε
IV. FINDINGS
4.1 Descriptive statistics
This section describes the findings about the general characteristics of the employee sample of the two HEIs that are shown in Table 2. The age of the respondents is divided into five groups. Almost half of the employees are older than 51 years, whereas only 4% are 30 years or younger. In addition, more than 40% of the employees have worked in the organization for more than 10 years. The respondents consist mainly of educational staff with 65%. Educational Table 2. Descriptive statistics sample
Variable Frequency (N=215) Percentage
Age 20-30 years 8 4% 31-40 years 53 25% 41-50 years 55 26% 51-60 69 32% 61+ 30 14% Gender Female 113 53% Male 102 47% Tenure (organizational) 0-5 years 67 31% 6-10 years 58 27% 11-20 years 52 24% 21-30 years 28 13% 31-40 years 9 4% 41+ years 1 < 1% Tenure (departmental) 0-5 years 88 41% 6-10 years 62 29% 11-20 years 44 20% 21-30 years 15 7% 31-40 years 5 2% 41+ years 1 < 1% Education Primary education 0 0% Bachelor’s degree 61 28%
Master’s degree or higher 129 60%
Secondary education 4 2%
Secondary vocational education 21 10%
Job type
Educational staff 139 65%
24
support staff made the remaining 35%. However, almost all respondents are highly educated, with 88% holding a bachelor’s degree or higher. The number of males and female employees is relatively equal in both organizations.
4.2 Early and late respondents
The surveys were sent out to the first organization starting from November 28 to December 12, 2017. For the second organization the surveys were sent from October 17 to December 30, 2017. In total, 75 employees responded from the first organization and 140 from the second organization. The response time of the respondents varied strongly. Early and late respondents are therefore determined by the first (early respondents) and fourth quantile (late respondents) of the response time. Table 3 shows the means and standard deviations of main constructs for early and late respondents of both organizations. Early and late respondents are analysed (1) to check if the two organizations show major differences in the means and standard deviations and (2) to test for non-response bias as late respondents can be a proxy for non-response. The means give first impressions about the representation of the lever in the MCS and the average level of autonomous motivation and performance of the respondents. Similar means of the four levers of control indicate that both organizations have a very balanced MCS with a PNR of around 1. Further, all employees show relatively high levels of autonomous motivation with means of 5.53 (First organization) and 5,65 (Second organization) and moderate performance with means of 3.56 (First organization) and 3.57 (Second organization). An independent 2-tailed t-test is performed for both organizations as
Table 3. Early and late respondents
First organization Second organization
Construct Total (n=75) Early respondents (1st quantile; n=19) Late respondents (4th quantile; n=19 Total (n=140) Early respondents (1st quantile; n=35) Late respondents (4th quantile; n=35) Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD
25
well as for early and late respondents of each organization. Results show that there are no significant differences in the means of the constructs in the two organization samples. Both samples can therefore be pooled. The second organization shows a significant difference in the mean of DIAGN (t(61.564)=-2.254; p=.028), which indicates that non-response bias could have been be an issue in the second organization and must be considered when interpreting the results.
4.3 Pearson correlation
The Pearson correlation matrix presents mutual correlations between the constructs that are used in the regression models as well as it can reveal multicollinearity issues between independent variables. All correlations between the variables are shown in Table 4. Significant correlations exist between the four levers of control. In addition, beliefs systems (BELIEFS) correlates significantly with autonomous motivation (AUTON_MOT). This effect is further tested in Model 1 (please see Table 5.). Performance (PERF) is the only variable that correlates significantly with autonomous motivation. Several control variables (8-12) correlate significantly with each other. Tenure and Age have the highest correlation coefficient (.593). However, all significant correlations in the variables that were used as independent variables are below the threshold .70 and should not cause multicollinearity problems in the regression analyses.
4.4 Hypothesis testing
The results from the multiple regression analyses to test the relationship between the PNR of the MCS and autonomous motivation are shown in Table 5. In Model 1 a multiple linear regression analysis was run to predict autonomous motivation from beliefs (BELIEFS), Table 4. Pearson correlation matrix
Construct (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (1) BELIEFS 1 (2) BOUND .361** 1 (3) DIAGN .364** .438** 1 (4) INTER .555** .485** .692** 1 (5) AUTON_MOT .210** -0,009 0,041 0,060 1 (6) EDUC_JOB -.183** -.206** -0,100 -.190** 0,122 1 (7) PERF 0,120 0,003 -0,017 0,012 .272** -0,009 1 (8) Age 0,057 0,015 0,007 -0,076 0,044 .138* -0,083 1 (9) Tenure 0,078 0,099 0,124 0,009 -0,074 -0,045 -0,041 .593** 1 (10) Contract_dummy -0,094 -0,045 -0,068 -0,129 -0,056 0,047 -0,069 .229** .344** 1 (11) Agreement_dummy -0,042 -0,032 -0,059 -0,110 0,061 -0,056 .243** -0,017 0,024 0,094 1 (12) Education_dummy -0,040 -.212** -0,051 -0,098 0,099 .430** -.138* 0,008 -.148* -0,036 -0,042 1
26
boundaries (BOUND), diagnostic controls (DIAGN) and interactive controls (INTER). A significant regression equation was found (F(9,205)=2.034, p<.05), with an R² of .082 (Adj. R²=.042). BELIEFS had a positive effect on autonomous motivation that was significant (p=.002). All other levers of control were not significant predictors of autonomous motivation.
In Model 2 a multiple regression was performed to examine whether the use of positive controls, relative to negative controls represented by the PNR of the MCS, and the job type (EDUC_JOB) could significantly predict employees’ autonomous motivation (AUTON_MOT) as proposed in H1 and H2. A significant regression equation was found (F(7,207)=1.961, p<.100), with an R² of .062 (Adj. R²=.030). The use of positive controls relative to negative controls had a positive significant effect on autonomous motivation (p=.017). This combined effect of the positive controls was higher than the direct effect of beliefs systems that was found
Table 5. Results of multiple linear regression analysis and moderation analysis Model 1 Coefficient estimate (Standard error) Model 2 Coefficient estimate (Standard error) Model 3 Coefficient estimate (Standard error) Model 4 Coefficient estimate (Standard error) Hypothesis Intercept 4.431*** (.469) 5.176*** (.401) (.486) 4.505 3.189*** (.031) BELIEFS .201** (.064) BOUND (.026) -.038 DIAGN (.068) .026 INTER (.076) -.039 PNR .427** (.178) (.339) .579* H1: supported
EDUC_JOB (.145) .192 (.421) .401 H2: not supported
PNR_x_EDUC (.398) -.210 H3: not supported
AUTON_MOT .134*** (.031) H4: supported Control variables Age (.007).010 (.007) .010 (.007) .009 (.003) -.004 Tenure (.009)-.014 (.009) -.010 (.009) -.010 (.004) .001 Contract_dummy (.209)-.061 (.208) -.150 (.210) -.138 (.096) -.098 Agreement_dummy (.125).141 (.125) .151 (.125) .150 .201*** (.057) Education_dummy (.198).208 (.215) .065 (.218) .048 -.213** (.090) R² .082 .062 .063 .162 Adjusted R² .042 .030 .027 .138 F value 2.034** 1.961* 1.745* 6.690***
27
in Model 1. The job type was not a significant predictor of autonomous motivation. Consequentially, H1 is supported and H2 must be rejected.
Model 3 examines the moderating effect of the job type as proposed in H3 in a moderation analysis. The dependent variable for the analysis is autonomous motivation (AUTON_MOT). Predictor variable for the analysis is the PNR of the MCS and the job type (EDUC_JOB). The moderator variable evaluated for the analysis is the educational job type (EDUC_JOB). The results show that there is no significant interaction effect between the PNR of the MCS and the job type. Hypothesis H3 must therefore be rejected.
Model 4
A multiple linear regression was performed to predict performance based on the degree of autonomous motivation as postulated in H4. A significant regression equation was found (F(6,208)=6.690, p<.001), with an R² of .162 (Adj. R²=.138). Autonomous motivation had a positive effect on performance that was significant (p<.0001). This strongly supports hypothesis H4. Further, results show a significant effect of the type of agreement on performance. Accordingly, fulltime employees perform better than part-time employees (p=.001). Finally, higher educated employees show a lower level of performance than lower educated employees. (p=.019). In sum, performance was predicted by autonomous motivation, type of agreement and educational level.
V. DISCUSSION AND CONCLUSION
28
impact individual’s motivation they point out three major differences. First, SDT expands the narrow focus of job characteristics as predictor for employee motivation by considering management style as major influential factor on autonomous motivation. Second, Job Characteristic Theory does not contemplate the compromising role of controlled motivation for autonomous motivation. Third, whereas the need strength that enhances motivation is central to Job Characteristic Theory, SDT is more concerned with the satisfaction of different needs that enable a specific type of motivation (Gagné & Deci, 2005). Since this study is only concerned with autonomous motivation both theories do not conflict with each other and can both be used to make hypotheses and interpretations. This thesis provides several significant findings both expected, based on theory and past studies, as well as surprising or somewhat unexpected.
First, the results show that the hypothesized relationship between management control and autonomous motivation indeed exist. In particular, findings indicate that an increased use of positive controls relative to negative controls in the MC package leads to more autonomous motivation. This outcome confirms Simons’ (1994) proposition that positive controls serve as a force to motivate, guide and provide freedom and are overall of supportive nature to the individual. STD explains this positive effect on autonomous motivation with the satisfaction of the three basic needs: autonomy, competence and relatedness. Another interesting finding was that solely a formal control system had a direct impact on autonomous motivation. As illustrated in the Model 1 (please see Table 5.) only beliefs systems had a positive effect on autonomous motivation that was significant, whereas all other control levers did not correlate with autonomous motivation directly. This complements a study in the Dutch public sector by Van der Kolk et al. (2019), who found that the communication of core norms and values (i.e. cultural controls) enhances intrinsic motivation. Different from a case study by Sutton & Brown (2016), who report positive effects of diagnostic controls such as performance evaluations on autonomous motivation of researchers in a university, I did not find a direct effect of the diagnostic or interactive use of control on autonomous motivation. Furthermore, there was no significant impact of boundaries, which indicates that boundaries were perceived as neutral rather than autonomy restricting. The examination of the MCS as a package showed an increase of autonomous motivation when more positive controls relative to negative controls were used, that was higher than the direct effect of beliefs systems. This could be explained by existing complementary effects of other MC elements within the system. However, more analysis is needed to make assumptions about what had caused this increased effect as I did not find a direct effect of interactive controls.
29
into two job types was chosen too broadly. For instance, one can imagine that teaching a first-year bachelor’s course with many participants involves different levels of task uncertainty and interdependence compared to teaching a small master’s course with only a few participants or supervising the writing of a master thesis. A similar spectrum of those two characteristics can be assumed in the variety of jobs that support the educational process and were examined as one job type in this study. Adler and Chen (2011) propose a third job type, that reflects both high levels of interdependence and task uncertainty. Although the hypothesis must be rejected in this study, I suggest more research on different job types.
Third, other than expected I could not find a significant interaction effect of management control and job type. The assumption that the MCS would have a stronger effect on the autonomous motivation of the educational staff can therefore not be declared correct. Instead, all employees perceived the MCS as equally need-supportive. Sheldon et al. (2003) state correctly that the control-oriented mindset of individuals that demands more structure and direction does not lead to individuals wanting to be more controlled and that all individuals benefit equally from more autonomy. Main argument for the assumption that the MCS affects the educational job type stronger than the educational support job type was the increased need of behavioural freedom due to higher task uncertainty and independence, which would be relatively easier constrained by an extensive use of negative controls and would at the same time benefit more from positive controls. Two factors could explain why this assumption was not supported. First, negative controls were not perceived as constraining at all. I did not find direct effects of any control elements on autonomous motivation other than beliefs systems which indicates that they were perceived as neutral. Second, there were very little outliers of strictly positive and strictly negative MCSs in the data sample. Overall, the MCS was perceived as very balanced by the employees indicating that both organizations made use of all four control levers to the same extent (please see Table 3.). This limited the effects of extreme uses of either of the opposing forces or single MC elements which could have constrained an individual. In sum, I have to reject the H3 with the remark that in other samples and settings results could have been different. Therefore, I suggest further research with a larger data sample.
Fourth, this thesis provides strong evidence that autonomous motivation is positively associated with performance and thus, substantiates past research on this topic (e.g. Van der Kolk et al., 2019; Sutton & Brown, 2016). This finding does not only underline the importance of having overall motivated employees in HEIs, but also confirms the significance of autonomous motivation as one specific type of motivation that drives performance.
30
for this effect could be that fulltime employees are overall more involved with their jobs compared to part-time employees as a meta-analysis by Thorsteinson (2003) confirms. Second, results show that employees with a lower educational level (secondary degree) rated their performance higher than those with a higher educational level (bachelor’s degree or higher), which supports a study by Kahya (2007), who reports a negative effect of the level of education on task performance. Finally, in my sample both performance and autonomous motivation were not influenced by age, tenure, or type of contract. These findings stand in contrast to the reviewed literature (Inceoglu et al., 2012; Ng and Feldman, 2010; Kinman et al. 1998).
This thesis makes several contributions to management accounting research. Most importantly is the positive effect of beliefs systems on autonomous motivation. Past research was mostly concerned with the use of the MCS represented by interactive and diagnostic controls but less with the role of beliefs and boundaries. The findings could stimulate more future research on those two formal control systems. Further, the examination of the MCS as a package contribute to past research on this topic. An important feature of this thesis is the examination of employee responses that allowed the investigation of human perception of management control. Scholars experimented with mixed surveys (e.g. Groen et al., 2017) to capture different perceptions on different issues (i.e. management and employee perception). For this thesis I followed Tessier and Otley’s (2012) recommendation to study employee responses instead of management response due to different perceptions. Accordingly, this thesis relied purely on employee data for management control, motivation and performance. This also allowed to better compare these three variables with each other. Additionally, the study of variables that were measured on an individual level such as motivation and management control in combination with a variable that was measured on a unit level (i.e. performance) add to these so-called ‘cross-level’ studies in prior research (e.g. Van der Kolk et al., 2019).
31
and give limited insight in human experience attached to this phenomenon. Future research could therefore investigate the relationships also in qualitative studies.
The aim of this thesis was to answer the question what the impact of the use of control on autonomous motivation for different job types is. In addition, this thesis aimed to substantiate the findings by providing evidence that autonomous motivation enhances performance. In regard to the research question I hypothesised a positive relationship between positive relative to negative controls and autonomous motivation (H1). Further, I hypothesized a relationship between job type and autonomous motivation so that the educational staff is more autonomously motivated than the educational support staff (H2). Furthermore, I assumed that an interaction effect between job type and management control exists (H2). In particular, I postulated a stronger perceived effect of the use of control on the educational staff (H2) than on the educational support staff. The findings support only H1. However, hypothesis H2 and H3 must be rejected. Lastly, the examination of the relationship between autonomous motivation and performance concluded that autonomous motivation and performance are positively associated, which supports H4. Prior research has focussed much attention on those control elements that determine the use of the MCS (i.e. interactive and diagnostic control) (e.g. Henri, 2006; Bobe & Taylor, 2010; Bisbe & Otley, 2004). Future research could continue this stream by focussing more on formal control systems (i.e. beliefs and boundaries) and on the opposing forces reflected by positive and negative controls. In addition, the comparison of different job types and their effect on different types of motivation could be studied. In particular, it would be interesting to study the effects on intrinsic and extrinsic motivation separately.
REFERENCES
Abernethy, M. A., & Lillis, A. M. (1995). The impact of manufacturing flexibility on management control system design.
Accounting, Organizations and Society, 20(4), 241-258.
Ahmad, A. R., Idris, M. T. M., & Hashim, M. H. (2013). A study of flexible working hours and motivation. Asian Social Science,
9(3), 208.
Ahrens, T., & Chapman, C. S. (2004). Accounting for flexibility and efficiency: A field study of management control systems in a restaurant chain. Contemporary Accounting Research, 21(2), 271-301.
Amabile, T. M. (1998). How to kill creativity Harvard Business School Publishing Boston, MA.
Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin, 117(3), 497.
Bobe, B., & Taylor, D. (2010). Use of management control systems in university faculties: Evidence of diagnostic versus interactive approaches by the upper echelons. Paper presented at the The Sixth Asia Pacific Interdisciplinary Research in Accounting
32
Buelens, M., & Van den Broeck, H. (2007). An analysis of differences in work motivation between public and private sector organizations. Public Administration Review, 67(1), 65-74.
Cerasoli, C. P., & Ford, M. T. (2014). Intrinsic motivation, performance, and the mediating role of mastery goal orientation: A test of self-determination theory. The Journal of psychology, 148(3), 267-286.
Cerasoli, C. P., Nicklin, J. M., & Ford, M. T. (2014). Intrinsic motivation and extrinsic incentives jointly predict performance: A 40-year meta-analysis. Psychological Bulletin, 140(4), 980.
Chenhall, R. H. (2003). Management control systems design within its organizational context: Findings from contingency-based research and directions for the future. Accounting, Organizations and Society, 28(2-3), 127-168.
Choi, J., Hecht, G. W., & Tayler, W. B. (2013). Strategy selection, surrogation, and strategic performance measurement systems.
Journal of Accounting Research, 51(1), 105-133.
Collier, P.M., 2005. Entrepreneurial control and the construction of a relevant accounting. Management Accounting Research
16, 321–339.
De Bruijn, H. (2007). Managing performance in the public sector Routledge.
Deci, E. L., & Ryan, R. M. (1985). Self-determination and intrinsic motivation in human behavior. EL Deci, RM Ryan.–1985. Frey, B. S. (2012). Crowding out and crowding in of intrinsic preferences. Reflexive Governance for Global Public Goods, 75, 78. Frey, B. S., Homberg, F., & Osterloh, M. (2013). Organizational control systems and pay-for-performance in the public service.
Organization Studies, 34(7), 949-972.
Gagné, M., & Deci, E. L. (2005). Self‐determination theory and work motivation. Journal of Organizational Behavior, 26(4), 331-362.
Gagné, M., Senecal, C. B., & Koestner, R. (1997). Proximal job characteristics, feelings of empowerment, and intrinsic motivation: A multidimensional model 1. Journal of Applied Social Psychology, 27(14), 1222-1240.
Georgellis, Y., Iossa, E., & Tabvuma, V. (2011). Crowding out intrinsic motivation in the public sector. Journal of Public
Administration Research and Theory, 21(3), 473-493.
Granlund, M., & Taipaleenmäki, J. (2005). Management control and controllership in new economy firms—a life cycle perspective. Management Accounting Research, 16(1), 21-57.
Groen, B. A., Wouters, M. J., & Wilderom, C. P. (2017). Employee participation, performance metrics, and job performance: A survey study based on self-determination theory. Management Accounting Research, 36, 51-66.
Grolnick, W. S., & Ryan, R. M. (1987). Autonomy in children's learning: An experimental and individual difference investigation.
Journal of Personality and Social Psychology, 52(5), 890.
Hackman, J. R., & Oldham, G. R. (1975). Development of the job diagnostic survey. Journal of Applied Psychology, 60(2), 159. Hackman, J. R., & Oldham, G. R. (1980). Work redesign.
Hagger, M. S., Hardcastle, S. J., Chater, A., Mallett, C., Pal, S., & Chatzisarantis, N. L. D. (2014). Autonomous and controlled motivational regulations for multiple health-related behaviors: between-and within-participants analyses. Health
Psychology and Behavioral Medicine: An Open Access Journal, 2(1), 565-601.
Hanaysha, J. R., & Majid, M. (2018). Employee motivation and its role in improving the productivity and organizational commitment at higher education institutions. Journal of Entrepreneurship and Business, 6(1), 17-28.
Hartmann, F. G., & Moers, F. (1999). Testing contingency hypotheses in budgetary research: An evaluation of the use of moderated regression analysis. Accounting, Organizations and Society, 24(4), 291-315.
Henri, J. (2006). Management control systems and strategy: A resource-based perspective. Accounting, Organizations and
Society, 31(6), 529-558.
Inceoglu, I., Segers, J., & Bartram, D. (2012). Age‐related differences in work motivation. Journal of Occupational and
33
Kahya, E. (2007). The effects of job characteristics and working conditions on job performance. International Journal of
Industrial Ergonomics, 37(6), 515-523.
Kinman, G., Jones, F., & Kinman, R. (2006). The well‐being of the UK academy, 1998–2004. Quality in Higher Education, 12(1), 15-27.
Koestner, R., & Losier, G. F. (2002). Distinguishing three ways of being highly motivated: A closer look at introjection, identification, and intrinsic motivation.
Kruis, A., Speklé, R. F., & Widener, S. K. (2016). The levers of control framework: An exploratory analysis of balance. Management
Accounting Research, 32, 27-44.
Kuchava, M. M., & Buchashvili, G. (2016). Staff motivation in private and public higher educational institutions (case of international black sea university, sokhumi state university and akaki tsereteli state university). Journal of Education &
Social Policy, 3(4), 92-100.
Lacy, F. J., & Sheehan, B. A. (1997). Job satisfaction among academic staff: An international perspective. Higher Education,
34(3), 305-322.
Malmi, T., & Brown, D. A. (2008). Management control systems as a package—Opportunities, challenges and research directions.
Management accounting research, 19(4), 287-300.
Marginson, D. E. (2002). Management control systems and their effects on strategy formation at middle‐management levels: evidence from a UK organization. Strategic management journal, 23(11), 1019-1031.
Marginson, D., & Ogden, S. (2005). Coping with ambiguity through the budget: The positive effects of budgetary targets on managers' budgeting behaviours. Accounting, Organizations and Society, 30(5), 435-456.
McInnis, C. (2000). The Work Roles of Academics in Australian Universities (Evaluations and Investigations Programme
Report No. 00/5). Canberra, ACT: AGPS.
McInnis, C. (1996). Change and diversity in the work patterns of australian academics. Higher Education Management, 8(2), 105-117.
Merchant, K. A., & Van der Stede, Wim A. (2007). Management control systems: Performance measurement, evaluation and
incentives Pearson Education.
Mohr, L. B. (1971). Organizational technology and organizational structure. Administrative Science Quarterly, , 444-459. Mundy, J. (2010). Creating dynamic tensions through a balanced use of management control systems. Accounting, Organizations
and Society, 35(5), 499-523.
Ng, T. W., & Feldman, D. C. (2010). Organizational tenure and job performance. Journal of Management, 36(5), 1220-1250. Nixon, W. A., & Burns, J. (2005). Management control in the 21st century. Management Accounting Research, 16(3), 260-268. Organisation for Economic Cooperation and Development, 2005, University Research Management: Developing Research in New
Institutions (OECD, Paris).
Parker, L., Guthrie, J., Milne, M., & Broadbent, J. (2008). Public sector to public services: 20 years of “contextual” accounting research. Accounting, Auditing & Accountability Journal,
Perrow, C. (1986). Economic theories of organization. Theory and society, 11-45.
Perry, J. L., Hondeghem, A., & Wise, L. R. (2010). Revisiting the motivational bases of public service: Twenty years of research and an agenda for the future. Public Administration Review, 70(5), 681-690.
Ryan, R. M. (1995). Psychological needs and the facilitation of integrative processes. Journal of Personality, 63(3), 397-427. Ryan, R. M., & Connell, J. P. (1989). Perceived locus of causality and internalization: Examining reasons for acting in two
domains. Journal of Personality and Social Psychology, 57(5), 749.
