Amsterdam Business School
The Move Towards a Total Rewards Strategy:
The Influence of Organization-based Self-esteem on Reward
Preferences
Name: Birgit Wingen Student number: 11089091
Thesis supervisor: Pr. Sander van der Triest Date: 17 June 2015
Word count: 15914
MSc Accountancy & Control, specialization [Control]
Statement of Originality
This document is written by student Birgit Wingen who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the
Abstract
In order to retain, motivate and attract employees necessary for organizational success, employers have to adapt more sophisticated reward strategies which pay more attention to individual preferences and offer a wide range of non-financial and financial rewards. In order to better understand individual preferences, the aim of this study is to observe whether personal employee characteristics such as organization-based self-esteem influence the preferences for rewards. The hypotheses are developed in accordance to the self-enhancement theory of Dipboye, which suggests that individuals with different levels of self-esteem possess different needs and thus have distinct preferences. Results indicate that the organization-based self-esteem of individuals in fact influences the reward preferences of employees. Individuals with a high level of organization-based self-esteem show a higher preference for non-financial intrinsic rewards, power rewards and individual pay for performance compared to employees with lower levels of organization-based self-esteem. The results of this study highlight the need for
organizations to distinguish between individual employees instead of applying a single reward strategy to all its employees, as a one-size fits all approach might waste organizations resources without leading to the desired benefits.
Table of Content
1 Introduction ... 7
1.1 Problem Definition and Contribution of the Paper ... 7
1.2 Course of Investigation and Methodology ... 8
2 Literature Review: Theory, Rewards and Organization-based Self-esteem ... 10
2.1 Theory ... 10
2.2 Rewards... 10
2.2.1 Definition of Rewards, Reward Systems and Reward Types ... 11
2.2.2 Financial and Non-financial Rewards ... 12
2.2.3 Goals and Benefits of Reward Strategies ... 13
2.2.4 Current Challenges of Reward Strategies ... 14
2.2.5 Reward Preferences ... 15
2.3 Organization-based Self-esteem ... 16
2.3.1 Stability and Dynamics of OBSE ... 16
2.3.2 Determinants and Consequences of OBSE and Their Connection to Rewards .... 17
2.3.3 Level of OBSE and its Influence on Needs and Preferences of Individuals ... 18
2.4 Summary and Theoretical Framework ... 20
3 Hypothesis Development ... 21
3.1 Preferences for Non-financial Intrinsic Rewards ... 21
3.2 Preferences for Non-financial Extrinsic Rewards ... 22
3.2.1 Support Rewards ... 22
3.2.2 Power Rewards ... 23
3.2.3 Relationships ... 24
3.3 Preferences for Financial Rewards ... 24
3.4 Preferences and Actual Rewards Provided ... 25
4.1 The Survey Method... 26
4.1.1 Sample Collection ... 26
4.1.2 Method of Data Collection... 27
4.2 Constructs ... 27
4.2.1 Reward Preferences ... 28
4.2.2 Organization-based Self-esteem and General Self-esteem ... 29
4.2.3 Control Variables ... 31
4.3 Further Statistical Analysis of the Data ... 31
5 Results ... 32
5.1 Sample Descriptives ... 32
5.2 Correlation... 34
5.3 Linear Regression – Hypothesis 1-5 ... 36
5.4 Linear Regression – Hypothesis 6 ... 38
5.5 Further Analysis... 39
6 Discussion ... 40
6.1 Summary and Practical Implications ... 40
6.2 Limitations... 42
6.3 Further Implications for Literature ... 43
References... 44
List of Figures
Figure 1: Classification of Financial Rewards According to Chiang and Birtch (2007) ... 12
Figure 2 : Classification of Non-Financial Rewards According to Chiang and Birtch (2007) ... 13
Figure 3: Theoretical Framework ... 20
Figure 4: Overview of the Hypotheses ... 25
Figure 5: Overview of the Demographics of the Sample ... 32
Figure 6: Descriptive Statistics of the Most Important Variables Investigated ... 33
Figure 7: Correlation Table... 35
Figure 8: Linear Regression Analysis – Hypothesis 1-5 ... 36
Figure 9: Linear Regression Analysis – Hypothesis 6 ... 38
Figure 10: Linear Regression Analysis – Further Results ... 39
List of Abbreviations FR – Financial Rewards GSE – General Self-esteem K – Kurtosis
KMO – Kaiser-Mayer Olkin M – Mean
NFE – Non-financial Extrinsic Rewards NFI – Non-financial Intrinsic Rewards OBSE – Organization-based Self-esteem PCA – Principal Components Factor Analysis SD – Standard Deviation
1 Introduction
In the following the topic Reward Strategy is introduced and the contribution of this thesis is depicted. Furthermore, a course of investigation and methodology is provided.
1.1 Problem Definition and Contribution of the Paper
Decreased employee motivation, high voluntary turnover, and scarcity of skilled personnel challenge organizations currently on a daily basis (Mc Kinsey, 2009; Chapman & Kelliher, 2012). Managing and motivating employees has therefore become key in retaining a qualified workforce (Lengnick-Hall, 2002).
Reward systems, as part of the performance measurement system, can serve this purpose as their main goal is to motivate employees and align employee and organizational objectives (Ferreira & Otley, 2009; Merchant et al., 2003; Malmi & Brown, 2008). Rewards represent “compensation […] an employee receives from an organization in exchange for his or her services” (p.1, Snelgar et al., 2013) and their spectrum ranges from financial rewards such as salary or bonuses to non-financial rewards such as job variety, feedback or work-life balance of a job position.
Until now employers and research mostly focused on financial rewards, however in order to stay competitive in a more complex environment this is no longer sufficient. To overcome current challenges organizations have to move towards a “total rewards strategy” (Luthans, 2000; Moore & Bussin, 2006; Demartini, 2013; Kerrin & Oliver, 2002). This strategy calls for an implementation of a wider range of rewards as well as to tailor rewards to individual employee preferences.
However, the reward preferences of individuals might differ and are not easily observable. Studies identified demographic variables such as age, gender or country to be explanatory for differences in reward preferences (see for instance von Bonsdorff, 2011; Gunkel, 2005; Kominis & Emmanuel, 2005). Despite some first studies, the literature is still fragmented, as little empirical evidence for different reward preferences exists (Gunkel, 2005; Kominis & Emmanuel, 2005). Specifically, most studies focus on demographic variables and leave personality traits of employees aside. This research study is one attempt to further advance literature which examines reward preferences of individuals based on their personal characteristic by studying the self-esteem of individuals.
I hypothesize organization-based self-esteem (OBSE) to influence the reward preferences of individuals and thus examine the following research question: Does the level of OBSE influence the reward preferences of employees?
OBSE is defined as the value organizational members believe themselves to have within an organization (Pierce et al., 1989). I specifically select OBSE as possible variable, as according to self-enhancement theory individuals with different levels of OBSE possess different needs. This discrepancy in needs is expected to lead to differences in individuals preferences. For instance, individuals with low-levels of OBSE constantly strive to increase their level of self-esteem. Thus, they might value support rewards, such as feedback to a greater extent than their high OBSE counterparts.
Furthermore, this research study will compare the individual reward preferences with the actual rewards employees receive. This will give additional insights on whether a reward mix which is aligned to the needs of employees can in turn increase the OBSE of individuals.
The research will contribute to both, theory and practice. As far as I know, there is no study which examines the relationship between OBSE and reward preferences. Thus, research still lacks insides on the influence of personal characteristics on reward preferences. Furthermore, the study does not only consider financial rewards, but also non-financial rewards which are rarely studied in research (Ferreira & Otley, 2009; Chapman & Kelliher, 2011). These theoretical contributions can further serve practitioners with valuable insides on how to structure reward systems of their employees most effectively.
1.2 Course of Investigation and Methodology
Firstly, underlying theories are introduced as these theories built the foundation for a further understanding of the research topic. Afterwards, the literature review provides an overview of definitions, classifications and current developments within the research area of “rewards” as well as OBSE. To round up the theoretical analysis, a summary and theoretical framework will be provided and hypotheses are developed based on self-enhancement theory. In the practical part of the thesis, the methodological foundations of the analysis will be examined, as this is necessary to retrace the practical assessment. Hence, detailed information on the sample collection, the method of data collection and the variable constructs is provided. Following, the results of the statistical analysis are depicted in detail. The thesis will end with an overall conclusion and with a discussion of further implications for both theory and practice.
For the theoretical review, an extensive literature analysis was conducted with the aim of including all relevant literature. Using a snowball technique, relevant information was gathered from databases and libraries. Different types of literature such as journals and books were examined. The work of Korman (1976), Dipboye (1977), Herzberg (1966, 1987), Maslow (1968, 1954), Pierce and Gardner (2004), Chiang and Birtch (2007) Armstrong (1999) and Chiu et al. (2002) has had a substantial influence on the field.
2 Literature Review: Theory, Rewards and Organization-based
Self-esteem
The goal of the theoretical part of this paper is to provide the reader with an overview of the wide range of literature. Underlying theories are introduced first, followed by a literature review on rewards and OBSE. Furthermore, a theoretical framework is provided which summarizes the relationships uncovered in the theoretical analysis.
2.1 Theory
One of the most prominent motivation theories is Maslow’s Hierarchy of Needs. Maslow (1968, 1954) distinguishes five different needs: physiological, safety, social, esteem and self-actualization. They occur in hierarchical order: Individuals first try to satisfy the more basic needs and once they are satisfied this needs have no motivational effect anymore.
Transferred to an organizational setting, Maslow’s theory stresses the need for differentiation of reward preferences. Different employees might be motivated by distinct rewards, dependent on their current hierarchical level. For instance, at a certain stage it might only be possible to increase employees’ motivation by implementing non-financial rewards, as pure financial rewards are not able to account for higher needs such as self-actualization once the employee reached a certain hierarchical level. This has to be taken into account by managers when designing the reward system (Deci, 1972).
In addition, the expectancy theory of Vroom (1964) is worth being mentioned as it expects that “motivation is not universal” (p.15, Gunkel, 2006). While working, individuals try to maximize their utility. Three different relationships are distinguished: the effort-performance, performance-reward, and reward-personal goal relationship (Gunkel, 2006). The first refers to the probability that an employee believes that his effort will lead into improved performance. The second describes how this improved performance is rewarded by the employer. The third relationship depicts how individuals are attracted by the rewards offered. Thus, different employees are not necessarily attracted by the same rewards and the motivational effect is dependent on how valuable the rewards provided are for the individual employee. (Gunkel, 2006)
2.2 Rewards
This section will introduce and define rewards, reward systems and reward types. The reward types most important for the study at hand will be compared in detail. In addition, the
importance of the reward system within the overall management control system will be highlighted by referring to the benefits of a well-designed reward system. Furthermore, the challenges organizations face in recent years regarding their reward strategies are summarized, as they illustrate the necessity to move towards new approaches. Finally, current research around reward preferences and possible antecedents are introduced.
2.2.1 Definition of Rewards, Reward Systems and Reward Types
Rewards can be defined as “compensation […] an employee receives from an organization in exchange for his or her services” (p.1, Snelgar et al., 2013). This compensation can include anything which is of value for the employee and thus the concept is broad in its very nature. Chiang and Birtch (2007) distinguish between reward systems and reward types.
The reward system refers more broadly to the method or strategy used by an organization to distribute rewards to its employees (Chiang & Birtch, 2007). For instance, organizations can decide to distribute their rewards either based on the employees’ performance or based on other criteria such as years of service. The most common distribution strategies are equity, equality, needs and seniority (Chen, 1995; Fischer, 2004). When distributing rewards based on an equity strategy, the proportion individual employees receive is dependent on their performance, while equality refers to an equal allocation between all members of an organization. The needs strategy suggests an allocation based on specific employee needs such as medical or financial problems, while seniority refers to the greatest allocation of rewards to older and more senior employees. Reward systems are one aspect of the overall management control system, which also includes other dimensions, such as culture, planning or organizational structure (Malmi & Brown, 2008; Ferreira & Otley, 2009; Merchant et al., 2003). However, especially in times in which retaining and motivating employees becomes key for organizational success, the reward system plays a vital role within the management control system (Snelgar et al., 2013). Employees are claimed to be one of the most important resources of organizations and can provide a competitive advantage, which is not easily imitable by competitors (Boyd & Salamin, 2001).
Reward type refers to the inherent nature of the rewards itself. Relevant literature classifies rewards as financial vs. non-financial (Chiang & Birtch, 2006; Ferreira & Otley, 2009; Armstrong, 1999; Luthans, 2000; Demartini, 2013; Kerrin & Oliver, 2002), direct vs. indirect (Chiang & Birtch, 2007), extrinsic vs. intrinsic (Kominis & Emmanuel, 2005; Malmi & Brown, 2008; Flamholtz et al., 1985), or incentive vs. non-incentive based (Chiang & Birtch, 2007).
2.2.2 Financial and Non-financial Rewards
Within this thesis the classification of financial and non-financial reward types is adapted, as this classification is most commonly referred to in research. While research merely focused on studying financial compensation aspects, current studies call for a consideration of different reward types and a move towards a “total rewards” strategy (Luthans, 2000; Moore & Bussin, 2006; Demartini, 2013; Chapman & Kelliher, 2011; Chiang & Birtch, 2007).
This total reward mix contains both, financial as well as non-financial rewards. Financial rewards are foremost wages, as well as options, bonuses, stocks, insurances or retirement benefits. Non-financial rewards are of multiple nature and range from employment securities, promotion possibilities, recognition and attention of the superior, status, social relationships, work conditions, feedback, job challenges to development opportunities (Chiang & Birtch, 2007; Mc Kinsey, 2009; von Bonsdorff, 2011; Luthans, 2000; Moore & Bussin, 2006). Chiang and Birtch (2007) further differentiate non-financial intrinsic and non-financial extrinsic rewards. The latter are “tangible awards attached to the job” (p.576, Chiang & Birtch, 2007). Relationships at work or work conditions are examples of non-financial extrinsic rewards. In contrast, non-financial intrinsic rewards arise from the nature of the job itself such as autonomy or challenge connected with one’s position. It is worth noting that research also identifies negative non-financial and financial rewards such as lay-off (Demartini, 2013). However, this thesis will focus on positive rewards and therefore negative rewards are not discussed in greater detail. The following tables provide an overview of the classification of reward types according to Chiang and Birtch (2007). For each category examples are provided.
Figure 1: Classification of Financial Rewards According to Chiang and Birtch (2007)
Indirect
Fixed Variable Family services
Base salary Individual pay for performance Health benefits Annual salary increase Team pay for performance Insurance benefits
Perks Leave
Financial Rewards Direct
Figure 2 : Classification of Non-Financial Rewards According to Chiang and Birtch (2007)
While literature mainly focused on studying financial rewards, current studies note the importance for also considering non-financial rewards. This is because compensation is one of the largest expenses of organizations and competitive pressure often leads to a cut in remuneration expenses (Chen, 1995). A study of Mc Kinsey (2009) noted a 15 percent decrease in the overall remuneration expenses of organizations (p.3). Non-financial rewards are thus a low cost alternative (Luthans, 2000). In addition, non-financial rewards can have a more long-term effect on motivation compared to financial rewards (Mc Kinsey, 2009). This is not to say that financial rewards are not important anymore, as money is still one of the strongest motivators for employees (Cox, 2000). The crucial point however is that both types of rewards can have a strong motivational impact and thus also foster employee performance (Luthans, 2000). The right mix of reward types can provide organizations with various benefits and advantages in competing others (Moore & Bussin, 2006).
2.2.3 Goals and Benefits of Reward Strategies
The benefits of reward systems highlighted in this section exemplify the importance of the reward system within the overall management control system and therefore also illustrate the motivation for this further research study within this area.
The three main goals of reward systems are to recruit, retain and motivate employees necessary for organizational success (Demartini, 2013). The more attractive the reward system, the more qualified individuals apply for open positions, which makes the recruitment process of organizations more efficient. In addition, an appealing reward system helps to retain the existing workforce of an organization. A higher retention rate itself brings next to direct benefits, such as cost reductions due to less recruitment effort, also indirect benefits (Frank et al., 2004). For instance, customers do not have to switch between contact persons during the customer service process or confidential information is kept inside the organization. Thirdly, both financial as well as non-financial rewards have a strong motivational effect on employees.
Intrinsic
Support Power Relationships Accomplishment
Training Promotion Rel. to superior Job nature Development Authority Rel. to customers Job challenge
Workload Title Rel. to colleagues Job responsibility
Decision-making Job variety
Nonfinancial Rewards Extrinsic
Hence, reward systems serve multiple purposes at the same time. Next to the main three benefits introduced above research also identified organizational commitment, job satisfaction, engagement, performance of individuals, legal compliance, labor cost control and perceived fairness of employees to be positively correlated to a well-developed reward system (Kerrin & Oliver, 2002; Luthans, 2000; Moore & Bussin, 2006). Moreover, reward systems serve as a communication tool between employers and employees (Demartini, 2013).
However, these benefits are only obtained when reward strategies and systems are designed well, which currently becomes more challenging. For instance, inadequate compensation is one of the main reasons for voluntary turnover (Snelgar et al., 2013).
2.2.4 Current Challenges of Reward Strategies
Organizations nowadays face multiple challenges in regard to their reward design and system. Due to these challenges organizations have to advance their reward strategies and suit it to a fast-changing and complex economic environment.
Research notes increased competition among organizations to retain high performing individuals, as skilled personnel becomes scarce (Frank et al., 2004; Mc Kinsey, 2009). Thus, employees themselves are in the position to become more demanding and they gain sufficient influence on the selection of rewards (Chapman & Kelliher, 2011). High voluntary turnover exemplifies this increased power of employees. Next to difficulties in retaining employees, research studies also note a general decline in engagement, trust, motivation and loyalty of employees (Mc Kinsey, 2009; Luthans, 2000). As there is a direct link between an organizations financial performance and employee engagement (Frank et al., 2004) this situation can be a major threat for organizations. One way to fight these developments and stay competitive across competing organizations is to let employees exploit their influence and consider their preferences to a greater extent.
In addition, cultural, life-stage and generational differences leave organizations often unaware of how employees are best retained and motivated, as great differences exist between different employee groups, cultures or generations. Organizations seem often to be reluctant to evaluate their reward strategies and thus gain knowledge on what employees really prefer (Armstrong et al., 2011). Thus, organizations can improve their competitive situation by measuring and evaluating their strategies regarding its success and the satisfaction of its employees.
Overall, these challenges demand that organizations rethink and adapt their current reward strategies in order to be able to retain, motivate and recruit the individuals necessary for organizational success (Frank et al., 2004; Chapman & Kelliher, 2011).
2.2.5 Reward Preferences
Due to current challenges a “one-size-fits-all” approach concerning reward systems is no longer sufficient and rewards should be “tailor-made to suit individual employees” (p.1, Snelgar et al., 2013; Moore & Bussin, 2006; Frank et al., 2004). Thus, individual reward preferences have to be considered and identified by employers. This research study contributes to the target of better predictions of individual employee preferences.
Previously, research often focused on identifying links between the reward strategy of an organization and different contingency factors. It was claimed that the reward strategy is at its best designed congruent to certain factors such as ownership, environmental uncertainty, size or business strategy. For instance, a good fit between the business strategy and the reward strategy is claimed to be most beneficial for organizations. Nevertheless, a contingency approach does only account for differences between organizations but does not account for individual differences in preferences within one organization and is therefore recently criticized in research. (Chapman & Kelliher, 2011; Cox, 2000)
In contrast, the “total rewards strategy” introduced in one of the previous chapters does not only call for considering a more diverse reward package in general, but also stresses the need to tailor rewards to individual organizational members. Research notes that especially the integration of employees into the strategy development process can contribute substantially to a successful compensation strategy (Cox, 2000). However, this can be challenging, especially for larger organizations, which would have to consider thousands of individual preferences. A possible solution is to build employee “cluster” and apply different reward strategies for each cluster. It is important to note that this is still a more sophisticated strategy than applying a reward strategy based on contingency theory, as employees within an organization are separated and treated according to their classification. Research investigated various demographic variables in order to be able to cluster employee preferences accordingly. For instance, job level, geographic location, cultures, gender differences, or generations are possible cluster solutions (see for instance Chiang & Birtch, 2007; Chiu et al., 2002; Gunkel, 2006; Kominis & Emmanuel, 2005; Rainey, 1982; Wine et al., 2005; Moore & Bussin, 2006).
However, research studies observing these demographic variables often reveal controversial results. For instance, while some studies found statistically significant differences between
generations, other studies did not (Bussin & Rooy, 2014). Moreover, some researchers claim that globalization is removing the cultural differences between countries, while individual differences within countries will still remain to exist (Snelgar et al., 2013). More research is necessary on these demographic variables in order for organizations to be able to base their strategies on these results.
While there are already some attempts in research to observe demographic variables and their influence on reward preferences, research still lacks considerable insides on how personal employee characteristics influence reward preferences. In the following, self-esteem is introduced as one possible variable that influences reward preferences in practice.
2.3 Organization-based Self-esteem
Pierce et al. (1989) define self-esteem as a “self-evaluation that individuals make and maintain with regard to themselves” (p.625). Transferred to an organizational context it is the value organizational members believe themselves to have within an organization and thus expresses the extent to which individuals consider themselves to be important, meaningful and worthwhile for the organization. (Pierce et al., 1989)
Self-esteem is a multidimensional concept and is to be found on all different levels and hierarchies. Pierce and Gardner (2004) specifically distinguish general or global, organizational-based and task-specific self-esteem, while other researcher further differentiate between emotional, social, intellectual, physical and moral self-esteem (Krampe, 2004; Dipboye, 1977). General self-esteem (GSE) refers to individuals overall self-evaluation of themselves and is often a function of the various situational and task-specific self-esteems individuals possess.
2.3.1 Stability and Dynamics of OBSE
The exact dynamics between different levels of self-esteem however are a big controversy in research (Krampe, 2004). Different forms of self-esteem are often correlated and influence each other but might also substantially differ.
For instance, if one has low confidence in his or her performance regarding individual tasks, one might also have less confidence regarding his overall job performance (Pierce et al., 1989). Congruently, research found mostly positive relationships between different forms of self-esteem (Pierce & Gardner, 2004; Jex & Elacqua, 1999). On the other hand, one might perceive him or herself to be important in one specific role – as a student or athlete, but not in another - as an employee (Bowling et al., 2010). Tenure is a good example for a variable that impacts the
relationship between GSE and OBSE over time in an organizational setting: The shorter an employee works within an organization, the more its OBSE might reflect his GSE, however over time as individuals develop and gain experiences within an organization the correlation between OBSE and GSE might become less strong (Pierce & Gardner, 2004). In addition, the weight individuals place on specific self-esteem forms and thus their overall influence on GSE is not necessarily the same for various individuals (Krampe, 2004).
Considering these dynamics, this study will test the influence of both OBSE and GSE on reward preferences. A positive relation is expected between OBSE and GSE, but dependent on certain factors, they might also differ. In the following, I will only directly refer to OBSE but similar results are expected for both OBSE and GSE.
2.3.2 Determinants and Consequences of OBSE and Their Connection to Rewards Research around OBSE investigated various antecedents and consequences of OBSE. Knowledge of both is vital for organizations: The consequences highlight the importance of OBSE, as it is positively correlated to various organizational benefits. Several determinants can help organizations to increase OBSE of their employees and thus positively affect organizational success.
Determinants of OBSE can be grouped in different categories (Pierce & Gardner, 2004; Schütz & Tice, 1997, Krampe 2004). Firstly, OBSE is influenced by signals derived from the organizational structure. For instance, high degrees of standardization and mechanistic work can indicate to employees that there is no trust in individuals and thus can negatively affect OBSE, while job complexity is positively correlated with OBSE (Vecchio, 2000; Chattopadhyay & George, 2001). Secondly, OBSE is formed around the messages derived from social interactions with other organizational members or from the organizational culture. Examples are relationships and communication with superiors or other employees, the job and pay-level of individuals or work-place discrimination (Kark et al., 2003; Pierce et al., 1989; Lee, 2003). Lastly, OBSE is influenced by personal experiences within the work environment itself. For instance, the successful completion of a project can positively influence the OBSE of employees. The probability that an organizational member is able to fulfil its tasks and projects successfully is in turn dependent on the job fit, socialization efforts by the organization, training or available resources (Pierce & Gardner, 2004; Riordan et al., 2001).
Plenty of the determinants identified above actually meet the definition of rewards and compensation. For instance, research found that financial rewards, such as the pay level as well as non-financial rewards such as superior recognition, feedback, training or job complexity
influence OBSE. Thus, we can conclude that rewards seem to be a major driver of the level of OBSE of employees.
In addition, various consequences of OBSE have been researched. These consequences highlight the importance of OBSE for organizational success, as high OBSE is associated with various positive outcomes. For instance, OBSE is positively correlated to motivation, job satisfaction, job involvement, job commitment, adaption of organizational change and performance (Dipboye, 1977; Pierce & Gardner, 2004; Pierce et al., 1989; Hui & Lee, 2000; Simpson & Boyle, 1975). In addition, research found OBSE to be negatively correlated to turnover, turnover intensions as well as depression and other physical diseases (Bowling et al., 2010; Vecchio, 2000). Most of the consequences listed above have already been identified in section 2.2.4 as several of the main challenges organizations currently face. This exemplifies the crucial importance for organizations to consider OBSE, as higher levels of OBSE can significantly reduce these threating challenges.
In summary, both forms of rewards, financial and non-financial ones, seem to be determinants of OBSE. In addition, higher levels of OBSE lead to positive economic outcomes for firms. However, while research clarified that rewards seem to influence OBSE, it remains unanswered whether high and low OBSE individuals have different needs and therefore value rewards differently.
2.3.3 Level of OBSE and its Influence on Needs and Preferences of Individuals
Two theories explore the relationship between the level of self-esteem and its implications for different needs of individuals: self-consistency and self-enhancement theory.
Self-consistency theory claims that people will engage in behavior that is consistent with their self-evaluation (Korman, 1970). In consequence, employees with a high level of OBSE will engage in actions that strengthen or demonstrate their positive self-evaluation, while employees with a low level of OBSE will withhold effort and engage in behavior that reinforces their negative image. However, critical work on consistency theory evolved, as self-enhancement is a viable alternative theory (Dipboye, 1977).
Self-enhancement refers to a “genre of theories assuming that humans have a fundamental need to achieve and maintain high levels of self-esteem” (p.11, Dipboye, 1977). Therefore, it assumes that individuals with a low level of OBSE will constantly strive to increase their self-image, while individuals with high OBSE will engage in behavior that maintains, represents or increases their level of confidence further. The rationale behind self-enhancement is the general tendency of
individuals to seek success and to minimize personnel failure. Dipboye (1977) concludes that individuals with a low level of self-esteem might even try to increase their OBSE to a greater extent than their high OBSE counterparts, as they intend to alter their current state.
Research studies generally reveal that the performance of individuals with lower levels of OBSE is worse compared to high esteem individuals. Both theories, enhancement and self-consistency theory, provide a viable explanation for this result: Korman (1979) argues that the lower performance is due to individuals acting consistently to their self-image, while Dipboye (1977) claims that individuals with lower levels of OBSE engage in so called “damage controls” (see also Pierce & Gardner, 2004). Thus, their main goal is to prevent that their negative self-image is confirmed by further failures. Therefore it is not a lack of ability or the reinforcement of their negative self-image that leads to lower performance, but rather their cautious behavior. In practice it seems most likely that both theories coexist in a way, as both provide viable explanations. However, there might be situations in which one of the two is more or less strong. Dipboye (1977) predicts that the greater the impact of a situation on GSE, the more likely it is that individuals will engage in self-enhancing behavior. Transferred to the organizational setting of this research study, self-enhancement theory seems more viable. The reason is that work still plays a decisive role in nowadays society and individuals life’s. People spend more time at work than with family and friends and therefore the OBSE of individuals might impact their GSE to a great extent. In addition, self-enhancement theory seems more reasonable in situations in which individuals are generally uncertain about their level of OBSE. Especially employees with a short tenure might be uncertain and therefore rather engage in self-enhancing behavior. Furthermore, employees are exposed to various organizational changes, such as new IT applications. Thus, they nowadays rather experience an unstable working environment and are therefore again more uncertain about their actual level of OBSE.
The following framework summarizes the relationships examined in this literature review and indicates the missing gap in literature. The main focus of this research study is to investigate the impact of individuals’ level of OBSE on their reward preferences. These differences are expected to exist as individuals with different levels of OBSE occupy different needs.
Figure 3: Theoretical Framework
2.4 Summary and Theoretical Framework
Sufficient reward strategies can help organizations to retain, motivate and attract employees. Due to current challenges, such as shortage of skilled labor or decreased motivation, employees themselves become more demanding in regard to the rewards they receive and organizations struggle in retaining, motivating and attracting employees. Thus, research recommends organizations to tailor rewards to the preferences of individual employees and offer a broader range of rewards, a strategy called “total rewards strategy”. However, research still lacks a clear picture on what characteristics influence the preferences of individuals. In this study special attention is given to OBSE as one variable which might influence individual preferences.
In addition, an investigation and comparison of individual preferences and the actual rewards employees receive in practice will reveal additional insides on how far organizations are in considering individual needs. Thus, next to reward preferences this study will also examine the actual rewards individuals receive.
3 Hypothesis Development
Due to the analysis in chapter 2.3.3 self-enhancement theory is used as underlying framework to develop the hypotheses.
According to the theory of self-enhancement, there is a general tendency of man to increase one’s level of self-esteem (Dipboye, 1997). Transferred to the organizational setting of this study, individuals with low levels of esteem will try to increase their level of organization based self-esteem and thus I predict them to prefer rewards that are able to fulfill this need. Also employees with a high level of OBSE have a general tendency to increase their level of self-esteem further. However, their needs still substantially differ from individuals with low levels of OBSE. As they already reached a high level of OBSE and therefore “like who and what they are” (p.592, Pierce & Gardner, 2004) they will more likely prefer rewards that help them maintain their level of OBSE, as well as prefer rewards which allow them to demonstrate their competencies to their environment (Krampe, 2014; Dipboye, 1997).
Therefore, it is expected that due to the different needs high and low OBSE individuals will differ in regard to their reward preferences. The circular relationship provided in the framework in section 2.3.3 reveals that if organizations consider these needs and the right set of financial and non-financial rewards is provided, this contributes positively to the OBSE of individuals and thus also increases motivation, satisfaction and performance.
The literature review introduced a variety of different financial and non-financial rewards. However, this thesis cannot provide hypothesis for all of the proposed reward types. Therefore, the study will focus on the reward categories developed by Chiang and Birtch (2007) which are expected to vary most for different levels of self-esteem.
3.1 Preferences for Non-financial Intrinsic Rewards
Among the non-financial intrinsic rewards Chiang and Birtch (2007) distinguish job nature, job challenge, job responsibility, skills used opportunities, job variety, job autonomy, accomplishment and job satisfaction (p.588).
These types of rewards are expected to be preferred by employees with a high level of OBSE in contrast to their low OBSE counterparts. The roots for this prediction can be found in the needs analysis introduced above. While employees with a low level of OBSE will generally seek to increase their OBSE, they are more careful to engage in any behavior that will expose them to risk and might confirm their low level of self-esteem (p.116, Dipboye, 1977). In contrast,
individuals with a high level of OBSE prefer rewards that help them to maintain their positive self-image and allow them to demonstrate their competencies to their working environment. Job challenges, responsibilities and skills used opportunities will perfectly serve this purpose.
In addition, individuals with a high level of OBSE are predicted to be more concerned about their job nature and job variety. This is because they generally perceive themselves to fulfill all job related requirements sufficiently and thus might value variety to a greater extent than their low OBSE counterparts.
Lastly, individuals with high levels of OBSE will by trend conceive their job as more important than low OBSE individuals. Again, the latter are concerned about increasing their overall level of self-esteem and therefore try to define themselves also through other competence areas in which they feel more secure. Thus, accomplishment and job satisfaction are predicted to be more important for individuals with a high level of OBSE.
Hypothesis 1: The higher the level of individuals’ OBSE, the greater their preference for non-financial intrinsic rewards.
3.2 Preferences for Non-financial Extrinsic Rewards
Chiang and Birtch (2007) divide non-financial extrinsic rewards in further subcategories: support rewards, power rewards and relationship rewards. For all categories hypotheses will be developed in the following.
3.2.1 Support Rewards
Chiang and Birtch (2007) group eight different reward items into the category of support rewards: team spirit, fair treatment, training and development, balance life and work, resources support, workload and work environment. Within this thesis the scale of Chiang and Birtch is extended by one more support reward: feedback. The reason is that research around rewards often includes this item and thus it should be integrated into the analysis as well.
Workload is expected to be more important for individuals with low levels of OBSE. Again, the same reasoning as for non-financial intrinsic rewards applies: They generally perceive themselves as not keeping up with the expectations and thus high workload can make them even more unsecure. Thus, they try to avoid increased workload. In contrast, employees with a high level of OBSE feel confident enough to fulfill more work at the same time and therefore workload might be of minor importance to them.
Training and development, resources support and feedback are also expected to be of greater preference for individuals with low levels of OBSE. These kind of rewards help the individuals to grow in their job and thus to feel more competent. In contrast, individuals with a high level of OBSE are less dependent on support by superiors or resources and thus will place less value on these reward types.
Similar relationships are expected for the preference for work-life balance, while contrary results are predicted for work environment and team spirit. As reasoned above, low OBSE employees will tend to define themselves more over other competence fields and thus work-life balance is of great importance for employees possessing low levels of OBSE. In contrast, individuals with a high level of OBSE place greater importance on their job and thus working environment and team spirit might be of greater importance.
No relationship is predicted for fair treatment, as the perception of fairness is not likely to differ dependent on individuals’ level of OBSE.
Most relationships expected for support rewards are negative and thus it is expected that individuals with lower OBSE have a greater preference for support rewards than their higher OBSE counterparts.
Hypothesis 2: The higher the level of individuals’ OBSE, the lower their preference for support rewards.
3.2.2 Power Rewards
Power rewards consists out of the following individual reward items: authority and power, promotion possibilities, title and status, decision-making possibilities and management style (Chiang & Birtch, 2007). Individuals with higher OBSE are expected to prefer power rewards to a greater extent than low OBSE individuals. This is because they perceive themselves as being “worth” to have a high position, status or a wider range of responsibilities and furthermore these reward types allow employees to demonstrate their competencies to others. Individuals with lower-levels of self-esteem are not concerned about demonstrating their level of self-esteem to their work environment and thus are expected to prefer power rewards to a smaller extent. In addition, more decision making possibilities might rather be perceived as a burden than a benefit. No relationship is expected between OBSE and management style.
Hypothesis 3: The higher the level of individuals’ OBSE, the greater their preference for power rewards.
3.2.3 Relationships
A slightly different reasoning is applied for developing hypothesis for relationship rewards. In addition to the self-enhancement theory introduced in section 2.3.3, behavioral plasticity theory of Brockner (1988) is used. He examines the “extent to which an individual is affected by external factors” (p.596, Pierce & Gardner, 2004). According to Brockner (1988) low self-esteem individuals are generally affected by external factors to a greater extent, as they are more uncertain about their work and thus will rely more on cues that guide them. “They seek acceptance and approval from others through conforming attitudinal and behavioral acts” (p.596, Pierce & Gardner, 2004). Thus, I expect relationships with superiors to be of greater importance for employees with a low level of OBSE in comparison to individuals with higher OBSE.
Hypothesis 4: The higher the level of individuals’ OBSE, the lower their preference for a good relationship to their superior.
3.3 Preferences for Financial Rewards
Basic salary and annual salary increase are expected to be equally important for both groups. Both, low and high OBSE individuals have a tendency to increase their level of self-esteem. Salary can serve this purpose, as it signals worthiness to the employees. However, it does not come at any cost for low OBSE individuals because it is not tied to performance and therefore cannot possibly enhance their negative self-image due to failures. Thus, no difference is expected between the preferences of both groups.
However, differences are expected for variable financial rewards. Individual pay incentives are expected to be of greater importance for individuals with a high level of OBSE, as employees with lower levels generally fear failure and thus variable pay (dependent on performance) comes at a risk.
In contrast, employees with a high level of OBSE are expected to show less preference for team performance incentives. The underlying reasoning is that they believe their own performance to be above average and therefore they might expect that the overall team evaluation is lower than their individual one.
Hypothesis 5a: The higher the level of individuals’ OBSE the greater their preference for individual performance incentives.
Hypothesis 5b: The higher the level of individuals’ OBSE the lower their preference for team performance incentives.
3.4 Preferences and Actual Rewards Provided
Within this study the reward preferences of employees and the actual rewards they receive are compared. The relationship between actual rewards and preferences of individuals is expected to differ dependent on the level of OBSE of individuals. High self-esteem individuals perceive themselves as valuable and important for the organization they work for. Thus, they might be in a stronger position to negotiate their preferences. Following, the rewards they receive might be relatively similar to their preferences. In contrast, employees with lower levels of OBSE do not believe themselves to have any value for the organization and therefore might not dare to negotiate their reward structure that much. Thus, the relationship between actual rewards received and their preferences might be less strong.
However, it has to be noted that in larger organizations certain policies might prevent the possibility of reward negotiation.
Hypothesis 6: The higher the level of individuals’ OBSE, the higher the perceived similarity between actual rewards and preferences.
The following table provides a summary of the hypothesis developed in this section.
Figure 4: Overview of the Hypotheses
Expected Sign
Reward Category Sub Category
H1 OBSE Non-financial intrinsic All non-financial intrinsic +
H2 OBSE Non-financial extrinsic Support Rewards
-H3 OBSE Non-financial extrinsic Power Rewards +
H4 OBSE Non-financial extrinsic Relationship (Superior)
-H5a OBSE Financial rewards Individual performance incentives +
H5b OBSE Financial rewards Team performance incentives
-H6 OBSE +
Dependent Variable Independent Variable
4 Research Design and Methodology
This section will provide further information on the research method applied. For the research question at hand a quantitative research approach is chosen. Specifically, the survey method is used as it suits best to answer the research question. Following, the sample collection and method of data collection are depicted in great detail. The variable constructs used within this research study are introduced next. For each construct a factor analysis and the Cronbach Alpha is provided. Lastly, some further statistical analysis of the data is provided in order to reveal information on the quality of the data obtained.
4.1 The Survey Method
The research question is observed by using a survey method. Cresswell (2003) notes the importance of a “match between the problem and approach” chosen (p. 21), as different research problems require different methods. The purpose of the current study is to understand the influence of OBSE on the preferences for certain rewards. Thus, the study aims to understand the predictors of an outcome for which a quantitative approach suits best (Cresswell, 2003; Field, 2009). Among quantitative research methods a survey method is chosen, as it is able to uncover specific respondents’ characteristics and personality traits which cannot be obtained by public databases. In addition, a survey offers further advantages such as flexibility, specialization and efficiency (Alrech & Settle, 1995).
4.1.1 Sample Collection
The survey method requires the collection of a sample out of the whole population. A sample can be characterized as representative, if the characteristics of the population are reflected in the sample in same proportions. However as research notes, this condition is difficult and seldom fulfilled in practice (Blaikie, 2009). Nevertheless, this study aims to be representative for employees working in the for-profit sector by including respondents from a wide range of educational backgrounds employed in different industrial sectors. In addition, the sample is diverse in regard to years of work-experience, gender and age.
Selection criteria have to be identified by the researcher to serve as rejection tool in order to identify a target population (Blaikie, 2009). A criterion chosen for this research study is the extent of organizational experience of participants. Participants with less than six month experience within an organizational setting are excluded. It is assumed that at least half a year experience is necessary in order to evaluate and identify one’s own preferences. In addition, only
employees working in the for-profit sector are included, because the non-profit and public sectors differ substantially from the for-profit sector, especially in regard to the motivation of employees.
The sample size refers to the number of participants included in a research study (Bajpai, 2011). The aim of the study was to collect around 100 participants. I collected 130 responses, with a dropout rate of 16.15%, which leaves a final sample of N=109.
Furthermore, the sampling technique chosen is crucial for the quality of a study. Research distinguishes between random and non-random sampling. In random sampling, each member of the population has the same probability of being selected, while this is not the case for non-random sampling (Bajpai, 2011). Due to feasibility issues a non-non-random sampling technique is chosen. Specifically, convenience sampling is used. Thus, the participants for the study are chosen based on the convenience of the researcher itself (Bajpai, 2011).
4.1.2 Method of Data Collection
There are several different methods of data collection such as a face-to-face method, mechanical method, mail, telephone or internet surveys (Bajpai, 2011; Dillman, 2008). Methods of data collection often are a source of error and all different methods have their advantages and disadvantages. For this master thesis self-administered internet surveys are used. Questions appear on the computer screen of participants and can be answered without the presence of the interviewer (Dillman, 2008). An advantage of self-completed questionnaires is clearly that the results of participants are not affected by the variability of an interviewer. In addition, they are convenient for the respondents themselves. However, self-completed questionnaires also contain some disadvantages, for instance respondents might experience difficulties in answering questions due to ambiguities. For this reason, the survey was tested with a psychologist and communication specialist to ensure that questions are as clear as possible. (Bryman, 2012)
4.2 Constructs
In order to measure the underlying phenomena already existing scales developed in research are used. The reward scale of Chiang and Birtch (2007) is used to measure the reward preferences of individuals, while the scales of Rosenberg (1965) and Pierce et al. (1989) measure the GSE and OBSE of individuals. Furthermore, three control variables are implemented in the study. The overall construct scores are calculated based on the average scores obtained on the individual items as this is the method most commonly used in research (Field, 2009).
A principal components factor analysis (PCA) is conducted on both the dependent and independent variables in order to measure the validity of the scales. Direct oblimin rotation is chosen as a rotation method as data is expected to be correlated and thus an oblique method reveals more accurate information (Costello & Osborne, 2005). Furthermore, the Cronbach Alpha is assessed in order to measure the internal reliability of the scales (Field, 2009).
4.2.1 Reward Preferences
The reward preferences of participants are measured according to a reward preference questionnaire developed by Chiang and Birtch (2007). Chiang and Birtch (2007) developed their scale with referring to primary works such as Maslow (1954) or Herzberg et al. (1959). I specifically chose the scale of Chiang and Birtch (2007) as it contains a wide range of financial and non-financial rewards. The scale has also been subsequently tested and used by other researcher. However, some single items were slightly adjusted in order for participants to better understand them after testing the survey with the communication specialist. For instance, “individual performance incentives” is labeled “individual pay for performance”, but these syntactical changes never changed the meaning of the item as such. The survey is conducted in two languages: German and English. However, there exists no German translation of the reward scale and thus I translated the reward items by myself. To deliver accuracy, the German version of the survey was again tested by the communication specialist to ensure that there are no differences between both versions. Appendix A and B contain the English and German version of the survey, while Appendix C provides a full overview of the translation and alterations of the reward preference construct. Single items are measured on a seven point Likert scale developed by Rensis Likert (Bajpai, 2011; Bryman, 2012). Hence, respondents have to indicate the level of importance of each item, ranging from “extremely unimportant” to “extremely important”. A PCA was conducted on the 19 individual reward items of the Chiang and Birtch Scale (2007) to investigate whether the categories such as non-financial intrinsic rewards in fact explain the same underlying phenomenon. The Bartlett’s test of sphericity is significant and thus indicates that the correlations between individual items are large enough to run the PCA (p<0.001). The Kaiser-Meyer-Olkin (KMO) measure reveals the adequacy of the sample size. The KMO value (KMO=0.845) is above the acceptable limit of 0.5 and values of KMO which exceed 0.8 are considered as being “great” in research (Field, 2009). In addition, all KMO values of the individual items exceed the threshold of 0.5.
The analysis revealed that the items load on three different factors, based on an Eigenvalue greater than 1. This matches the categorization of Chiang and Birtch in three different categories: non-financial intrinsic rewards, power rewards and support rewards. Together, the factors explain 57.6 % of the variance. Factor 1 explains 34.9 % of the variance, while factor 2 and 3 explain 14.86% and 7.83% respectively.
The rotated component matrix provides the factor loadings of the individual items. All items of the non-financial intrinsic rewards load sufficiently on factor 1, as all values exceed the threshold of 0.5 (Field, 2009). However, while items (3), (4), (5), (6) and (8) of the support rewards load sufficiently on factor 2, items (1), (2) and (7) do not. Thus, it seems like team spirit, fair treatment and work environment do not fit the category of support rewards perfectly as they are explained by another underlying item. In addition, the items of the power rewards seem to load on the same component than the non-financial intrinsic rewards. Therefore, power rewards seem to not discriminate from non-financial intrinsic rewards (Shiu et al., 2011) and the preferences for these two reward types overlap. This however does not cause problems of multicollinearity as both reward categories are tested independently in a separate analysis as dependent variable. Categories are retained as proposed by Chiang and Birtch (2007) as the scale has successfully been tested and used in research. Appendix D provides an overview of the individual factor loadings retained from the PCA.
In addition, the Cronbach Alphas of the reward categories are analyzed. The Cronbach Alpha of the non-financial intrinsic reward category is 0.847, while support and power rewards have an alpha of 0.765 and 0.724 respectively. Thus, all categories exceed the desired threshold of 0.7 (Field, 2009) and therefore all individual reward items are kept for the further analysis.
4.2.2 Organization-based Self-esteem and General Self-esteem
The scale of Rosenberg (1965) is used for an assessment of the GSE of respondents. The scale consists out of six different items, for instance “I am able to do things as well as most other people”, “I certainly feel useless at times” or “On the whole, I am satisfied with myself” (Rosenberg, 1965). The scale has been empirically tested and approved by relating research (Eisenberg et al., 2015). Respondents have to indicate their level of agreement with each item, ranging from “strongly agree” to “strongly disagree” (Bajpai, 2011). For the German version of the survey, the translation of the scale by Herzberg and Collani (2013) is applied. Three of the items of the GSE scale are negatively worded and thus are recoded (“At times I think I am no good at all”, “I feel I do not have much to be proud of” and “I certainly feel useless at times”).
Pierce et al. (1989) have been the first to develop a scale for measuring self-esteem specifically in an organizational setting. The scale was derived from comments made by organizational members within their specific organizational setting. Each item measures the extent to which employees belief they are “valuable, worthwhile, [and] effectual members of their employing organization” (p.634, Pierce et al., 1989). Items include “I count around here”, “I am taken seriously around here”, “I am important around here”, “There is faith in me around here” or “I can make a difference around here” (p.634). The scale of Pierce et al. (1989) was subsequently tested in seven different studies and has also been used by other researchers (see for instance Liu et al., 2013). The translation of Kanning and Hill (2012) is used for the German version of the survey. Again, the single items are measured on a seven point Likert scale ranging from “strongly agree” to “strongly disagree”.
The PCA is conducted on 17 items of the OBSE and GSE scale. Again, the Bartlett’s test of sphericity is significant, the KMO values are above the acceptable limit (KMO=0.888), and the KMO values for all individual reward items exceed the threshold of 0.5.
The PCA of the 17 items reveals three factors based on the criterion of an eigenvalue greater than 1 (Field, 2009). The factors together explain 66.17% of the variance. The first component explains 48.97% of the variance, while the second component explains 10.32%. All items of the GSE load sufficiently on factor 2 with values above 0.5. However, while item (1), (2) and (8) of the OBSE scale load on factor 3, items (3), (4), (5) and (7) load on factor 1. Items (6), (9) and (10) seem to load equally on both factors and thus their values do not exceed the threshold of 0.5. Thus, the individual items of the OBSE scale seem to be explanatory for two underlying phenomenon. However, as factor 3 does only explain a small percentage of the variance and the main aim of the factor analysis was to verify that the different items used in this study indeed reflect different factors, factor 3 was not retained for the further analysis. Therefore, the PCA was rerun with limiting the number of factors to the number of constructs. In this second stage of the PCA items (8) and (10) are slightly below the threshold of 0.5 (0.469 and 0.477), while all other items exceed the threshold. However, items (8) and (10) are kept for the further analysis and are not deleted. The reason is that both items are only slightly lower than the threshold and the scale has been proven to be valid in research (Pierce et al., 1989). Appendix E and F provide an overview of the individual factor loadings retained from the PCA in both stages.
The Cronbach Alphas of the GSE and OBSE scales are 0.873 and 0.913 respectively, which are above the threshold of 0.7. Further analysis revealed that the Cronbach Alphas of both scales
could not have been improved by deleting any item of the scale. Due to this reason and a strong Cronbach Alpha in general all items are retained for the further analysis.
4.2.3 Control Variables
In order to assure that demographic variables such as age or gender do not bias the results, control variables are implemented in the model. Based on a correlation analysis the demographic variables which seemed to influence reward preferences most have been selected as control variables.
For instance, age is negatively correlated (p<0.01) to annual salary increase, individual pay for performance, promotion possibilities, title and status, training and development, job challenge and skills-used-opportunities.
Thus, age seems to be a predictor of reward preferences and is therefore implemented in the model. Next to age, gender and years of work-experience are selected as control variables because according to prior research they are also determinants of reward preferences (Kominis & Emmanuel, 2005).
4.3 Further Statistical Analysis of the Data
In addition to calculating the Cronbach Alpha and running the PCA, an early-late response bias test was conducted to ensure that late respondents do not significantly differ from early respondents.
As data was first collected on the English version of the survey and in a second step on the German version, the early-late respondents test was only conducted for participants who filled in the English version of the questionnaire, as these are the majority of the overall sample. A non-parametric test, the Man-Whitney test, is chosen for the comparison of both groups as data is ordinal and shows slight deviations from a normal distribution (Field, 2009). For instance, the variable GSE seems to significantly deviate from a normal distribution (SK=-5.41; K=2.102). The Man-Whitney test reveals that there are no significant differences between early and late respondents (p>0.05) for GSE, OBSE and all non-financial and financial reward categories.
5 Results
In the following the main results of the research study are depicted in detail. Firstly, demographics and descriptive statistics of the sample are introduced in order to provide an overall picture of the obtained sample. Afterwards, the hypotheses are tested by running a regression model.
5.1 Sample Descriptives
Figure 5 provides an overview of the most important sample characteristics.
Figure 5: Overview of the Demographics of the Sample
Demographic variable Sub-category
n % Gender Male 53 48.6 Female 56 51.4 Age (years) 18-25 58 53.2 26-35 27 24.8 36-45 10 9.2 46-59 14 12.8 60+ 0 0
Industry / Sector Energy 3 2.8
Construction 15 13.8 Logistics 5 4.6 Banking / Finance 16 14.7 Retail 7 6.4 Healthcare 10 9.2 Media 5 4.6 Telecommunications 1 0.9
Travel and leisure 4 3.7
Professional services 21 19.3 Others 22 20.2 Department Administrative 7 6.4 Human resources 3 2.8 Manufacturing 6 5.5 IT 2 1.8 Consulting 10 9.2 Accounting / Finance 22 20.2 Marketing / Communication 16 14.7 Engineering / Technical 8 7.3 R&D 5 4.6 Others 30 27.5 Total
The demographics of the sample are important as the sample aims to be representative for employees working in the for-profit sector. In order to be representative a wide range of employees, which differ in regard to their demographic variables, have been included. The sample is not only diverse in regard to gender (48.6% male and 51.4% female), but also in regard to age (ranging from 18 to 59), industry (includes employees working in at least 11 different industries) and departments (includes employees working in at least 10 different departments). Furthermore, respondents are diverse in regard to nationality (employees from 11 different countries are included), educational background (ranging from High School to Doctorate), job level (ranging from intern to executive), degree of oversight responsibilities (ranging from no oversight responsibility to oversight responsibility over more than 15 employees) and length of work-experience (ranging from half a year to over 20 years of working experience). Appendix G provides a detailed overview of all the demographic variables.
In addition, the following table provides mean, standard deviation, minimum and maximum for the main variables investigated in this research study.
Figure 6: Descriptive Statistics of the Most Important Variables Investigated
Measured on a seven-point Likert scale, the average level of OBSE is 5.52 (SD=0.85), while the average level of GSE is 5.51 (SD=1.01). Thus, the general level of OBSE and GSE of individuals tends to be similar, which will also be revealed by a further analysis of the correlation between both constructs. Most participants at least “to some extent” agree that they are valued and worthwhile in a personal as well as organizational context.
Mean and standard deviation of the scale developed by Chiang and Birtch (2007) provide some valuable insides on which rewards, both financial and non-financial, are most preferred by employees. Relationship to one’s superior (M=6.01, SD=0.88) are rated as being most important, followed by support rewards (M=5.67; SD=0.72) and non-financial intrinsic rewards (M=5.64; SD=0.83). These results clearly indicate the importance employees place on non-financial
Minimum Maximum Mean St. Deviation
OBSE Scale 3.3 7 5.52 0.85
GSE Scale 1.33 6.83 5.51 1.01
Non-financial intrinsic rewards 3 7 5.64 0.83
Support rewards 3.57 7 5.67 0.72
Power Rewards 2.2 6.8 5.18 0.95
Relationships (Superior) 3 7 6.01 0.88
Individual performance incentives 1 7 4.88 1.6
Team performance incentives 1 7 4.23 1.56
Similarity actual rewards vs. preferences 1 7 4.47 1.51
