Employee’s willingness to cooperate.
The effect of (Non) – Monetary incentives, and Personal Values. Behavioral Foundations of strategy
January 18, 2016 Master Thesis
MSc Business Administration Faculty of Economics and Business
University of Amsterdam Supervisors: Dr. J.W. Stoelhorst Dr. F. Bridoux Author: Panagiotis Zogopoulos 10389725
2
This document is written by Student Panagiotis Zogopoulos who declares to take full
responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources
other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion
3
Abstract
Designing optimal internal reward mechanisms within an organization is of high importance because it can lead to better performance for the firm by influencing behaviors at the individual level.
This study aims to shed light in the relationships between monetary/nonmonetary rewards and employee willingness to cooperate. This thesis is conceptualizing an online research, combination of vignette and survey technics to investigate observations of these relationships and the influence of employees’ personal values among other heterogeneous characteristics. The results of the analyses add to existing studies that solely choice between monetary and nonmonetary rewards will not necessary affect employee’s willingness to cooperate and may even disrupt it by interfering with intrinsic and extrinsic types of behavior. Moreover, this study empirically presents that personal values (Self-transcendence & self-enhancement) and heterogeneity, trigger various behaviors regarding cooperation. In case of self-transcendence specifically, it was found that positively relates to willingness to cooperate independently of type or reward in whereas self-enhancement did not influence these relationships.
Keywords: willingness to cooperate; monetary rewards; nonmonetary rewards; heterogeneity; personal values; self-enhancement; self-transcendence.
4 Table of Contents Abstract ... 3 1. Introduction ... 5 2. Literature Review ... 9 2.1 Cooperation ... 9 2.2 Motivation theory ... 10 2.2.1 Intrinsic motivation ... 12 2.2.2 Extrinsic motivation ... 12
2.3 Monetary - Non Monetary rewards ... 13
2.3.1 Monetary Rewards ... 13
2.3.2 Non- Monetary Rewards ... 15
3. Theoretical Framework ... 18
3.1 Monetary & non-monetary rewards and willingness to cooperate ... 19
3.2 Social values as moderators. ... 20
3.3 Gender & education as moderators ... 21
4. Methodology & research design ... 22
4.1 Method ... 22
4.2 Sample ... 23
4.3 Design ... 23
4.3.1 Dependent and independent variables. ... 23
4.3.2 Social values as moderators ... 26
4.3.3 Age as control variable ... 27
4.4 Analysis ... 27 5. Results ... 28 6. Discussion ... 49 6.1. Main findings ... 49 7. Conclusion ... 56 References ... 57 Appendices ... 64 Appendix I: ... 64 Appendix II: ... 67
5
1. Introduction
“There are only three measurements that tell you nearly everything you need to know about your organization’s overall performance: employee engagement, customer satisfaction, and cash flow. It goes without saying that no company, small or large, can
win over the long run without energized employees who believe in the mission and understand how to achieve it.” – Jack Welch, former CEO of GE
Resilience is what companies are aiming for. In other words, to create and sustain a competitive advantage over time, which will determine the success or failure of a firm based on the very same reason it exists, competition (Porter, 1985). The credits of creating this advantage over time are being attributed, by many scholars, to firms’ resources (Barney, 1991; Peteraf, 1993). Coff (1997, 1999), makes a deeper analysis on the resources’ nature by distinguishing them in inanimate and animate assets-humans. Considering the lack of ‘free will’ in inanimate resources, we can easily rule out any kind of objection against those who own or wish to exploit these resources, in order to gain from their performance. This is not the case though when we talk of animate assets, meaning humans. Based on the high complexity of human nature, and on the numerous variables that determine their behavior, the adherence to firm’s norms and desired outcomes is highly dependent on their cooperation. However, being motive-related, humans’ willingness to cooperate should not be taken for granted.
Despite Coff’s (1997) recognition that employees are not like all other resources, it is only recently that strategy researchers have started paying attention to the need of understanding what individuals are ‘made of’. “Individuals matter and micro-foundations are needed for explanation in strategic organization” (Felin & Foss, 2005,). Felin & Foss (2005) argue that for an in depth organizational analysis of identity, capabilities or knowledge, one must begin the analysis with the individual actor under focus, and research the nature of this individual, his
6
choices, purposes, expectations, motivations, as well as heterogeneity across individuals. A view of Strategic Organization without the proper attention to individuals and their unique characteristics would certainly be less accurate, lacking the fundamentally understanding of the main element of organizations. Even if firms are considered as entities, their characteristics and routines emerge from the individual members they employ. Thus, in recent strategy research, individuals considered as resources have come to centre stage and the need for micro-foundations analysis has become clear (Foss, 2011).
Scholars in the strategy field have linked firms’ performance to the level of cooperation between employees, estimating that the higher the level of cooperation the better the performance (Kosfeld & Siemens, 2011). Cooperation can be defined as an organized collective effort from individuals who want to achieve a common goal (Pinto & Prescott, 1993). Achieving business goals would be really difficult, if employees could not coordinate their efforts and knowledge for the greater good and instead were aiming at selfish goals (Kollock, 1998). If an employee acts under no control or guidance, then he develops the trend to behave selfishly, since he will not be accountable for his actions (Kollock, 1998). This is the reason why managers take under serious consideration the way they control, guide, coordinate and motivate the behavior of their employees (Mitchell, 1982). The task to inspire and maintain team cooperation has been always a responsibility for firms’ management (Tabibnia & Lieberman, 2007).
When strategy and business plans or major decisions related to firms’ future are under consideration, we tend to focus on the potential outcomes of these decisions and we evaluate them according to success or failure (Eisenhardt, 1999). These actions though, are the result of individual thinking. The analysis of this thinking process depends on several variables, as usually happens when human beings become the subject of research. Communications between
7
employees, routines that are created through time, norms of behavior and knowledge articulation within a firm are what determine its performance (Eisenhardt & Martin, 2000).
The uniqueness of every individual regarding their personality, characteristics and values requires a better understanding for controlling their behavior. Managers use various systems to control employees’ performances and motivate them to cooperate under the firm’s main goals (Christensen et.al, 2006). Frequently used motives fall under the categories of intrinsic or extrinsic, with the latter further categorized to monetary and nonmonetary rewards (Luthans & Stajkovic, 1999). The choice of each motivational strategy would have a significant effect on the employee’s behavior and especially at their willingness to cooperate under interdependent tasks (Coff, 1997).
In this research I will focus mainly on the extrinsic motivational strategies and especially on the different effects that monetary rewards have in contrast to the non-monetary as an alternative choice of rewards. Moreover, I will examine the individuals in terms of personal characteristics, such as social values and demographic characteristics, and how this heterogeneity can influence the relationship between motivators and cooperation intentions.
Providing a deeper analysis on micro-foundations in strategic organization, and detailing out the key elements in the direct and moderating relationships among the variables, I aim to further support management with the know-how in conceiving and executing strategies, leveraging the results and achieving sustainable outperformance.
The structure of this research is based on the following sequence. After the Introduction, the literature review follows (chapter 2) where I present an overview of the existing relevant research on individual’s motivation, previous findings in the field of cooperation, and supporting
8
evidence from research on personal values. In chapter 3 I present the theoretical framework of my research and the connections among my variables that will be tested. In chapter 4 and 5, I present the research design, the methods of the research and its results. At the end of this paper you can find the discussion and conclusion of the research, respectively, in chapter 6 and 7.
9
2. Literature Review
“A group of people get together and exist as an institution we call a company so they are able to accomplish something collectively that they could not accomplish separately – they make a contribution to society, a phrase which sounds trite but is fundamental.” – David Packard, late
co-founder of Hewlett-Packard 2.1 Cooperation
Following the logic of the statement above but also adhering to natural rules of existence, people are the most cooperative life form, “From hunter-gatherer societies to nation-states, cooperation is the decisive organizing principle of human society” (Nowak, 2006, p. 1560). Moving from the biological identity of humans towards an organizational-corporational environment, individuals are defined, or at least are not excluded, as resources to be exploited for the firm’s benefit (Barney & Clark, 2007). According to resource-based theory, the ability of a firm to gain a superior advantage lays on its resources performance and ‘co-operation’ (Barney, 1991). Considering an organization without any human resources, it could be easily assumed that this firm would exploit 100% of its asset’ performance, without any resistance (Alchian & Demsetz, 1972). On the other hand though, when human resources join in, the level of exploitation by the firm cannot be guaranteed at the same levels (Kim & Mahoney, 2007). Due to the fact that previous scholars, focusing on resource-based theories, did not investigate deeper in the human factor, a better understanding of how individuals can contribute to the scenery, has not been provided yet (Foss, 2003).
Cooperation is an essential factor for productivity inside a firm. The development of specific routines and norms between employees can increase productivity and underlies the difference between a successful and not successful company. Furthermore, if this ability to cooperate in a productive way can be sustained through time, it will consist part of the firm’s dynamic capabilities (Eisenhardt & Martin, 2000). Individuals differ by nature in every aspect
10
that concerns them (Schwartz 2004). Every person has their own way of thinking, their special skills, unique characteristics and personality, as well as diverse ways to perceive the changes in their environment and, ultimately, choose the best action to respond (Maslow 1943). This level of diversity among employees affects their willingness to cooperate in tasks, where teamwork is a prerequisite (Felin & Foss, 2005). That is why a mix of motivational strategies should be carefully chosen according to the character of each employee, aiming to a higher level of cooperation between them (Bridoux et al., 2011).
However, to the same stream of literature, further research is needed, in order to understand better when, why and which motivational strategy should be chosen, including factors as the environment of the employees, and most importantly for this research, their personality.
2.2 Motivation theory
“Managers can use a variety of carrots and sticks to
encourage people to work together and accomplish change. Their ability to get results depends on selecting
tools that match the circumstances they face”(Christensen et.tal., 2006)
Indeed, trying to influence people to outperform themselves is a challenging task assigned to managers, and the effort to crack-the-knot on how to do so or what is exactly ‘it’ that motivates individuals, “is a centuries-old puzzle” (Nohria, Groysberg, Lee, 2008, p1). Although we have a significant fundamental theory on the reasons of our behavior, mainly due to contribution of science ‘fathers’ like Aristotle, A. Smith, S. Freud, A. Maslow, the conclusions were the outcome of direct observation and lacking the advantages of modern scientific fields (neuroscience, psychology, etc.), insights and available data (Nohria, Groysberg, Lee, 2008). Still though, the complexity of human behavior requests a similarly complex and in depth
11
analysis of what managerial systems and to whom should be applied, in order to drive a firm’s performance (Nohria, Groysberg, Lee, 2008).
People have the tendency to be motivated by different things or have different expectations from a potential job (Tabibnia & Lieberman, 2007). Even the same individual can present different behavior based on their age, level of education or current status, and different approaches to what is important at a certain point or not based on their values (Schwartz, 2004). Individual pursuit of goals is expressed through personal values and affects the coordination with others in related teams (Schwartz 1995). Managers have always tried to control the behavior of their employees in order to guide the collective effort to achieve the firms’ goals (Nohria, Groysberg, Lee, 2008). That is why knowledge about their employees is very important, and through time there have been various theories regarding how managers can encourage employees to pursue the firm’s goals. Performance evaluation and minimizing the divergence of preferences among organizational members have been suggested as the most effective control strategies (Ouchi, 1979).
Nohria et.al (2008), in their research, suggest that there are four instinctive needs that guide humans behavior, ‘[t]he drives to acquire (obtain scarce goods, including intangibles such as social status); bond (form connections with individuals and groups); comprehend (satisfy our curiosity and master the world around us); and defend (protect against external threats and promote justice)’ (Nohria, Groysberg, Lee, 2008, p1). Further in their model, Nohria et all. 2008 support that there are two main systems that organizations can use to cover employee’s needs, the reward system and the (company) culture. This distinction of systems, in accordance with the existing literature, and the categorization of motives into external (extrinsic) and internal (intrinsic), where extrinsic motivation is when the outcome of an action leads to a distinct
12
reward, whereas in intrinsic motivation performing the action itself can lead to a rewarding feeling (Deci and Ryan, 1985).
2.2.1 Intrinsic motivation
Here motivation comes from inside the individual and refers to a task that will give pleasure to the employee by actually performing that task (Deci 2000). The intrinsic motives are in the mind of every individual and they vary depending on their personality (Deci 2000). An intrinsically motivated person will work hard just for the satisfaction of fulfilling the given task instead of awaiting a reward (Deci 2000). Psychological research has shown that intrinsic motivations are related to the need to feel component and self-determining (Deci 1985). Individuals will act for the good feeling delivered by their task and the joy of completing it. (Deci 1985)
2.2.2 Extrinsic motivation
The rewards used for this kind of motivation are usually tangible like money, bonus etc. and come from outside the individual. This kind of rewards is to provide the required satisfaction to the employee in order to continue performing the same tasks even if he finds no interest in this (Deci, 2000). A person motivated by extrinsic rewards will work hard on possibly unpleasant task anticipating the promised reward. The reward itself can be anything from a single demonstration of approval by the employer to a big bonus, either way it depends on the character of the individual. (Deci &Ryan, 2000)
Agency theory argues that the principal’s (manager’s) task is to engage the agent (employee) to projects necessary for the firm’s business, to develop control mechanisms to measure the agent’s performance and to provide the right motives or punishments that will convince the agent to fulfill his task, which in turn will contribute to the collective goal of the
13
firm (Eisenhardt, 1985). However, motivating multiple agents to cooperate is not an easy task as they have to deal with diverse and complex interactions (Ostrom, 2000). Thus, the call for further analysis in the extrinsic motives to cooperate and the need for in-more-depth look in micro-foundations (Felin & Hesterly, 2007).
2.3 Monetary - Non Monetary rewards
Recent scholars have discussed that individuals’ behavior can be driven from a variety of motives (Gottschalg & Zollo, 2007). The level of cooperation in an organization is linked to the individuals’ motives and the motivational strategy, the system which is used by the management as a reward for employees’ contribution to achieving the firm’s common goals (Mitchell, 1982). 2.3.1 Monetary Rewards
“When people are financially invested, they want a return… (Simon Sinek)
Human beings have always been motivated by money as reward in a variety of situations, like a child getting money to do some chores, a customer gaining a cash-back offer for his purchase or an employee rewarded with a raise or a bonus for good performance (Tabibnia & Lieberman, 2007). Indeed, there is a strong relationship among monetary rewards, employee motivation and other positive (desired) results (Jewell & Jewell, 1987). There is empirical evidence of different levels of motivation between rewarded and unrewarded employees, (Stajkovic & Luthans, 2001), along with multinational presence and application of that (motivational) system (Du & Choi, 2010, Cadsby et al, 2007, Campbell et al, 2009).
Monetary rewards come in many forms, salaries, bonuses, equity, long-term earnings etc. (Aguinis, 2013), each of them with a strong impact on the individuals’ behavior, according to scholars defending these motives (Locke et al, 1980), with practical results (30% average increase) in employee’s contribution (Aguinis, 1980 ; Locke, Feren, McCaleb, Shaw & Denny,
14
1980). Among other things, monetary rewards, such as (greater) salary, can determine the acceptance, or not, of a vacancy (Feldman & Arnold, 1978), can provoke the response, and eventually retention, of top-end candidates (Rynes et al, 2004), can provide great returns on (resources) investments, (Brown et al, 2003), can fulfill primary (food, home etc.) but secondary also (social acceptance, feeling of belonging etc.) human needs (Long & Shields, 2010), and can also be perceived as a symbolic representation of the individual’s success (Trank et al, 2002).
Someone could probably argue, based on the evidence so far, that monetary rewards, as management’s choice for motivation strategy, would most likely be the best-in-class solution, applicable in most (if not all) situations and bringing the desired results. However, this would be far from the truth, as there is the dark-side of this choice, where monetary rewards unveil their flaws, lead to occasions with opposite than the desired results, and apparently, according to employee surveys, are not chosen as the most important motivation trigger (Aguinis, 2013).
Beer & Cannon, (2004) say that rewards of significant monetary value don’t always achieve the expected motivational boost, and Harris & Bromiley (2007) come to add that (generous) rewards can even lead to decrease of productivity. Chib et al. (2012) relate this downfall to the enhanced employee’s fear of not ‘delivering’ under the (over)pressure feeling of those generous rewards. On the other side of this coin, employees may get used to a certain level (amount) of rewards, and when this is not met, they could start slip in their behavior or even leave the firm (Schaubroek et al, 2008).
Although monetary rewards can cause increase in employee motivation , do not necessarily improve and develop employees and as a result the company itself (Dierdorff &Surface, 2008). Relying on a single-style method of rewards, in this case monetary is not wise and not effective either (Mitchell, 1997). Thus, it is of vital importance to better understand and
15
examine which rewards systems to use, in what degree, when to use them and most importantly to whom (personality, values, individual etc.) (Aguinis et al. 2013).
2.3.2 Non- Monetary Rewards
…when people are emotionally invested, they want to contribute.” - Simon Sinek
Recently attention has turned to non-monetary incentives with studies like Masclet et al (2002) reporting that non-monetary motives have a positive impact on cooperation. Monetary incentives do not have the ability to link workers with the organization, they cannot be shared by all the employees and they create conflicts of interest, because, for example, the more money the employees receive the less is left for the organization and vice versa (Sorauren, 2000). Non-monetary incentives respond to higher psychological needs, such as the needs for self-actualization and esteem, according to Maslow’s (1943) hierarchy. Studies have thoroughly covered the effect that nonmonetary rewards have on individuals’ behavior, although personality as a variable has been neglected it from scholars and attracted less analysis in empirical level (Good 1972).
According to Dugar (2007), types of non-monetary rewards (i.e. social recognition) can be effective as a motivation to cooperate and also have been found to be effective in infinitely repeated interactions (Rand et al. 2009). In addition, Hollander (1990) provided an early influential model with contributions under peer-to-peer approval showing cooperation with positive contributions can exist. Such non-economic motives have recently started to be identified as important factors to be examined in order to better understand the cooperation process (De Cremer, 2003, Milinski et al, 2002).
Individuals come into an employment relationship with their organization believing that their efforts are going to be rewarded by their employer and that there exists a mutual obligation
16
between the employee and the employer.(Kosfeld & von Siemens, 2011) Besides the formal, written contract that is agreed upon once the person joins an organization’s working force and that explicitly states the rules of employment, perhaps whar is more important is the informal, unwritten, psychological contract that develops constantly (Tabibnia & Lieberman, 2007). The psychological contract is characterized by the employee’s subjective perspective of what they are promised to receive from the organization in exchange for their work behavior (Rousseau, 1989). The employees provide the organization with their skills, time and effort and expect from the organization in return to help them fulfill their needs and achieve their goals. The psychological contract is, thus, created on the part of the individual, not the organization (McDonald and Makim, 2000). According to exchange theory, when employees feel that their organization meets their expectations and fulfills their working needs they respond with increased commitment, which in turn leads to more rewards, resulting in a cycle beneficial to both employees and the organization (Malhotra, Budhwar and Prowse, 2007). Furthermore, the way the employer responds to these expectations has been found to affect the outcomes of job satisfaction and work performance (Robinson, 1996).
Considering all the above studies we can highlight the importance of non-monetary incentives in fostering cooperation and furthermore, the importance of individuality as a (non)-success factor. Hence, the remaining thing now to do in this paper is to answer the following questions.
17
“Q1- How do nonmonetary rewards, as motivators, influence employees cooperation in comparison to monetary rewards?”
“Q2 -How do personal values (self-enhancement, self-transcendence) moderate the relationship in Q1?”
“Q3- What is the moderating effect of individual’s demographics (age, gender, education) in the relationship between non-monetary motives and cooperation?”
18
3. Theoretical Framework
With the current thesis, an argument is made that monetary and nonmonetary rewards, both extrinsic motivators, play a significant influential role to the levels of employees’ intention to cooperate, and moreover that the relationship between non-monetary rewards and cooperative behavior, is expected to be the strongest. In deeper analysis, the pre-mentioned relationship will be tested under the moderation of individual’s personal values, namely self enhancement and self-transcendence, with the assumption that the first will strengthen the relationship between monetary rewards and levels of cooperation, whereas the later will do the same to non-monetary rewards and level of cooperation. In the opposite combination, of values and rewards, the relationship between rewards and cooperation is expected to be weakened.
Lastly, additional factors, that this paper argues, can influence and shift the ‘relationships’ above, are the descriptive characteristics of individuals, specifically gender, age and level of education. Thus the following research model.
Personal Values
Self-Enhancement (Achievement, Power) Self-Transcendence (Benevolence, Universalism)
Nonmonetary Rewards VS
Monetary Rewards
Employee’s Intention for Cooperation
Demographics Gender & Education
H1a, H1b H1c, H1d H1b H2a, H2b, H2c, H2d H3a, H3b, H3c H1b
19
3.1 Monetary & non-monetary rewards and willingness to cooperate
Influencing employee’s intention to cooperate can be achieved by factors that are not in the control of the employee, thus extrinsic motivation (Deci & Ryan 2007). Enhanced or weakened employee cooperation depends on the type of incentives selected each time, with the monetary rewards being, traditionally, the most common choice to increase levels of cooperation (Bridoux et al, 2010). Nevertheless, non-monetary rewards have been proven an equal effective motivator for employees to be more cooperative (Deci & Ryan 2007). Levels of cooperation can be measured by monitoring employee’s behavior, for example the willingness to share with others knowledge (Cabrera & Cabrera, 2002). Based on the fact that monetary and non-monetary rewards can positively influence knowledge sharing behaviors (Lin 2007), thus cooperative behavior, in this research I expect that:
H1a Monetary rewards have a positive relationship to employee’s willingness to cooperate. H1b Non-monetary rewards have a positive relationship to employee’s willingness to cooperate.
Although monetary rewards can boost employee motivation by providing the means to fulfill needs (Long & Shields, 2010) they come with restrictions and even counterproductive results (Kerr, 1975, Harris & Bromiley, 2007). Employees also seek for a deeper meaning and an enriched job, that monetary rewards cannot provide (Grant & Parker, 2009). According to the above and the recent turn of attention to non-monetary rewards (Masclet at all, 2002), it is expected that:
H1c Non-monetary rewards have a stronger positive relationship to employee’s willingness to cooperate, than monetary rewards.
H1d No rewards have a weaker positive relationship to employee’s willingness to cooperate, than the monetary rewards.
20 3.2 Social values as moderators.
Despite the moderating effect of chosen motives, monetary or non-monetary, it is important to consider, and include in the equation the factor of individual heterogeneity, under the prism and classification of social values (Kollock, 1998). For example, in a situation where extrinsic motives, solely focused in monetary types, were deployed to employees, then there is a risk of increased selfish behaviors (Bowles, 2008), competitive behaviors (Bock & Kim, 2002), thus low levels of cooperation. By nature individuals can be categorized by two behavioral characteristics, those who benefit other people or society as a whole, and those who focus on individualistic goals and desires (Liebrand et al., 1986). Additionally, behaviors like the former pre-mentioned, have been related to self-transcendence values, benevolence & universalism, whereas, behaviors like the latter, related to self-enhancing values, power & achievement (Caprara & Steca, 2007). Based on the above I expect the following:
H2a: The positive relationship of monetary rewards on willingness to cooperate will be moderated by Self-Enhancement, such that it will be stronger for people that score high on self-enhancement.
H2b: The positive relationship of non-monetary rewards on willingness to cooperate will be moderated by transcendence, such that it will be stronger for people that score high on self-transcendence.
H2c: The positive relationship of monetary rewards on willingness to cooperate will be moderated by transcendence, such that it will be weaker for people that score high on self-transcendence.
H2d: The positive relationship of nonmonetary rewards on willingness to cooperate will be moderated by enhancement, such that it will be weaker for people that score high on self-enhancement.
21 3.3 Gender & education as moderators
There has been evidence in literature, that men and women perceive job incentives as motives, in a different way (Veroff et. al., 1980). Warr (2008) i.e. is mentioning that achievement and power (Page & Baron, 1995) are ranked higher in men’s priorities than in women’s. Bakan (1996), says that individuals can be classed in two categories, gender related and in terms of behavioral intentions, namely those who value high the concern for others, in contrast with more individualistic behaviors. It has been noticed, based on the job positions that end up, that men are perceived to act more individualistic whereas women described as more communal (Williams & Best, 1990). In combination with the content of $3.2 we can easily assume that self-enhancement values will appear more in men and self-transcendence values will appear more in women. Thus the following hypothesis for this research:
H3a The positive relationship between monetary rewards and employee’s willingness to cooperate is weaker for women, than for men.
H3b The positive relationship between non-monetary rewards and employee’s willingness to cooperate is weaker for men, than for women.
Last but not least, another demographic characteristic, that affect job results, and needs to be examined, is employees’ level of education (Warr, 2008). It has been observed, although limited (Warr, 2008), evidence of correlation, among employee’s level of education and job attributes (Vansteenkiste et al., 2007). Furthermore, highly educated employees, value purposed work higher than increased salaries and feeling of safety in the position, in contract with less-educated employees, where level of income and feeling of permanency, ranked higher in their preferences (Lacy et al., 1983). In these terms, the following relationship will be examined in this research:
H3c The relationship between non-monetary rewards and intention of cooperation is stronger for highly educated employees, than employees with lower level of education.
22
4. Methodology & research design
A detailed and explanatory, step by step, analysis of method and design of this research is unfolded in the following paragraphs of this chapter. A combination of two research techniques, those of vignette and survey, were used to measure the dependent/independent and moderating variables, respectively. The first part of the survey consists of the vignette, introducing an imaginary scenario to measure extrinsic motivation and the intention to cooperate. Next, the survey questionnaire is introduced to measure the moderating variables: personal values, gender and education, along with age as control variable. Additionally, data collection and sample description is depicted. The actual survey is illustrated, in full in Appendix I.
4.1 Method
To test the hypotheses mentioned in chapter three, and subsequently answer the research question derived from the literature review, the vignette technique was used as a reliable method, (Alexander & Becker, 1978), to capture individual responses. The independent variables, monetary and non-monetray rewards, are manipulated in the vignette and followed next by directed questions regarding the dependent variable (intention to cooperate - through knowledge sharing behavior), to unveil respondent's personal views and behavioral intentions (Schoenberg & Radval, 2000). Despite the lack of ‘natural’ setting, where ‘real’ behavioral attitudes could be observed, vignettes are still a proven method of triggering and capturing these behaviors (Hammersley & Atkinson , 1995), are frequently introduced by researchers for motivation and cognition, among various topics (Stolte, 1994), and with the ability to isolate and depersonalize, responder’s unique situation (Finch, 1987).
Additionally, and in combination with this factorial research design, the more common survey questionnaire was used to measure individuals’ personal values, moderators between independent/dependent variables in our case, and further demographic characteristics that
23
influence this relationship. As Saunders et al. (2009) mention, by using this research tool, scholar can obtain access to sufficient amounts of quantitative data, and benefit from the highly standardized format of this method, and perform statistical analysis and illustration of collected data.
To conclude, all responders were invited to fill in the survey in digitalized form that was send via e-mail, in order, to take also advantage of the higher levels of responses and quicker reaction time, associated with online surveys (Lonsdale et al., 2006). Personal network of connections was also used to further promote the questionnaire. Data were collected during a period of four weeks (August, 2015).
4.2 Sample
The survey was conducted among a diversified group of people, all across various European countries, industries and management levels. In total, 277 people responded to the survey, with 208 of them completing the questionnaire in full. For the purpose of this research, the participants were asked to state their employment status. Out of these 208 completed questionnaires, 49 responders were unemployed, and after removing these cases, the final dataset under analysis was at the level of 159, employed, responders, and 76.4% of completed questionnaires. One of the 3 different versions of the vignettes was, automatically and randomly, presented to each responder. All responders were informed regarding the academic purpose of this study.
4.3 Design
4.3.1 Dependent and independent variables.
The independent variables, monetary and nonmonetary rewards as types of extrinsic motivation and the dependent variable, employee’s intention to cooperate, are measured by the
24
introduction of a vignette. In this vignette, an imaginary situation is presented, describing a work-related scenario, where co-workers from cross-function teams and diversified expertise’s are requested to collaborate in order to successfully fulfill a complex and demanding project. Thus, levels of intention for cooperative behaviors, like sharing knowledge will be key success factors. This project is on top of employee’s already heavy workload, with strict deadlines to be met. The outcome of their cooperation will be measured in terms of quality and completion time, and respectively, employees will be rewarded. To manipulate the influence of independent variables, and test hypothesis H1a-c, the vignette is been presented, randomly, in 3 different versions. Describing the same scenario each time, upon the completion of the project there is a different reward for the employees. First version is with no reward, second version is with monetary reward, and third is with nonmonetary reward. The table below presents the way these versions are introduced to responders.
Vignette version
Type of reward Manipulation
1 No reward ---
2 Monetary
reward
If the project ends to success, team members will be awarded with a ,30% of their salary, bonus.
3 Nonmonetary
reward
If the project ends to success, team members and their success will be announced and recognized, via company’s internal communication channels, personally by company’s CEO.
The dependent variable, in our case intention to cooperate, has been associated with different behavioral expressions (Lazenga & Patterson, 1999, Bock et al., 2005) representing social dilemmas, such as the willingness to share knowledge (Cabrera & Cabrera, 2002). Following Bock et al. (2005), in order to measure employees’ intention for cooperation a 7-point
25
Likert scale was used to evaluate responses related to experience, documents, knowledge and overall sharing willingness. Additionally, investigating the level of sacrifice in the name of cooperation, it was adopted from Lazenga & Patterson (1999), one more item in the scale, regarding willingness to work extra hours. Table1 illustrates all the items and the 7-point scale. Table 1 Dependent variable, intention to cooperate:
How willing would you be to behave as illustrated in the statements below? Please indicate the most appropriate answer:
7-point scale
1 - I am willing to share all my experience or know how with the colleagues working in this project.
1=Very unwilling 2 3 4 5 6 7= Very willing 2 - I am willing to share all my helpful information and ideas with the
colleagues working on this project.
3 - I am willing to share copies of all my manuals and methodologies with the colleagues working on this project.
4 - I am willing to share my experience ideas and documents, knowing that they have a problem.
5 - I am willing to work extra hours.
26 4.3.2 Social values as moderators
In order to test the Hypotheses 2 and 3, mentioned in the previous chapter, I used a traditional questionnaire survey to measure the moderating variables, which are: age - level of education - gender, and the personal values, self- enhancement and self-transcendence.
The social values concept has been repeatedly considered in situations where individualistic trend behaviors were measured in relation with operationalization of enterprises or communities (Schein, 1985; Schwartz, 2001). The personal values concept is a proven theory in studying the heterogeneity of individuals and their behavioral traits (Schwartz, 2001). As we described before individuals can mainly be classified as those who care for collectively progress and those who care for individualistic progress (Liebrand et al., 1986). In this research, the focus is on the self-transcendence and self-enhancing personal values, associated with the pre-mentioned classifications.
To identify and measure these values from survey’s responders, I used a validated instrument, the Portrait Values Questionnaire (PVQ), to capture responder's ranking in one of the following personal values characteristics, universalism, benevolence (self-transcendence), and power, achievement (self-enhancement). The responder was given a description of a person or certain behavior and desire, and the responder had to indicate in a 6-point scale, how much like him/her the described person was. The full description of this part of the survey is presented in the Appendix I. PVQ items were chosen among other instruments, like Social Value Survey (SVS), due to its relatively simplicity, lack of numerical values, which may be confusing for responders, and due to the fact of being more concrete by providing descriptions of people rather than vague terminology of values. Additionally it is not obvious to the responder that he/she is questioned about personal values (Schwartz, 2001).
27 4.3.3 Age as control variable
Differences in work behaviors, based on age have been investigated from scholars, although traditionally focused on skills-related effect (e.g. Cavanaugh & Blanchard-Fields, 2006; Schaie, 1996, 2005). In their research Kanfer & Ackerman (2004), highlight the role of age-related differences, in relation with motivational incentives and the effect in work outcomes. Nowadays, there is evidence from recent studies, that individuals will become less or more motivated, in their aging course (Kooij et. al., 2011). Inceoglu’s et.al research (2012) provides additional evidence of the effect of age in motivational behaviors, and Warr (2008) specifies, i.e., that salary increases may become less important to older employees than younger age groups. Thus in the current research, age was used as a control variable.
4.4 Analysis
In order to proceed with the investigation and interpretation of the results, the collected data should go through a preliminary analysis to be evaluated in terms of normality, distribution and reliability for respective variables. Using Cronbach’s alpha the independent and dependent variables will be tested on reliability. As a next step will be a representation of the descriptive statistics for variables, giving an overview and separating categorical from continuous one. Continuing on the analysis, correlations among variables will be presented by highlighting the significant ones, considering Pearson correlation coefficients. Last but not least in order to test the hypothesis mentioned in chapter 3, a regression analysis will be executed to unveil any expected relationships, moderating or direct, between the variables. The aforementioned steps will be presented in detail in the following chapter. The program SPPS was used for the analysis.
28
5. Results
5.1 - Checking reliability:
In the current research design, I used a scaled instrument to measure 7 of dataset’s variables. These are: the dependent variable willingness to cooperate, and the moderating variables, benevolence, universalism, achievement, power, and transcendence and self-enhancement.
Starting with the dependent variable willingness to cooperate, the 7-point scale, showed a Cronbach’s alpha coefficient of 0.90, higher than 0.70 (Appendix II, Table 1), indicating that the internal consistency was good for this variable.
Regarding the moderating variables and starting with the Benevolence one, the Cronbach’s alpha coefficient was at levels of 0.78, showing also high internal consistency (Appendix II, table 1). In the case of Universalism the Cronbach's alpha coefficient was slightly higher, with 0.85 (Appendix II, table 1). The variables Achievement and Power showed significantly higher internal consistency with Cronbach’s alpha coefficient of 0.93 and 0.88 respectively (Appendix II, Table 1). Moving to self-transcendence variable, the Cronbach’s alpha coefficient was 0.87 showing high internal consistency (Appendix II, table 1), and self-enhancement showed even higher internal consistency with Cronbach’s alpha being at 0.92 (Appendix II, table 1).
Concluding, all the aforementioned variables have demonstrated higher Cronbach’s alpha coefficient than the threshold of 0.70, in order to validate good internal consistency and permit re-coding into separate variables.
29 5.2. Descriptive Statistics
In our analysis below, first the categorical variables are illustrated and then the continuous variables.
For this research’s needs, the categorical variables that were used, are firstly the motivational incentives as independent variables, measured by 3 versions of a vignette, and split into non-monetary reward, monetary reward and no reward, and secondly the moderating variables, gender, age and level of education. Frequencies and percentages are presented in Table 2 below.
Out of a total of 158 questionnaires that were completely filled in by responders under employment, the 3 versions of the vignette were almost equally distributed in 30+% for each version. Specifically, for the version with no reward: 55 responders, for the version with monetary reward: 52 responders, and for the version with nonmonetary reward: 51 responders (Table 2).
36% of the responders were found to be between 18 and 34 years old, 25% between 35 and 44 years old, 29% between 45 and 54 years old, with the remaining 9.5% being over 55 years old.
Most of the responders had gone through a higher education level, like 4-year college (22.8%) or master degree (60.1%), where the remaining ~13% had a 2-year college or lower level education. Out of this sample 81 responders (51.3%) were male and 77 were female.
30 Table 2: Categorical variables frequencies
Variable Frequency Percent Cumulative percent
Version vignette: Non-monetary reward Monetary reward No reward Total 51 52 55 158 32.3% 32.9% 34.8% 100.0% 32.3% 65.2% 100% Age 18-14 years 25-34 years 35-44 years 45-54 years 55-64 years Total 9 48 40 46 15 158 5.7% 30.4% 25.3% 29.1% 9.5% 100.0% 5.7% 36.1% 61.4% 90.5% 100% Gender Male Female Total 81 77 158 51.3% 48.7% 100.0% 51.3% 100.0% Level of Education
Less than High School
High School / GED
Some College
2-year College Degree
4-year College Degree
Master Degree Doctoral Degree Professional Degree (JD, MD) Total 0 7 6 7 36 95 5 1 158 0% 4.4% 3.8% 4.4% 22.8% 60.1% 3.2% 0.6 100% 0% 4.4% 8.2% 12.6% 35.4% 95.5% 98.7% 100.0%*
*0.7% of no selection in education question
Next step in the analysis was the descriptive statistics for the seven continuous variables, that were measured for this research, namely the dependent variable Willingness to cooperate, and the moderating variables, Benevolence, Universalism, Achievement and Power as Personal Values.
31
The dependent variable, willingness to cooperate was measured and ranked values from 1 to 7 with a mean of 5.71 and a standard deviation of 1.30 (Table 3). The skewness of -1.77 and the kurtosis of 3.56 indicate a cluster of values to the right with a rather peaked distribution (Table 3).
For the variables Benevolence, Universalism, Achievement and Power, I measured the results at the level of personal values clusters, namely Self-Transcendence (benevolence + universalism), and Self-Enhancement (achievement + power).
Self-Transcendence as moderating variable, had values ranking from 2.83 to 6, with a mean of 4.90 and standard deviation of 0.70. Given the levels of skewness at -0.49 and kurtosis at -0.89 the values were clustered at the right and the distribution was rather flattened.
Self-Enhancement as moderating variable, had values ranking from 1 to 6 with a mean of 3.6 and standard deviation of 1.14. Skewness was at -0.27 and kurtosis at -0.51, meaning that the values were clustered to the right at the high end scores and its distribution flat.
In the bottom part of Table 3, the descriptive statistics for the dependent variable, willingness to cooperate, are also given per vignette type.
Table 3 Frequencies for continuous variables
Variable N Min. Max. Mean Std. Dev Skewness Kurtosis Willingness to Cooperate 158 1.00 7.00 5.71 1.30 -1.774 3.569 Self - Transcendence 158 2.83 6.00 4.90 0.70 -.495 -.089 Self - Enhancement 158 1.00 6.00 3.63 1.14 -.272 -.513 Willingness to Cooperate – Frequencies per vignette type
Vignette 1 - Non-monetary 51 1.00 7.00 5.40 1.68 -1.453 1.390 Vignette 2 -Monetary 52 1.40 7.00 5.71 1.25 -1.632 2.952 Vignette 3- No reward 55 4.20 7.00 6.00 0.79 -0.439 -0.811
32 5.3 Testing normality
The next step in analyzing the data was to test the variables in terms of normal distribution.
The variable willingness to cooperate showed a mean of 5.71 and a 5% trimmed mean of 5.87, an indicator that the mean could be influenced from the extreme scores (Appendix II, Table 2). Based on the negative scores of skewness and positive scores of kurtosis (Appendix II, Table 2), values were centered and clustered to the right, as it can been also seen in the histogram (Appendix II, graph 1). Although the values showed a normal distribution in the Q-Q plot table (Appendix II, graph 2), I further investigated regarding extreme scores and the outliers (Appendix II, graph 3) have been excluded from the dataset.
Excluding the outliers it was observed that the mean of 5.93 was closer to the 5% trimmed mean 5.99 (Appendix II, Table 3) and the values are following a rather straight line normally distributed around it (Appendix II, graph 4).
The scores of the Kolgomorov-Smirnov statistic (Appendix II, Table 6) were suggesting a normality violation. By plotting the values of the dependent willingness to cooperate, had as a result the rather normal distribution of scores, around a straight line (Appendix II, Graphs 2, 3, 4), concluding that the extreme values do not influence the mean.
5.4 Correlations
In this part of the analysis, the correlations, assessed with the use of Pearson's correlation coefficient, are illustrated with the most important of them, between dependent and independent variables, discussed. In order to check the correlations, the transformation of all categorical variables to dummies had to proceed.
33
By looking at the table (Table 4), a first observation is that the independent variable, nonmonetary reward incentives, negatively correlates (not significantly) to the dependent variable willingness to cooperate (Table 4). Similarly, a negative correlation is observed between monetary rewards and the dependent variable. We see though, that the independent variable, no reward, positively related to willingness to cooperate, and among the three incentives versions, it has the strongest relationship with the dependent variable.
Willingness to cooperate is also positively related to Self-Transcendence (Pearson correlation coefficient = 0.32; Sig = 0.00), and age (35-44) (Pearson correlation coefficient = 0.21; Sig = 0.00). From this we can suspect that people clustered as self-transcendence based on their personal values, will be more willing to cooperate with their colleagues, and also people aged 35 to 44 will demonstrate increased willingness to cooperate. On the contrary, willingness to cooperate is negatively correlated (Pearson correlation coefficient = -0.21; Sig = 0.00)(Table 4) with the age group 25 to 34 years, indicating the lower levels of employee’s willingness to cooperate at this age group.
Other observations that we can made regarding correlations, is the positive relationship between age group 55-64 and nonmonetary rewards (Pearson correlation coefficient = 0.19; Sig = 0.00), while also positively related to Self-Transcendence (Pearson correlation coefficient = 0.18; Sig = 0.01), suggesting that people at that age will be more socially-oriented and more interested in non-tangible rewards for their work.
34
Table 4: Correlation Table for the dependent variable willingness to cooperate
Variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 1 Wil. to Cooperate 2 Nonmonetary -.010 3 Monetary -.049 -.463** 4 No Incentives .057 -.498** -.538** 5 S-Transcendence .320** .160 -.074 -.080 6 S- Enhancement -.093 .026 -.069 .043 -.036 7 Male .116 -.061 -.019 .077 -.020 .077 8 Age 18-24 -.063 .141 -.060 -.076 -.147 .089 -.035 9 Age 25-34 -.214** -.088 -.041 .124 .97 .138 -.105 -.168* 10 Age 35-44 .212** -.132 .053 .073 -.65 -.018 .014 -.152 -.401** 11 Age 45-54 -.055 .023 .074 -.094 -.078 -.085 .058 -.155 -.408** -.370** 12 Age 55-64 .153 .190* -.081 -.101 .186* -.134 .083 -.081 -.213** -.193* -.197* 13 Edu: H. School .084 -.059 .000 .056 -.021 -.002 .063 -.052 -.062 .031 .104 -.065 14 Edu: S Col/ge -.078 .041 -.053 .013 .005 -.116 -.042 -.047 -.123 -.112 .219** .068 -.038
15 Edu: 2yr College .005 .131 -.022 -.103 -.96 -.155 .152 -.056 -.079 .010 .006 .146 -.045 -.041
16 Edu: 4yr College -.151 .109 .033 -.136 -.026 .000 -.046 .121 -.001 .014 -.029 -.073 -.115 -.104 -.124
17 Edu: Master .132 -.148 .029 .113 .070 .069 .022 -.023 .100 .062 -.140 -.019 -.250** -.227** -.271** -.688**
18 Edu: Doctor -.053 .041 -.131 .090 .003 .158 -.116 -.047 .038 -.112 .136 -.060 -.038 -.034 -.041 -.104 -.227**
19 Edu: Prof. Degree .099 -.054 .116 -.062 .004 -.97 -.084 -.021 -.054 -.049 -.050 .255 -.017 -.015 -.018 -.046 -.100 -.015
**. Correlation is significant at the 0.01 level (2-tailed) *. Correlation is significant at the 0.05 level (2-tailed)
35 5.5 Regression Analyses.
In this part of the analysis, the hypotheses that were presented in Chapter 3 are going to be tested to validate if they can be supported or not based on the data. For running the regression analysis, and in order to test the direct and moderating relationships on the dependent variable, Willingness to Cooperate, I used a hierarchical multiple regression to test the hypothesis. Thus, I recoded all variables to dummy ones. In the models below, the direct effect of the variable is illustrated in Model A by including the control variables, then in Model B the independent variable is added, followed by Model C and Model D were the moderating effect is illustrated, in the tables below.
The first hypothesis to be tested was the positive relationship between the independent variable, monetary motivation, and dependent variable, employees’ willingness to cooperate. From the results on Table 7, we can see that the 9.2% is explained by the control variable age. The age group 25-34 was used as baseline for all the following models. The model was significant at the 10% (F change = 3.670; Sig. F Change = 0.07). By introducing to the model, the independent variable monetary rewards, we see just a 0.3% increase in the R2 (Table 7). Model B was not significant at 10% level (F Change = 0.422; Sig. F Change = 0.50). Monetary rewards, showed a negative regression coefficient and not significant (B = -0.053, Sig = 0.505) (Table 7). Although the significance was low for the independent variable monetary rewards, it had a negative influence to the dependent variable willingness to cooperate, thus Hypothesis H1a was not supported. The control variable age, showed positive effect at 5% level for age group 35-44 (B = 0.298; Sig = 0.00), and also positive effect for age group 55-64 (B= 0.233; Sig = 0.00) (Table 7).
36
The second hypothesis to be tested was the positive relationship between non-monetary rewards as motivators and employees willingness to cooperate. From the results on Table 8, we can see that the 9.2% is explained by the control variable. The model was significant at the 10% (F change = 3.670; Sig. F Change = 0.07). By introducing to the model, the independent variable nonmonetary rewards, we see no increase in the R2 (Table 8). Model B was not significant at Table 7: Regression results (DV: Willingness to Cooperate)
Model A Control Variables Model B Monetary Rewards B Sig B Sig (Constant) 5.648 .000 5.678 .000 Monetary Rewards -.053 .505 Control Variables Age 18-24yrs .017 .840 .015 .861 Age 35-44yrs .298 .002 .301 .001 Age 45-54yrs .103 .267 .108 .249 Age 55-64yrs .233 .008 .230 .009 R .303 .308 R2 .092 .095 Adjusted R2 .067 .063 F change 3.670 .007 .447 .505
37
10% level (F Change = 0.066; Sig. F Change = 0.80). Nonmonetary rewards, showed a negative regression coefficient and not significant (B = -0.021, Sig = 0.798) (Table 8). Although the significance was low for the independent variable nonmonetary rewards, it had a negative influence to the dependent variable willingness to cooperate, thus Hypothesis H1b was not supported. The control variable age, showed positive effect at 5% level for age group 35-44 (B = 0.297; Sig = 0.00), and also positive effect for age group 55-64 (B= 0.237; Sig = 0.00) (Table 8).
Table 8: Regression results (DV: Willingness to Cooperate) Model A Control Variables Model B Nonmonetary Rewards B Sig B Sig (Constant) 5.648 .000 5.658 .000 Nonmonetary Rewards -.021 .798 Control Variables Age 18-24yrs .017 .840 .020 .811 Age 35-44yrs .298 .002 .297 .002 Age 45-54yrs .103 .267 .105 .262 Age 55-64yrs .233 .008 .237 .008 R .303 .304 R2 .092 .092 Adjusted R2 .067 .061 F change 3.670 .007 .066 .798
38 Table 9: Regression results (DV: Willingness to Cooperate)
Model A Control Variables Model B No Rewards B Sig B Sig (Constant) 5.648 .000 5.587 .000 No Rewards .073 .369 Control Variables Age 18-24yrs .017 .840 .025 .765 Age 35-44yrs .298 .002 .300 .002 Age 45-54yrs .103 .267 .114 .224 Age 55-64yrs .233 .008 .243 .006 R .303 .312 R2 .092 .097 Adjusted R2 .067 .066 F change 3.670 .007 .812 .369
The third hypothesis to be tested was the estimation that the relationship between no-monetary rewards and willingness to cooperate, would be positively stronger that the relationship between monetary rewards and willingness to cooperate. Based on the results of the regression analysis on Table 7 & 8, it is clear that there isn’t any positive relationship between these two independent variables and dependent variable, willingness to cooperate. However, regardless the insignificant coefficients, the effect of monetary rewards were slightly stronger (negatively) than
39
the effect of nonmonetary reward. Therefore the hypothesis H1c is not supported. Table 10, presents also the comparison among the three types of independent variable (monetary/nonmonetary rewards and no rewards)
Table 10
Independent Variable B Sig
Nonmonetary reward
-.021 .798
Monetary reward
-.053 .505
No reward .073 .369
The forth hypothesis to be tested was that no rewards as independent variable, would have weaker positive relationship, with willingness to cooperate, than monetary rewards. As we see in Table 9 though, the Model B was not significant at 10% level (F Change = 0.812; Sig F Change = 0.36), and the independent variable, no-rewards, had a non-significant but positive regression coefficient (B= 0.073; Sig= 0.36). Even though the positive relationship of no-rewards to willingness to cooperate, and also not significant, we see that the relationship was stronger than the monetary rewards (Table 10), thus Hypothesis H1d is partly supported.
Regarding the effect of Social Value, Self-Enhancement, as moderator in the relationship between monetary rewards and willingness to cooperate, the fifth hypothesis to be tested, was that the positive pre-mentioned relationship would be stronger for Self-Enhancement people. The results of the regression analysis for Model C, including control variables, monetary rewards as independent variable, and Self-Enhancement, showed that there wasn’t any significant relationship between Self-Enhancement and willingness to cooperate (Table 11). The model was not significantly improved overall (F Change = 0.456; Sig F Change: 0.501)(Table 11), and R2
40
increased only 0.3%. Going further with the analysis and introducing the interaction effect the results showed that the model was improved at 10% level of significance (F Change = 3.605; Sig F Change: 0.060), and R2 increased by 2.2 %. Although the significant effect of the interaction of self-enhancement, it was negatively related with willingness to cooperate (B= -0.152; Sig= 0.06)(Table 11), therefore Hypothesis H2a was not supported.
Table 11: Regression results (DV: Willingness to Cooperate)
Model A Control Variables Model B Control Variables + Independent Variable Model C Control Variables + IV + Moderator Model D Control Variables + IV + Moderator + Interaction Effect
B Sig B Sig B Sig B Sig
(Constant) 5.648 .000 5.678 .000 5.847 .000 5.868 .000 Monetary Rewards -.053 .505 -.057 .478 -.063 .433 Self-Enhancement -.055 .501 -.057 .480 Interaction Self-enhancement -.152 .060 Control Variables Age 18-24yrs .017 .840 .015 .861 .016 .846 .026 .756 Age 35-44yrs .298 .002 .301 .001 .296 .002 .285 .003 Age 45-54yrs .103 .267 .108 .249 .100 .292 .075 .427 Age 55-64yrs .233 .008 .230 .009 .219 .014 .207 .019 R .303 .308 .312 .346 R2 .092 .095 .098 .120 Adjusted R2 .067 .063 .060 .077 F change 3.670 .007 .447 .505 .456 .501 3.605 .060
41
The sixth hypothesis to be tested was that the (positive) relationship between nonmonetary rewards, as motives, and employee’s willingness to cooperate, would be stronger for Self-Transcendence people. The results of moderating effect are illustrated in the Table 12 below. The control variables were added in model A, and the independent variable in Model B. Self-Transcendence variable was put in Model C, and in case the main effect was significant, it was further tested by adding the interaction variable in Model D. Regarding the moderating effect of Self-Transcendence on the direct relationship between nonmonetary rewards and employee’s willingness to cooperate, the regression results of Model C, including control variables, independent variable, and moderator, showed that Self-Transcendence was positively related to willingness to cooperate, and significant at 5% level (B= 0.341; Sig= 0.00) (Table 12). Furthermore, by introducing Self-Transcendence in the model, R2 increased by 10.6% and the model significantly improved (F Change = 18.845, Sig F Change= 0.00)(Table 12). However, when the interaction effect was introduced in Model D the model was not increased significantly overall (F Change = 0.366; Sig F Change= 0.54) (Table 12). Therefore, Hypothesis H2b was not supported.
42 Table 12: Regression results (DV: Willingness to Cooperate)
Model A Control Variables Model B Control Variables + Independent Variable Model C Control Variables + IV + Moderator Model D Control Variables + IV + Moderator + Interaction Effect
B Sig B Sig B Sig B Sig
(Constant) 5.648 .000 5.658 .000 3.485 .000 3.448 .000 Nonmonetary Rewards -.021 .798 -.076 .338 -.085 .294 Self-Transcendence .341 .000 .346 .000 Interaction Self-transcendence .047 .546 Control Variables Age 18-24yrs .017 .840 .020 .811 .088 .283 .096 .251 Age 35-44yrs .298 .002 .297 .002 .333 .000 .332 .000 Age 45-54yrs .103 .267 .105 .262 .150 .092 .149 .095 Age 55-64yrs .233 .008 .237 .008 .206 .015 .206 .015 R .303 .304 .445 .447 R2 .092 .092 .198 .200 Adjusted R2 .067 .061 .164 .161 F change 3.670 .007 .066 .798 18.845 .000 .366 .546
Testing the effect of Self-Transcendence on the relationship between monetary rewards and willingness to cooperate, the results of regression test in Model C, including control variables, independent variable monetary rewards, and Self-transcendence, showed a positive relationship between Self-Transcendence and willingness to cooperate (B= 0.326: Sig=
43
0.00)(Table 13). By introducing Self-Transcendence in the model, R2 increased by 9.9% and the model was significantly improved at 5% level (F Change = 17.605; Sig. F change = 0.00) (Table 13). However by introducing the interaction effect in Model D, the model was not significantly improved (F Change = 0.410; Sig. F Change = 0.52), thus Hypothesis H2c was not supported.
Table 13: Regression results (DV: Willingness to Cooperate)
Model A Control Variables Model B Control Variables + Independent Variable Model C Control Variables + IV + Moderator Model D Control Variables + IV + Moderator + Interaction Effect
B Sig B Sig B Sig B Sig
(Constant) 5.648 .000 5.678 .000 3.561 .000 3.559 .000 Monetary Rewards -.053 .505 -.034 .657 -.037 .632 Self-Transcendence .326 .000 .326 .000 Interaction Self-transcendence -.048 .523 Control Variables Age 18-24yrs .017 .840 .015 .861 .072 .375 .074 .364 Age 35-44yrs .298 .002 .301 .001 .336 .000 .340 .000 Age 45-54yrs .103 .267 .108 .249 .146 .103 .147 .100 Age 55-64yrs .233 .008 .230 .009 .190 .023 .189 .023 R .303 .308 .440 .443 R2 .092 .095 .194 .196 Adjusted R2 .067 .063 .160 .157 F change 3.670 .007 .447 .505 17.605 .000 .410 .523
