Reward systems and the intention to join firms : the influence of Dark Triad personalities

(1)

Reward systems and the intention to join

firms: The influence of Dark Triad

personalities

Student:

Sellina Menso

Studentnumber:

6130356

First supervisor:

Mw. dr. K.J.P. Quintelier

Date of submission: 02-07-2014

Version:

First version

Programme:

MSc Business Studies - Strategy

(2)

Page | 2

List of Tables

Table 1:Principle Components Analysis using Oblique Rotation ... 24

Table 2: Reliability Statistics ... 25

Table 3: Descriptive Statistics ... 27

Table 4: Correlation table ... 29

Table 5: Coefficients and statistics ... 30

List of Figures

Figure 1: Conceptual Model ... 17

(4)

Page | 4

Abstract

This study examined whether job applicants’ intentions to join a firm are influenced by the offered reward system of that particular firm and the degree to which this relation is influenced by Dark Triad personalities. In this study graduate students (N = 297) were randomly assigned to one of the two vignettes in which one variable, a firm offering either a variable pay or a fixed pay, was manipulated. Students then had to indicate their intention to join the firm. This was followed by questions serving to identify Dark Triad personality traits. Results showed that the offered reward system did have an effect on job applicants, and that their intention to join a firm was enhanced for firms offering fixed pay. These results disagree with the current increase of variable pay in the world. Next, the results showed that this relation is not influenced by Dark Triad personalities; therefore there are many other factors that influence job applicants’ intention to join a specific firm.

(5)

Page | 5

Chapter 1: Introduction

The success and growth of firms depends on high quality staff. Therefore they need to attract high quality job applicants (Rynes & Barber 1990). However, given the changing demographics and economic factors, this will become increasingly more difficult in the coming years (Chapman et al. 2005). Researchers have even spoken of a “war for talent” (Michaels, Handfield-Jones, & Axelrod, 2001). An important aspect of attracting high quality applicants is increasing their intention to join the firm. Many researchers have written about factors that influence job applicants’ intention to join specific firms and their reaction to recruitment activities (Bhattacharya et al. 2012; Zhang & Gowan 2012; Rupp et al. 2013; Evans & Davis 2011; Turban et al. 1998; Taylor & Bergman 1987; Breagh & Starke 2000). All these articles have contributed to a better understanding of job applicants and their motivations to join firms. Many questions still remain, but we can conclude that there are many factors that influence job applicants’ intentions to join a specific firm.

One of the most important factors in deciding on accepting employment is remuneration (Bretz & Judge 1994). Next to this, the reward system that a firm offers to its employees is considered another important factor when deciding on whether to join a firm (Reynolds 1951; Bretz & Judge 1994). Reward systems specify what an employee can expect for his or her contributions and performance and thereby represent a company’s values and beliefs (Kerr and Slocum 1987).

Personality is another important factor that influences the intention to join a firm. In the literature it is assumed that people are attracted to those firms that best suit their beliefs, values and norms (Kristof 1996; Cable & Judge 1996; Schneider 2001; Carless 2005). While most literature on personality traits and intention to join a firm has focused on the Big Five personality traits - Extraversion, Agreeableness, Conscientiousness, Neuroticism and Openness to Experience - little research has investigated the so-called dark personalities, also known as Dark Triad personalities (Paulhus & Williams 2002). Dark Triad personalities consist of three related personality traits: Machiavellianism, narcissism and psychopathy and are associated with a number of negative traits. Dark Triad personalities share a number of features, namely they all entail a socially malevolent character with behaviour tendencies toward self-promotion, emotional coldness, duplicity, and aggressiveness (Paulhus and Williams 2002, p. 557). Individuals scoring high on Dark Triad personalities are considered to be more toxic employees (Jonason et al. 2012) and a strong relation between Dark Triad personalities and counterproductive behaviour has been found (Heijnis 2009).

(6)

Page | 6 Attracting individuals that with Dark Triad personalities might have several negative consequences for an organization once they are employed (Crysel et al. 2013; Griffith et al. 2009). However, people with explicit Dark Triad personalities are often found in managementt boards and high positions (Babiak & Hare 2006; Babiak et al. 2010; Gudmundsson & Southey 2011; Lipman 2013; Chatterjee & Hambrick 2007). Individuals in such positions have substantial power in the firm and can thus significantly influence its performance. In addition, these individuals are also associated with risky behaviour (Crysel, et al. 2013), and thus may do whatever it takes to achieve an objective, despite potentially detrimental consequences for the company. Thus it is important to investigate the interaction between dark triad and pay system.

Another reason for investigating the possible interaction between these variables is the call for research of Cable and Judge (1994). Their study on pay preferences and job search decisions discovered that job search is indeed affected by the offered reward system, and suggested to review other job applicants groups and to examine their pay preferences (ibid. p.346). The present study will therefore investigate the influence of the reward system on job applicants’ intentions to join a specific firm, and whether this is influenced by their Dark Triad personality traits. Thus the research question that this study aims to answer is the following:

What is the influence of reward systems on the intention to join a specific firm for individuals with explicit Dark Triad personalities?

The theoretical contribution of the present study is that the study fills a gap in the existing literature by giving information on how reward systems that firms offer can influence job applicants’ intention to join a firm and whether this relation is influenced by Dark Triad personalities. This study will also shed more light on Dark Triad personalities and discover correlations between these personalities and the intention to join a specific firm. The most important managerial implication is to show them what the potential risks and opportunities are for having Dark Triad personalities in a firm. Next to this, Dark Triad personalities are mostly associated with negative consequences. Therefore knowing how reward systems can influence individuals with Dark Triad personalities in their intention to join a firm, can help firms adjust their recruitment activities so that these individuals are less or more motivated to join their firm. More contributions are discussed in the discussion section.

The study is presented as follows. First, the literature review follows; here the intention to join a firm, reward systems and the Dark Triad personalities will each be discussed separately. Then the

(7)

Page | 7 conceptual model is presented. In chapter three the methodology used for the study and the operationalization of the variables will be given. In chapter four the results of the study are given. In chapter five a discussion will follow where the results are discussed, here also limitations to the study and directions for future research are discussed, as well as the managerial implications. Lastly the conclusion of the study is discussed.

(8)

Page | 8

Chapter 2: Literature review

This chapter discusses the existing literature on applicants’ intentions to join firms, followed by reward systems that firms offer. This is followed by a discussion on Dark Triad personalities.

2.1 Intentions to join a firm

In this thesis, “the applicant’s desire to join a specific firm” is used as a definition for intention to join a firm. In the literature different synonyms have been used for intentions to join a firm, e.g. “desire to pursue employment with an organization” (Williamson et al. 2003). Furthermore, organizational attractiveness is defined by Tsai and Yang (2010, p. 49) as “applicants’ willingness to pursue jobs and to accept job offers in an organization.” Even though the definitions of intention to join a firm and organizational attractiveness look similar, they are not identical. The similarity between the two concepts is that they both look at an applicant’s desire and willingness to join and/or pursue a job (at) a firm. The difference between the two concepts lies in the part that relates to what influences that desire and willingness. With organizational attractiveness, this desire is influenced by the attractiveness of the firm, meaning whether the firm has certain factors which are important for applicants. According to Rynes (1991) the size of the firm has a role in organizational attractiveness. There are more factors that could influence the intention to join a firm, so here not only the organizational attractiveness, but also other factors such as social influences by friends and family or a person’s personality play a decisive role. Highhouse et al. (2003) take a different approach and believe that the intention to join a firm is one of the characteristics that assess an organizations’ attractiveness. However, because these two concepts in the literature are often intertwined, and because organizational attractiveness has an influence on job applicants’ intention to join a firm, literature on organizational attractiveness has also been used in this study.

In order for a firm to recruit a high quality workforce it should create and increase its organizational attractiveness (Tsai & Yang 2010), because that will lead to an enhanced intention for potential job applicants. The organizational attractiveness of a firm is related to its properties and performance, e.g. ethical leader behaviour increases the organizational attractiveness and the intention to join that particular firm (Strobel et al. 2010). Organizational attractiveness could be increased if the firm would highlight those aspects that are important for job applicants. Another factor of organizational attractiveness is the offered reward system. Applicants pay attention to different factors when they decide on the attractiveness of a firm; individuals also differ in the factors that they find important in

(9)

Page | 9 an organization. The differences in the organizational attractiveness perceived by different individuals can be explained by their diverse personalities and background (Schneider et al. 1998; Lievens et al. 2001). Individual stages of life and specific situations, such as e.g. being unemployed and needing money immediately, may further defer the significance of organizational attractiveness. However, in general the more attractive an organization is, the more likely it is that a high quality job applicant will want to join that firm. This also means that when a firm is less or not at all attractive to job applicants, fewer people will want to work there. It is therefore mandatory for firms to enhance job applicants’ motivation to join their firm by increasing their organizational attractiveness.

2.2 Reward systems

The following definition of reward systems by Kerr and Slocum (1987, p.99) will be used in this study: “The reward system defines the relationship between the organization and the individual member by specifying the terms of exchange: It specifies the contributions expected from members and expresses values and norms to which those in the organization must conform, as well as the response individuals can expect to receive as a result of their performance.”

Reward systems represent a company’s values and beliefs and are therefore important to understand a company’s culture (Kerr & Slocum 1987, p. 99). Indeed, reward systems influence the level of firm attractiveness for job applicants, as pay seems to be one of the few factors applicants keep in mind when setting out minimal standards for deciding on whether or not to apply for a job (Reynolds 1951). Also Bretz and Judge (1994, p. 541) have found that the pay level and a company’s reward system are amongst the most significant predictors for job offer acceptance. Especially the amount of remuneration is very important; when unemployed workers are searching for employment they may still reject an offer due to what they believe to be insufficient remuneration (Montgomer 1992, p. 587).

There is a wide range of reward systems that companies can use to reward their employees. In this study the focus is on variable and fixed pay reward systems for the individual, because these are the forms of compensation that are most used by organizations (Merriman 2014). In the following two sections these two types of reward systems will be discussed more extensively.

(10)

Page | 10

2.2.1 Fixed pay

Fixed pay is also known as base pay and salary. This thesis uses the following definition for fixed pay:

“The part of the total salary or pay that does not depend in a given year on the individual's job performance.” (Chauvin & Ash 1994, p. 636)

Thus, fixed pay is the pay that employees receive on a regular basis and is not directly related to performance. Fixed pay is often used when it is difficult to measure the output of individual performance (Lazear 1986; Erikkson & Villeval 2008). Changes in fixed pay can be based on employees’ performance and occur when individuals are given a promotion or demotion. However sometimes increases in fixed pay are determined through a formal salary plan (Kerr & Slocum 2005 p.132). Changes in fixed pay usually take place on an annual basis, or less frequently. The amount of increase in base pay is determined by the supervisor’s performance evaluation of that employee (Frisch & Dickinson 1990).

There are several advantages to having a fixed pay reward system in an organization, both for the company and the employee. A fixed pay reward system leads to a higher self-perceived status of the employees when they see and/or know that they earn more than in comparable positions (Thierry 2001). Employees that tend to feel more appreciated and valued by the organization are rated as high performers, which reflect positively on the organization (Gardner et al. 2004, p.316). Base pay level is also less influenced by temporal factors (Gardner et al. 2004), which can benefit employees, as their remuneration is not affected by sudden economic downturn. Another possible benefit of having a fixed pay reward system in place is that it will probably attract the more risk-averse individuals (Cadsby et al. 2007). This can benefit the company as it will lead to employees taking fewer risks when performing their tasks, which in turn will reduce risks affecting the performance of the organization.

2.2.2 Variable pay

Variable pay is pay for performance; the employee receives a lower percentage of a base salary, but, subject to performance, can earn more money with variable pay in the end. This thesis uses the definition of Kurdelbusch (2002, p. 326) for variable pay:

(11)

Page | 11 “Variable pay includes all wage components that are contingent either on a firm’s performance (or the performance of units within the firm), such as profit-sharing, annual bonuses or employee ownership plans, or on individual workers’ achievements (that is, performance-related pay), ranging from traditional piecework to different systems of reward linked to appraisal and income-relevant forms of management by objectives.”

Variable pay can be given when employees’ productivity exceeds a specific standard (Frisch & Dickinson 1990). However, performance below specifications would negatively affect the amounts received. Variable pay is, as Madhani (2011) says a “one-time earning”, the earnings must be re-earned for each month or cycle. Because of this, variable pay could be linked and associated with individuals more prone to risk taking because when earning variable pay, it is not known at the start of the pay cycle how much they will receive (Kuhn & Yockey 2003). The only thing that they are certain of is their base pay.

Variable pay is often part of the reward system for executives and directors and in jobs where results-based performance measures are available (Gerhart & Fang 2014). An example of the latter is sales based organizations. According to Madhani (2013) variable pay is often found in these organizations because sales employees need to be compensated in a way that will motivate them to sell as many products or services as possible, and this motivation can best be provided by such a reward system. The pay of CEOs also tends to consist of about 50% of bonuses (Jensen & Murphy 2001, 138). Variable pay is less likely to be found in lower level positions (Gerhart & Fang 2014) or in the manufacturing industry (Stroh et al. 1996).

In the past years, variable pay has become a more common concept in the business world (Gerhart & Fang 2014). Kurdelbusch (2002, p. 326) believes that this increase in variable pay is based on the assumption that variable financial rewards provide strong incentives to work harder and/or better, which benefits the organization. Kuvaas (2006, p. 366) also supports this theory and adds that organizations do this because they will “attract and retain highly valued employees and [it] will generate favourable attitudes toward the organization”. Another reason why variable pay may be interesting for organizations is because it is a way to share the losses and profits from the organization (Cardon & Stevens 2004). Benefits of having a variable pay structure in an organization are best illustrated with an example from a study from Lazear (2000). This sample shows that when a company changed their pay structure from a fixed pay reward system to an individual incentive, i.e. a variable pay system, there was an increase of 44% in productivity. Eriksson & Villeval (2008) also found a positive effect when they studied the link between variable pay and performance. However

(12)

Page | 12 there are also disadvantages of having a variable pay system. Such a disadvantage could be that employees may only perform the work for which they are rewarded, neglecting other tasks which are also part of their function. Thus, a variable pay system could also encourage dysfunctional behaviour (Gupta & Shaw 2014). Variable pay could even cause employees to do whatever it takes to achieve the targets, even if their actions would negatively affect the company (Jensen 2003). Next to these disadvantages, variable pay could also cause tension between employees and cause employees to regard each other as competitors, especially when they are individually rewarded as opposed to as a group (Meyer 1976).

Overall, the researcher believes that job applicants in general will prefer fixed pay because the risk of them not getting the same pay the next month is very small when receiving such a pay. Next to this, even though variable pay in general is subject to individual performance and an employee is therefore able to influence his or her pay, other factors can also influence the performance of the employee, such as the economy in general or the environment. In addition, several authors have found that individuals in general prefer fixed pay over variable pay (Cable & Judge 1994). Therefore the following hypothesis is that:

Hypothesis 1: The intention to join a firm is higher when a firm offers fixed pay than when it offers variable pay

2.3 The Dark Triad traits

This chapter commences by discussing each of the three traits of the Dark Triad personalities, followed by the Dark Triad as a whole and their potential influence on applicants’ reaction to reward systems.

2.3.1 Machiavellianism

Machiavellianism is a construct from Richard Christie’s (1970) selection of statements from Niccolo Machiavelli’s original books (Paulhus & Williams 2002). Niccolo Machiavelli was an influential man in the Italian Renaissance. He wrote the handbook ‘The Prince’ (1532) for those trying to seize and retain political power (O’Bryan et al. 2012) after having observed the political fight in Italy over political supremacy and the strategies that were deployed (Hawley 1982). One of the strategies he observed was the tendency towards manipulation. This thesis uses the following definition for a Machiavellian personality:

(13)

Page | 13 “Machiavellian personality is defined by three sets of interrelated values: an avowed belief in the effectiveness of manipulative tactics in dealing with other people (e.g., “Never tell anyone the real reason you did something unless it is useful to do so”), a cynical view of human nature (e.g., “It is safest to assume that all people have a vicious streak and it will come out when they are given a chance”), and a moral outlook that puts expediency above principle (e.g., “It is hard to get ahead without cutting corners here and there”). “

(O’Boyle et al. 2012, p. 558)

Individuals that score high as opposed to those that score low on Machiavellism tend to behave in a cold and manipulative way in both laboratory and real world studies. They prefer to dominate situations and have control in their own hands rather than to accept someone else’s control (Christie & Geis 1970). Next to this, Machiavellians also tend to strategize on how to get to where they want to be (Hawley 1982). Another feature of the high-scoring Machiavellians is that they like to be praised by others (Pandey & Rastogi 1979). Machiavellians also tend to help themselves to someone else’s money; however it is unlikely that they are attracted to high risk situations (Jones 2013). Therefore individuals that score high on Machiavellism might be more tempted to join a firm where employees will earn a fixed pay, because a variable pay system tends to be risky (Kuhn & Yockey 2003). The following hypothesis follows from this line of argument:

Hypothesis 2: The intention to join a firm is higher when a firm offers fixed pay than when it offers variable pay. This relation will not be influenced by individuals’ scores on Machiavellism.

2.3.2 Psychopathy

Psychopathy is a personality disorder (Andrews & Furniss 2009). This thesis uses the definition by Glenn et al. (2011, p.372) for psychopathy:

“[It’s a] personality type describing individuals who demonstrate a pronounced lack of guilt, remorse, and empathic concern for others. Psychopaths appear to lack emotional distress and are impervious to distress in others. In addition, they are superﬁcially charming, manipulative, egocentric, and grandiose. They tend to be impulsive, risk-taking, and fail to plan for the future...”

(14)

Page | 14 Individuals that are psychopathic have a lack of affect and tend to be dishonest and insensitive (Crysel et al. 2013). They also are thrill-seekers (Gudmundsson & Southey 2011; Horvath & Zuckerman 1993). Psychopathy has had the attention of researchers for over 60 years since Cleckley, H. (1941) wrote his book, ‘The Mask of Sanity’, on psychopathy (Salekin et al. 2001). Most previous work on psychopaths has been written about criminal populations (Harris et al. 1991; Patrick et al. 1993; Reidy et al. 2011). In the general population there is an incidence of 1% of psychopaths, the incidence in the business world is 3.9% (Babiak et al. 2010), while in prison populations this is estimated at around 15% (Lipman 2013). This trait of the Dark Triad personality is seen as the most negative (Heijnis 2009).

Having psychopaths in a company could lead to problems, especially when the psychopath has a higher rank or function (Boddy 2010; 2011). Psychopaths have a lack of guilt and empathic concern for others (Glenn et al. 2011) and might be inclined to do whatever it takes to achieve certain goals (Jensen 2003), which could have devastating consequences for firms. An individual without the lack of guilt prevalent in psychopaths would choose altruistic actions and hence would cooperate more with others (Ketelaar & Au 2003). As this is not the case, a psychopath may make decisions entirely on his own, with potentially disastrous results.

Taking into account the description of psychopaths by Glenn et al. (2011) in which they refer to psychopaths as being risk-takers and failing to plan for the future, it could be expected that psychopaths prefer to work at a firm with a variable pay system, because by receiving such pay they would risk not earning the same pay in the next pay cycle (Madhani 2011; Kuhn & Yockey 2003). Another reason why psychopaths could prefer variable pay over fixed pay is because they are often found in leadership positions (Babiak & Hare 2006; Babiak et al. 2010; Gudmundsson & Southey 2011; Lipman 2013), and there most employees receive variable pay (Jensen & Murphy 1990; Gerhart & Fang 2014). Therefore the following hypothesis follows from this line of argument:

Hypothesis 3: The intention to join a firm is higher when a firm offers fixed pay than when it offers variable pay. This relation will be reduced for those individuals scoring high on psychopathy.

2.3.3 Narcissism

Narcissism, just like Machiavellianism and psychopathy, is a personality disorder (Cooper & Ronningstam 1992). The term narcissism comes from the Greek myth of Narcissus, who fell in love

(15)

Page | 15 with his self-reflection in the water and then stayed there to gaze at it until he died (Lubit 2002; Salman Akhtar & Thomson, 1982). This thesis uses the following definition from Brunell et al. (2008, p. 1664) for narcissism in which he characterized narcissism into three basics:

“(a) positive and inflated views of the self, (b) a pervasive pattern of self-regulation that maintains positive self-views—often at the expense of others, and (c) interpersonal relationships that lack warmth and intimacy.”

Other literature on narcissism has described narcissists as individuals that will do almost anything, and for whom it is very important to enhance their positive self-view (specifically to achieve status and esteem) (Campbell et al. 2004, p.298). Narcissists also tend to see themselves as being better than others and like to exploit others (Greenhalgh & Gilkey 1997), they also like to be admired by others (Brunell et al. 2008), just like individuals that score high on Machiavellianism. This could explain why narcissists, like psychopaths, are often found in positions of leadership (Brunell et al. 2008). Consequences for firm performance when narcissists are in leadership, such as in executive positions have been studied by Chatterjee & Hambrick (2007, p.351) who found that those firms either performed extremely well or experienced a multitude of failures and problems. Next to this, these firms experienced extreme fluctuations in their performance. Reasons for these extreme fluctuations in firm performance and the big losses and wins for these companies with narcissistic CEOs can be explained by the preference of these CEOs for “bold actions that attract attention” (ibid. p.376). This could also be a reason for why people that score high on narcissism have a higher intention to join a firm that offers variable pay. Because if everyone in the firm would earn fixed pay, there would be no way for narcissists to distinguish themselves from others in the form of pay.

Another reason why narcissists might be more attracted to work in firms that offer variable pay can be provided by a study conducted by Campbell et al. (2004, p. 297). This study found that narcissism predicted overconfidence. People with a high level of overconfidence are more likely to take risky actions (Nosic & Weber 2010) related to risky behaviour, which is associated with variable pay (Kuhn & Yockey 2003). Therefore the following hypothesis follows from this line of argument:

Hypothesis 4: The intention to join a firm is higher when a firm offers fixed pay than when it offers variable pay. This relation will be reduced for those individuals scoring high on narcissism.

(16)

Page | 16

2.3.4 The Dark Triad personalities

The Dark Triad personalities refer to Machiavellianism, narcissism, and psychopathy. The concept was first introduced in an article by Paulhus and Williams in 2002. The Dark Triad personalities are linked to a wide range of negative outcomes (Crysel et al. 2013; Griffith et al. 2009). According to Paulhus and Williams (2002, p. 557) the Dark Triad personalities share a number of features, namely they all entail a socially malevolent character with behaviour tendencies toward self-promotion, emotional coldness, duplicity, and aggressiveness. This malevolency directly affects interpersonal behaviour (O’Boyle, et al. 2012). The Dark Triad personalities are also associated with a willingness to gain at the expense of others (Jones 2013).

In a work environment, increases in Machiavellianism and psychopathy are associated with a reduction in the quality of job performance; however, this can be managed by contextual factors such as authority and culture (O’Boyle et al. 2012). What makes the concept of the Dark Triad important for businesses is that they are associated with poor business outcomes and working behaviours (Jones 2013), and can be harmful for the organization (Griffith et al. 2009; Heijnis 2009; Crysel et al. 2013). However, certain traits of the Dark Triad can be linked to good performance: often people with Dark Triad personalities have high functions in organizations. Especially narcissists and psychopaths are found in high functions in organizations such as in boards and as CEOs (Brunell et al. 2008; Gudmundsson & Southey 2011; Chatterjee & Hambrick 2007). Researchers have come up with several reasons for this. What they all have in common in their answers is that in order to rise in the ranks of a company, an employee should be able to favourably impress the direct manager (Boddy 2010, 2011; Brunell et al. 2008; Griffith et al. 2011). In order to be able to do this individuals should possess certain traits, and several of these traits are found in the Dark Triad personalities. For example, in the characteristics of psychopaths, charm and grandiosity could be mistaken for vision or confidence (Lipman 2013; Babiak et al. 2010). These two characteristics are important traits to be able to impress your supervisor. Psychopaths have even been referred by Babiak & Hare (2006) as ‘the masters of impression management’. Machiavellians are very good at manipulation.

Taking all information regarding the Dark Triad personalities together, we could say that individuals scoring high on the Dark Triad might be more inclined to join a firm with a variable pay reward system. Reasons for this preference are the various traits that are associated with the individual personalities underlying the Dark Triad personality. Psychopaths are the risk-takers and fail to plan for the future (Glenn et al. 2011). Narcissists have high levels of overconfidence and are therefore also more likely to take risky actions (Nosic & Weber 2010), which according to Kuhn & Yockey (2003)

(17)

Page | 17 is associated with variable pay. Moreover, Machavellians like to be praised by others (Pandey & Rastogi 1979), and excelling in their work and thereby getting more money for their performance will probably lead to them being praised by their colleagues and even their supervisor. Next to all the reasons for the individual personalities, combined they can be very dangerous to a firm. Also, they are often found in leadership positions and there the usual reward system is that of variable pay (Babiak & Hare 2006; Babiak et al. 2010; Gudmundsson & Southey 2011; Lipman 2013; Jensen & Murphy 1990; Gerhart & Fang 2014). Therefore it will probably not come as a surprise that the following hypothesis for the Dark Triad personalities of argument:

Hypothesis 5: The intention to join a firm is higher when a firm offers fixed pay than when it offers variable pay. This relation will be reduced for those individuals scoring high on the Dark Triad.

(18)

Page | 18

2.4 Conceptual model

In the previous sections literature on the intentions to join a firm, reward systems and the Dark Triad personalities have been discussed and hypotheses for this study have been given. In this section the conceptual model is illustrated in Figure 1.

Figure 1: Conceptual model

H4 H3 H2 H5

Dark Triad

Machiavellianism Psychopathy Narcissism

Reward systems

Intention to join

_{a firm}

Fixed pay Variable pay

(19)

Page | 19

Chapter 3 Methodology

This chapter discusses the methods used to test the developed hypotheses. First, the sample is discussed, followed by the research design. Second, the procedure of the experiment is discussed and the operationalization of the variables. Then the manipulation check will be discussed. Last, the measures and validations of the constructs are given.

3.1 Sample

The participants for this study are higher education students and graduates from both universities of applied sciences and universities. College graduates are often used in studies researching job applicants (Heilman et al. 1988; Powel 1984). College graduates have also been used in studies on Machiavellism (McHoskey 2001), narcissism (Twenge et al. 2008; Penney & Spector 2002) and psychopathy (Williams et al. 2007; Judge & Cable 1997). Participation for the study is voluntary and the given answers are confidential. The study was sent to 8399 students from the Faculty of Behaviorial studies of the University of Amsterdam.

3.2 Research design

In order to test the hypotheses, a quantitative vignette study was used to collect the data (see appendix 1). This experiment-type consists of a traditional survey and a vignette experiment (Atzmüller & Steiner 2010). Vignettes enable the researcher “[…] to identify and assess the importance of those vignette factors which causally affect individual responses to the contextualized but hypothetical vignette settings” (ibid., p.129). Those vignette factors are the independent variables - the rewarding system - and are used to measure their influence on the dependent variable - the intention to join the firm. Thus, by using vignettes, participants can be presented with “simulations of real events which can be used in research studies to elicit subjects' knowledge, attitudes or opinions according to how they state they would behave in the hypothetical situation depicted” (Gould 1996, p. 207). Therefore this design was chosen to measure the respondents’ intentions. Vignettes produce valid and reliable measures for finding out the respondent’s opinion (Alexander & Becker 1978, p. 93). However according to Stolte (1994) a vignette experiment sometimes can turn into an unreliable measure, because it could encourage satisficing, which could lead to respondents reading the vignette less carefully then they would under the ideal conditions

(20)

Page | 20 (Stolte 1994; Krosnick 1991; Simon 1957). In order to make the vignettes in the study more reliable, some manipulation check questions were included in the survey (see appendix 1).

For the study two vignettes have been developed about a hypothetical firm wherein all variables, except one, regarding the environment of the firm, location, firm size and other firm-related variables which could also have an effect on a respondent’s intention to join a particular firm are kept constant. The reason for this is because previous research has shown that there are different variables that can influence a person’s choice to join a firm (Chapman et al. 2005; Kochanski & Ledford 2001). The only manipulated variable that sets the two vignettes apart from each other is the reward system that the firm offers, because pay is one of the few factors that applicants keep in mind when setting out standards for deciding on whether or not to apply for a job (Reynolds 1951). The type of vignette experiment that has been used is a between-subjects design, because each participant was presented with one of the two vignettes (Atzmüller & Steiner 2010), in order not to bias decision making.

The traditional survey, which is also part of the quantitative vignette study, enabled the researcher to ask for demographic information about the respondents (e.g. information regarding their background, current status of employment, living situation) and the independent variable, the Dark Triad personalities. The traditional survey also has high external validity, which refers to the generalizability of research findings (Lucas 2003, p.238). In order to increase the validity for this measure in the study all items in the survey, except for the demographics questions, have been used from already existing scales that have been tested and validated.

3.3 Procedure

Participants were approached via Facebook and personal contacts, and a large number were approached through the university e-mail of the University of Amsterdam. Reasons for using an online over a paper-based vignette study is because it is faster, cheaper and the results can be analyzed immediately. Next, online surveys are quite appealing for the target group of the study and Internet usage under them grows exponentially (Sax et al., 2003). However collecting data online also has certain disadvantages, since people that do not have Internet access will be excluded from the experiment (Gjestland 1996).

By also including higher education students and people from the researcher’s personal network it was hoped to obtain a large number of respondents. A large number of respondents facilitate generalizations of the results (Polit & Beck 2010). The questionnaire was distributed via Qualtrics, because that program enabled the even division of the two vignettes among all the respondents and sending the respondents an online link.

(21)

Page | 21 In the experiment, respondents were first presented with the vignette and then asked to indicate their attraction for that organization. When they had completed this, the twelve-item Dark Triad measure (Jonason & Webster 2010) was presented. Finally, demographic information was requested as outlined in the above.

3.4 Operationalization of the variables

The used variables have been operationalized before in other studies and have thus been validated. Therefore existing measures have been used to test the hypothesis.

Dependent variable. The dependent variable was the intention to join a firm; this was measured via the 4-item scale from Greening and Turban (2000). In this scale respondents were asked to rate the following statements: “I would put in a great deal of effort to work for ABC”, “I would be interested in pursuing a job application with ABC”, “I am likely to send my resume (CV) to ABC” and “I am likely to accept a job offer from ABC”. In the vignettes ‘ABC’ has been changed to company X. Respondents responded on a 7-point Likert scale ranging from 1 “Not true for me” to 7 “Very true for me”. In order to increase the reliability and validity of the survey all points on the scale have been labelled with words, so that not only the endpoints have been labelled with words (Krosnick 1999; Krosnick & Berent 1993; Peters & McCormick 1966).

Independent variable. The independent variable in the study was the reward systems that the firm offered. This variable was manipulated in the study by having two vignettes with one describing a firm offering a fixed reward system and one offering a variable reward system. In the latter the variable reward system was based on individual performance. The two vignettes can be found in appendix 1.

Moderator. The Dark Triad personality is the moderator and has been measured via the twelve-item Dark Triad scale also known as the ‘Dirty Dozen’ (Jonason & Webster 2010). This measure makes use of a nine-point Likert scale ranging from 1 “strongly disagree” to 9 “strongly agree”. In this study the Likert scale has been shortened to one ranging from 1 “strongly disagree” to seven “strongly agree”, because the researcher believed that it would make it easier for students to choose from a seven- point Likert scale then from a nine-point Likert scale. Next, the researcher did not want to confuse respondents by showing them two different Likert scales, since the used Likert scale for the dependent variable also had seven options. In order to increase the reliability and validity of the scale, all points on the scale have been labelled with words, so not only the endpoints have been

(22)

Page | 22 labelled with words (Krosnick 1999; Krosnick & Berent 1993; Peters & McCormick 1966). The first four questions of the Dark Triad scale are related to Machiavellism, the second four questions are related to psychopathy and the last four questions are related to narcissism.

Control variables: Since there are many other variables that might influence a job applicants’ decision on whether or not to join a firm; some of them have been controlled in this study in order to make clear predictions on the influence of a reward system on the intention to join a firm. The control variables used in this study are the following: age, gender, nationality, labour market experience, education level and major.

Age

Respondents have been asked to indicate their birth year, as the older participants will be more likely to have worked in firms with variable and fixed pay. In addition, preferences for job attributes can change during a person’s life (Tolbert & Moen 1998).

Gender

The reason for using gender as a controlling variable is that women and men consider different factors when deciding on whether or not to join a company (Jurgensen 1978; Bigoness 1988).

Nationality

Respondents have been controlled for nationality because the choices that people make are influenced by the cultural environment they were raised in (Hofstede 1984). Hofstede (1984) found that people in western countries, e.g. countries in Western Europe, are more individualistic, which makes them prioritize self-interests and find individual success important. Therefore individuals in the western countries might be more inclined to work for variable pay based on individual performance, because that will enable them to demonstrate their individual performance (as compared to that of other employees).

Labour market experience

Questions related to labour market experience related to years of working experience; whether they have ever worked for variable and/or fixed pay before and whether it motivated them to perform their job as well as possible. If respondents had worked for a company before where they were offered one of the two, and had a positive experience with one of them, this might increase their intention to join a firm that also offers the same type of reward system.

(23)

Page | 23 Education level and major

Questions related to education were about the level of education they have completed, whether they were still students, their year of study and the subject of their major.

3.4 Manipulation checks

A manipulation check had been designed to see whether the respondents clearly understood all the questions and whether they had read the vignette. Three manipulation check questions were asked; two of them were about the reward system that the firm in the vignette offered and whether the firm offered a job. The manipulation check questions were asked immediately after the respondents had answered the questions about the vignette.

3.5 Measurements and validation of constructs

Prior to testing the hypotheses the measures have been tested. This is done by performing an exploratory factor analysis and reliability analysis. A reliability analysis tests the internal validity of items intended to measure a construct by calculating a Cronbach’s Alpha. A Cronbach’s Alpha gives a number between 0 and 1. A number of ≥.7 is considered acceptable in the literature (Cortina 1993). An exploratory factor analysis is used to identify dimensions of a test (Tate 2003).

A reliability analysis was carried out for all the items intended to measure the different variables related to the Dark Triad personalities – the Dark Triad, psychopathy, narcissism, Machiavellism – to assess the internal consistency. In table 1 an overview is presented of all the items intended to measure the different variables and their Cronbach’s Alpha. The Cronbach‘s Alpha coefficients for the Dirty Dozen, narcissism and Machiavellism was calculated above .7, and therefore had a good internal validity. However the Cronbach’s Alpha for the four items intended to measure psychopathy was calculated at .688. The reliability analysis suggested that the Cronbach’s Alpha for this variable would only become higher if the item “I tend to be cynical” would be deleted, the Cronbach’s Alpha would then become .757. However even though the change in the Cronbach’s Alpha would be significant, the researcher had decided to keep the item in the dataset, as was proposed in the literature (Jonason & Webster 2010). Scale scores were then calculated by averaging all the items intended to measure the variables. A score of 1 on the scales related to the Dark Triad personalities meant that respondents believe that the measured variable is not true for them and a score of 7 means that they believe that it is very true for them, and therefore also means that they score high on Dark Triad personalities.

(24)

Page | 24 Subsequently, an exploratory factor analysis was performed to see whether the twelve items for the Dark Triad personality scale would fall into three factors, because Jonason & Webster (2010) have said that the items would fall into three factors. They also have said that the factor loadings for the items of the Dark Triad personality should be ≥.30, therefore small coefficients have been suppressed when they had a value of below .30. Of the twelve items on the Dark Triad scale, the first four items are related to Machiavellism, the second four are related to psychopathy and the last four are related to narcissism (Ibid., p. 429).

Before doing the factor analysis, the Direct Oblimin and the Kaiser-Meyer-Olkin (KMO) measure had been turned on. The latter measure “assesses the extent to which the items are psychometrically related, as determined by their intercorrelations” (Dunk 1993, p. 404). The KMO measure of Sampling Adequacy was .837, a KMO of 0.5 is acceptable (Kaiser & Rice 1974). The principal components analysis suggested that there were indeed three factors. The three factors together explained 63% of the total variance of the underlying variable, the Dark Triad personalities. Then for each factor was checked whether the proposed items by Jonason & Webster (2010) were in the right factor. Unfortunately the factor analysis did not show the same results as described in the literature, see table 1. The factor analysis showed that all the four items of narcissism were put together in one dimension, factor 2. However item three was also placed in the dimension of narcissism, but the factor loading was not so big for this dimension. Item three also had a factor loading of ≥.30 for the Machiavellissm scale and therefore was the only variable that came across in two dimensions. Other items that had a factor loading of ≥.30 for the Machiavellism scale were the first three items, see factor 3, and item eight which should not have had a high factor loading on this scale, but for the scale of psychopathy. Item four which should have had a factor loading of ≥.30 for Machiavellism had a factor loading which was just under .3 (-.269) and therefore was not assigned to this scale. Item four however did have a factor loading of ≥.30 for psychopathy. Other items for the psychopathy scale were item five to seven, item eight which should have had a high enough factor loading for this scale did not have it, see factor 1. Item eight was also the item that caused the Cronbach’s Alpha to be lower than .7 for the items intended to measure psychopathy. However the researcher decided to retain the proposed factors and their items as has suggested by Jonason & Webster (2010).

(25)

Page | 25

Table 1 Principle components analysis using Oblique rotation.

Items

Factors

1 2 3

1. I tend to manipulate others to get my way (M) ,270 -,005 -,654

2. I have used deceit or lied to get my way (M) ,101 -,084 -,831

3. I have used flattery to get my way (M) -,106 ,370 -,543

4. I tend to exploit others towards my own end (M) ,585 ,132 -,269

5. I tend to lack remorse (P) ,859 ,040 ,033

6. I tend to be unconcerned with the morality of my actions (P) ,828 -,040 ,002

7. I tend to be callous or insensitive (P) ,707 -,003 -,025

8. I tend to be cynical (P) _-,035 _{-,017 -,670}

9. I tend to want others to admire me (N) -,112 ,787 -,147

10. I tend to want others to pay attention to me (N) ,047 ,910 ,157

11. I tend to seek prestige or status (N) -,050 ,833 -,020

12. I tend to expect special favours from others (N) ,239 ,680 -,011

Note: The numbers are factor loadings. Factor loadings ≥ |.30|are shown in bold.

Note: in the article of Jonason & Webster (2010), questions 1-4 of the Dark Triad scale are related to Machiavellism, questions 5-8 are related to psychopathy and questions 9-12 are related to narcissism. In the table this is shown by putting the first letter of the personalities in brackets behind the item.

A reliability analysis of the four items intended to measure the intention to join the firm was also carried out to assess the internal consistency. A Cronbach’s Alpha of .95 was calculated. An exploratory factor analysis was also performed for these four items. The KMO measure of Sampling Adequacy was .868 and therefore acceptable (Kaiser & Rice 1974). The factor analysis revealed that there was only one factor. This factor explained 87.1% of the variance in the underlying variable. A composite measure was then created to test the hypotheses on intention to join the firm by averaging the mean ratings of the four items.

(26)

Page | 26 Table 2. Reliability statistics

Variable N Items Cronbach’s Alpha

Intention to join a firm

3 1. I would put in a great deal of effort to work for company X

2. I would be interested in pursuing a job application with company X 3. I am likely to send my resume (CV) to company X

4. I am likely to accept a job offer from company X

.950 Dark Triad personalities (N=12) M P* N 4 4 4

1. I tend to manipulate others to get my way 2. I have used deceit or lied to get my way 3. I have used flattery to get my way

4. I tend to exploit others towards my own end 1. I tend to lack remorse

2. I tend to be unconcerned with the morality of my actions 3. I tend to be callous or insensitive

4. (I tend to be cynical)

1. I tend to want others to admire me

2. I tend to want others to pay attention to me 3. I tend to seek prestige or status

4. I tend to expect special favors from others

.837** .763

. 688 (.755)

.832

Note: M = Machiavellianism, P = pscychopathy, N = narcissism.

*The items in brackets are for the item and Cronbach’s Alpha what would have been if that item were to be excluded for the variable psychopathy

** This is the Cronbach’s Alpha for the Dark Triad personalities

Chapter 4 Results

Before running analyses to test the hypothesis, a series of preliminary tests had to be performed to identify and delete possible errors or abnormalities. First, missing data had to be handled with. All individuals that did not complete the questionnaire were excluded from the dataset. Then their answers on the manipulation check questions have been checked, and individuals that answered wrongly were also excluded. Then the dataset was cleaned by excluding those respondents that had missing values on all the items intended to measure intention to join the firm, the Dark Triad personalities and the control variables - ‘gender’, ‘birth year’, ‘nationality’, ‘labour market experience’, ‘currently a student’, ‘experience variable pay’, ‘experience fixed pay’, ‘major’, ‘highest completed education’, ‘years of work experience’ and ‘worked full-time’. Furthermore, the variables ‘gender’, ‘reward system’, ‘currently a student’, ‘experience fixed pay’, ‘experience variable pay’ and ‘worked full-time’ had to be recoded. For ‘gender’ a 0 meant that the individual was a female and 1 for male. This variable was then renamed to ‘male’, so effects of this variable would be easier to read. ‘Reward system’ was also recoded to 0 for a fixed pay and 1 for a variable pay and was renamed to ‘variable pay’ for the same reasons as those for the variable ‘male’. All other dichotomous variables were recoded to 0 for a yes and 1 for a no. The researcher then centered the

(27)

Page | 27 predictor variables; ‘Dark Triad’, ‘Machiavellianism’, ‘psychopathy’, ‘narcissism’ and ‘valuable pay’. Centering was achieved by subtracting the mean from each score (Robinson & Schumacker, 2009: p.6), so for example ‘Mean_Mach’ was centered by the formula Mean_Mach – 2.8961 (which is its mean, see table 3). Centering of the variables was performed because all the Dark Triad personalities are highly correlated with each other because the three personalities are all part of the Dark Triad and by centering the variables the collinearity will become lower. Interaction variables are necessary because according to Jaccard & Turissi (2003, p.3) because “they are said to exist when the effect of the independent variable on the dependent variable differs depending on the value of a third variable”, which is the case in this study for the moderator variable Dark Triad. So after centering the variables, the interaction variables have been made by multiplying the centered variable ‘Dark Triad’ and the centered variable ‘variable pay’. Other interaction variables that were made were multiplied with the centered variable, variable pay and ‘Machiavellism’, ‘narcissism’ and ‘psychopathy’. The interaction variables have been made to test hypotheses two to four.

4.1 Descriptive statistics

Of all the individuals that opened the questionnaire (N = 850), 279 respondents have been used for the analysis. From them, 129 (46.2%) have seen the vignette of fixed pay and 150 (58.8%) have seen the vignette on variable pay. The sample primarily consisted of young adults of Dutch nationality (74.2%, N = 207), with a mean age of 24.9 years (SD = 4.093, Min = 18 years, Max = 50 years). The percentage of women in the study was 72.4% (N = 202). From the respondents, 96.1% (N = 268) were students at the time they filled in the questionnaire. The majority of the respondents were studying or have completed a major in behavioural studies 45.9% (N = 128). A majority of the respondents had completed their education at the level of university bachelor 43% (N = 120), followed by high school 35.5% (N = 99). From the respondents, 58.1% (N = 162) have ever worked for fixed pay and 49.1% (N = 137) of the respondents have ever worked on a variable pay basis. The mean years of work experience was between two to four years (SD = 1.616), and 44.8% (N = 125) have previously had full-time working experience. The participation in the study took place on a voluntary basis and the survey was in English.

In table 3 the means, standard deviation, minimum and maximum scores of the variables are listed.

(28)

Page | 28 Table 3. Descriptive Statistics

N Min. Max. Mean SD

Variable pay 279 0 1 ,537634 ,499

Male 279 0 1 ,28 ,448

Year of birth* 279 2 34 8,85 4,093

Currently a student 279 0 1 ,04 ,195

Highest completed education 279 2 7 4,04 1,591

Major 279 1 5 3,19 1,567

Experience with fixed pay 279 0 1 ,42 ,494

Experience with variable pay 279 0 1 ,51 ,501

Years work experience 279 1 6 4,52 1,616

Worked full-time 279 0 1 ,55 ,498

Intention to join a firm 279 1,00 7,00 5,2437 1,58148

Machiavelissm 279 1,00 6,75 2,8961 1,14484

Narcississm 279 1,00 6,50 3,8154 1,24187

Psychopathy 279 1,00 7,00 2,2921 ,94914

Dark Triad personality 279 1,50 6,56 3,4875 ,72256

Valid N (listwise) 279

Note: For the variable Male, 1 means that the respondent was male. For all other dichotomous variables a 0 means ‘yes’ and a 1 means ‘no’.

*a birth year of 2 means that the individual was born in 1996, a birth year of 3 means that the individual was born in 1995, and so on

4.2 Correlations

After having carried out the reliability and factor analysis, the correlations between the independent, dependent and control variables have been checked. The correlations are shown in table 4. Table 4 shows that variables ‘Machiavellissm’ (r.= .814; p <.01), ‘narcissism’ (r.=.718; p <.01) and ‘psychopathy’ (r. =.700; p <.01) which are intended to measure the Dark Triad personalities all are highly correlated with the Dark Triad, which is not strange since they each are part of that variable. These variables are also correlated with each other, but the only statistically significant correlations between them are not found between psychopathy and Machiavellism (r. = -.029, p = .630).The only other variable that Dark Triad highly positively correlates with is the variable ‘Male (r= .346; p <.01). The dependent variable, intention to join a firm, is statistically significantly correlated with the variable ‘major’ (r.= -.129 ;p < 0.05), ‘experience with variable pay’ (r.= -.128 ;p <.05), ‘variable pay’ (r.= -.409 ;p <.01), ‘narcissism’ (r.= .164 ;p <.01) and Dark Triad (r.= .146 ;p <.05). What is remarkable is that gender is the only control variable that correlates positively significant with all the Dark Triad personalities, ‘Machiavellism’ (r. =.268 ;p <.01), ‘narcissism’ (r. = .199 ; p <.01), ‘psychopathy’ (r. = .351 ; p <.01) and the Dark Triad (r. = .346 ; p <.01). Other variables that correlated statistically significant at p<.01 were ‘year of birth’ and ‘male’, ‘highest completed education’ and ‘currently a student’, ‘highest completed education’ and ‘year of birth, ‘major’ and ‘currently a student’,

(29)

Page | 29 ‘experience with fixed pay’ and ‘currently a student’, ‘experience with variable pay’ and ‘experience with fixed pay’, ‘years of work experience’ and ‘year of birth’, ‘years of work experience’ and ‘experience with fixed pay’, ‘worked full-time’ and ‘year of birth’, ‘worked full-time’ and ‘highest completed education’, ‘worked full-time’ and ‘experience with fixed pay’ and ‘variable pay’ with ‘experience in variable pay’.

(30)

Table 4: Bivariate Correlations

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

1. Intention to join a firm Pearson Correlation 1 Sig. (2-tailed)

2. Male Pearson Correlation -,034 1

Sig. (2-tailed) ,568

3. Year of birth Pearson Correlation ,002 ,208** 1 Sig. (2-tailed) ,967 ,000

4. Currently a student? Pearson Correlation -,046 -,001 ,071 1 Sig. (2-tailed) ,445 ,980 ,239

5. Highest completed education Pearson Correlation -,052 -,004 ,477** ,193** 1 Sig. (2-tailed) ,386 ,949 ,000 ,001

6. Major Pearson Correlation -,129* -,017 ,102 ,164** ,095 1 Sig. (2-tailed) ,032 ,775 ,089 ,006 ,112

7. Experience with fixed pay Pearson Correlation -,006 -,118* -,295** -,098 -,129* ,015 1 Sig. (2-tailed) ,923 ,048 ,000 ,104 ,031 ,805

8. Experience with variable pay Pearson Correlation -,128* -,131* ,089 ,015 ,121* -,039 -,240** 1 Sig. (2-tailed) ,033 ,028 ,137 ,806 ,043 ,519 ,000

9. Years work experience Pearson Correlation ,016 ,139* ,219** -,122* -,137* ,081 -,175** -,035 1 Sig. (2-tailed) ,789 ,020 ,000 ,041 ,022 ,177 ,003 ,564

10. Worked full-time Pearson Correlation -,065 -,105 -,408** -,114 -,202** -,068 ,357** ,067 -,125* 1 Sig. (2-tailed) ,278 ,080 ,000 ,058 ,001 ,259 ,000 ,268 ,036

11. Variable pay Pearson Correlation -,409** -,006 -,090 ,040 -,070 ,046 ,074 ,168** -,067 ,003 1 Sig. (2-tailed) ,000 ,915 ,136 ,505 ,247 ,442 ,216 ,005 ,267 ,961

12. Machiavelissm Pearson Correlation ,091 ,268** ,011 -,082 -,049 -,119* -,021 -,052 ,102 -,036 -,138* 1 Sig. (2-tailed) ,129 ,000 ,853 ,170 ,417 ,047 ,724 ,390 ,090 ,547 ,021

13. Narcissism Pearson Correlation ,164** ,199** -,060 ,004 -,001 -,069 ,061 -,064 -,030 -,021 -,124* ,438** 1 Sig. (2-tailed) ,006 ,001 ,321 ,945 ,990 ,253 ,313 ,288 ,623 ,730 ,039 ,000

14. Psychopathy Pearson Correlation ,086 ,351** ,143* ,015 ,037 -,035 -,042 -,023 ,132* -,074 -,029 ,548** ,251** 1 Sig. (2-tailed) ,152 ,000 ,017 ,799 ,537 ,561 ,489 ,708 ,028 ,217 ,630 ,000 ,000

15. Dark Triad Pearson Correlation ,146* ,346** ,052 -,004 ,010 -,103 -,027 -,060 ,078 -,075 -,128* ,814** ,718** ,700** 1 Sig. (2-tailed) ,015 ,000 ,388 ,941 ,868 ,085 ,656 ,318 ,192 ,212 ,033 ,000 ,000 ,000 *. Correlation is significant at the 0.05 level (2-tailed).

(31)

4.3 Hypotheses testing

In order to test all the hypotheses a linear regression analysis was used. A regression analysis aims to describe the relationship of a dependent variable with independent, explanatory variables (Liang & Zeger 1993, 43). Hypotheses have only been proved at p < 0.05 level. For the linear regression analysis three models were executed. In the first model only the control variables were used; in the second model the control variables, the independent variable ‘variable pay’, Machiavellism, narcissism, pyschopathy and their interaction effects were added. In the last model the variable ‘Dark Triad’ and the interaction variable Dark Triad was added. The ‘Dark Triad’ and the interaction effect of the Dark Triad was not added in model 2, so that it would not correlate with the other variables which are part of the Dark Triad – Machiavellism, narcissism and psychopathy- in this way the researcher tried to avoid multicollinearity. The second model was used to test hypotheses two, three and four. The third model was used to test hypotheses five. In table 5 the regression models and their coefficients are shown.

Table 5: Coefficients and statistics

Variable Model 1 Model 2 Model 3

B Sig. B Sig. B Sig.

(Constant) 6,182 ,000 6,363 ,000 6,382 ,000

Control variables

Male -,256 ,247 -,331 ,130 -,324 ,140

Year of birth ,014 ,633 ,003 ,919 ,003 ,916

Currently a student -,191 ,707 -,121 ,796 -,118 ,802

Highest completed education -,050 ,481 -,089 ,178 -,088 ,182

Major -,136 ,029 -,108 ,061 -,110 ,059

Experience with fixed pay -,068 ,757 ,048 ,814 ,055 ,789

Experience with variable pay -,442 ,029 -,164 ,387 -,167 ,379

Years work experience ,003 ,968 -,023 ,694 -,026 ,661

Worked full-time -,198 ,368 -,310 ,128 -,314 ,125 Independent variables Variable pay -1,258 ,000 -1,259 ,000 Machiavellism -,106 ,290 -,109 ,466 Narcissism ,150 ,062 ,136 ,308 Pscychopathy ,182 ,115 ,173 ,277

Variable pay x Machiavellism ,048 ,811 -,075 ,801

Variable pay x Narcissism ,062 ,699 -,055 ,834

Variable pay x Psychopathy ,128 ,567 ,002 ,996

Dark Triad ,023 ,957

Variable pay x Dark Triad ,466 ,577

R Square ,046 ,223 224

F 1,433 4,711 4,178

R Square Change ,046 ,178 ,001

F Change 1,433 8,563 158

(32)

Page | 32 The constant, which would mean that all the possible predictor variables were to be zero, the average intention to join a firm for model 1 would be 6.182, which is already quite high taken into account that the dependent variable, intention to join a firm has a minimum value of 1 and a maximum value of 7, see table 3. The constant for all other models were also above 6, with model two having a constant of 6.363 and model three a constant of 6.382. The regression analysis showed that the first model with the control variables explained the dependent variable, intention to join a firm, not statistically significant (F = 1.433; Sig. =.174). However by adding the other variables in model two and three, the model does become statically significant. Model two had a Sig. F. Change of .000, meaning a that there indeed is a significant effect when the variables that were added in this table in comparison to the control variables in model one. So even when the researcher controlled for the control variables, the other variables seemed to be more important in explaining the dependent variable. The regression analysis showed an R square of .046 for the first model, which means that this model explains 4.6% of the variance of the dependent variable. So 95.4% is explained by other factors which are not included in the model. The R Square increased with .177 when independent variables were included in the second model and with .178 when all the predictor variables were included. Model three explained 22.4% of the variance of the dependent variables.

For analyzing potential influences of each variable the researcher looked at the Unstandardized Coefficients, these are expressed in the table with the letter B. A positive B means that for every unit increase of the independent variable, the dependent variable – intention to join a firm - also increases. Unfortunately no statistically significant regression coefficients have been found for all the control variables in the three different models. In model two, the independent variable ‘variable pay’ was statistically significant (B = -1.258; Sig. =.000). The variable ‘variable pay’ was also the only statically significant variable in model three (B = -1.259; Sig. =.000). Therefore, consistent with the expectation that the intention to join a firm for job applicants is higher when the firm offers them a fixed pay instead of a variable pay, the first hypothesis could be approved. Thus if a firm were to go from 0 (fixed pay reward system) to 1 (variable pay system), the intention to join that particular firm would be lower, because the Unstandardized Coefficient B is negative. It can be said that there is a significant effect of individuals to join a firm with a fixed pay. In contrast to the expectations, none of the variables related to the Dark Triad personalities - narcissism’ ‘psychopathy’, ‘Dark Triad’ and ‘Machiavelliansim’ - were statistical significant. Therefore hypotheses 3, 4 and 5 had to be rejected, because those hypotheses expected a statistically significant effect. The hypothesis on Machiavellianism did not expect a statistically significant effect and therefore this hypothesis, just as hypothesis 1, can be supported by the results.

Reward systems and the intention to join firms : the influence of Dark Triad personalities