Machiavellianism and CWB: The Moderating Role of
Transformational and Transactional Leadership
Master thesis Business Studies
Student:
Krijn Verdaasdonk
Student number: 10011803
Supervisor:
F. D. Belschak
Date:
June 30
th, 2014
1 Abstract
Machiavellianism is seen as the manipulative personality (Paulhus & Williams, 2002). In this master thesis its relationship with voluntary behavior that harms the organization (CWB) was investigated. The role of transformational and transactional leadership as possible moderators in this relationship was investigated as well. High transformational leaders and high
transactional leaders were expected to weaken the relationship between Machiavellianism and CWB. Contrary to this expectation both transformational and transactional leadership actually strengthened the relationship. Possible explanations for this strengthening are discussed, as well as the implications of this finding. The results offer various openings to new research areas, of which several are elaborated upon in the last part of this thesis.
2 Table of Contents
Abstract p. 1
Introduction p. 3
- Machiavellianism p. 4
- Machiavellianism and Performance p. 6
- Machiavellianism and Extra-role Behaviors p. 7
- Leadership, CWB and Machiavellianism p. 10
- Transformational Leadership p. 11
- Transactional Leadership p. 14
Method
- Sample and Procedure p. 17
- Measures p. 18
- Analytical Strategy p. 20
Results p. 20
- Descriptives and Correlations p. 21
- Hypotheses Testing p. 22
Discussion p. 25
- Implications p. 36
- Limitations and Future Research p. 37
- Conclusion p. 39
References p. 40
Appendices
3 Introduction
Since the financial crisis which began in late 2008 and early 2009 companies and governments are looking more and more to work efficiently and save money where possible (Visser, 2012). In light of this current development, it is important to notice that
counterproductive work behavior (CWB) costs organizations billions of dollars annually (Holtz & Harold, 2013). CWB can most readily be described as voluntary behavior that harms the organization (Sacket, Berry, Wiemann & Laczo, 2006). Reducing CWB, and investigating its causes and antecedents is therefore currently more relevant than ever. The personality trait of Machiavellianism is known to be positively related to CWB (Dahler, Whitaker and Levy, 2009; Zagenczyk, Restubog, Kiewitz, Kiazad, & Tang, 2014). Investigating this personality trait and its relationship with CWB might therefore help in reducing the costs associated with CWB. In the past few decades, the construct of Machiavellianism emerged in psychology and business. It is based on Niccolò Machiavelli’s thoughts. Niccolò Machiavelli was a 16th
century Italian statesman and writer. In his book Il Principe, the Prince, he writes about how to successfully rule a princedom. The bases of his thoughts on ruling different kinds of princedoms are manipulation and deceit (Machiavelli, 1513, translated by Thomson, 1992). People who score high on Machiavellianism, tend to think cold, pragmatic, immoral and cynical, they tend to have agentic motivations and strategic long-term planning and use deceit and exploitation (Christie & Geis, 1970). The relationship of Machiavellianism with
performance seems somewhat unclear. For example, Machiavellianism seems to lead to more sales but poor management reviews (Ricks and Fraedrich, 1998). Machiavellianism also seems to have a positive relationship with harmful extra-role behaviors such as CWB (Dahler et al., 2009; Zagenczyk et al., 2014). The relationship between Machiavellianism and CWB takes a central part in this research, therefore these concepts will be explained in further detail later on.
4
The goal of this research is to investigate whether certain types of leadership can moderate the relationship between Machiavellianism and CWB. More precisely, whether transformational and/or transactional leadership can moderate this relationship. Although current literature acknowledges a relationship between Machiavellianism and CWB (Dahler et al., 2009; Zagenczyk et al., 2014), it does not yet offer answers on how to mitigate this
relationship, or on how to constrain the negative consequences. If a moderation of these leadership styles can be found, some of the negative consequences of having an employee with high Machiavellianism might be avoided. At the same time, the positive consequences, like the earlier named rise in sales, can be held intact.
Machiavellianism
Machiavellianism is seen as one of the three personalities in the dark triad of personality (Paulhus & Williams, 2002). This dark triad consists of Machiavellianism, subclinical narcissism and subclinical psychopathology. Whereas subclinical narcissism has to do with dominance, superiority, entitlement and grandiosity, subclinical psychopathology has to do with low anxiety and empathy, combined with thrill-seeking and impulsivity. Machiavellianism is in short seen as the manipulative personality (Paulhus & Williams, 2002). Although the origins of these three personality traits are different, they do share a number of features. They all have some kind of negative social character with tendencies to behave towards emotional coldness, aggressiveness, duplicity and self-promotion (Paulhus & Williams, 2002). Paulhus and Williams (2002) even proposed that these three personality traits reflect one core, bigger, personality trait, but they could not confirm this hypothesis. Rauthman and Kolar (2012) later investigated how bad this dark triad really was, and found that while narcissism was considered not that evil by others, Machiavellianism and
5
Although this manipulative personality trait is considered evil by others, it does not mean that its consequences are solely negative. Machiavellianism shares a positive
relationship with motives for control, relaxation and impression management (Walter, Anderson, & Martin, 2005; Becker & O’Hair, 2007). These motives can lead to positive consequences despite of the negativity by which Machiavellianism is generally surrounded. As stated in the introduction and will be mentioned in the next section as well,
Machiavellianism can lead to more sales (Aziz, May & Crotts, 2002; Aziz, 2004; Aziz, 2005; Ricks and Fraedrich, 1998), whereas an increase in sales is clearly a positive thing for
companies.
However, most motives that are associated with Machiavellianism are not positive at all, and are therefore unlikely to lead to positive consequences. For example,
Machiavellianism has a negative relationship with the motives for affection with leaders, prosocial values and organizational concern, as well as a positive relationship with the escape motive (Walter, Anderson, & Martin, 2005; Becker & O’Hair, 2007). These motives might explain why Machiavellianists are seen as cold, pragmatic, immoral and cynical, and why they tend to have agentic motivations, strategic long-term planning, a strong goal focus and use deceit and exploitation (Christie & Geis, 1970). Furthermore, when people score high on Machiavellianism, they are not only less truthful and manipulative, their truthfulness is also harder to judge by others. For example, in research by Geis and Moon (1981), judges could not tell whether a high Machiavellianist was lying or not.
From an evolutionary perspective, Machiavellianism can be seen as a behavioral strategy, that can be useful in some situations but not in others (Wilson, Near, & Miller, 1996). Note that in some cases the motives to escape, control, relax, or manage impressions can be useful. Not showing affect to a leader, having low organizational concern and
6
the organization. The usefulness of Machiavellianism depends on situations, although it is generally interpreted as negative in current society (Paulhus & Williams, 2002).
Machiavellianism and Performance
In business, it is important to know what Machiavellianism, when possessed by people in an organization, can do. One of the main links investigated with Machiavellianism is performance. Performance is an important goal for organizations. Without performance most companies would cease to exist due to a lack of means. Several connections between
Machiavellianism and performance are found in previous research. There seems to be a positive relationship between Machiavellianism and sales performance with stockbrokers, automobile salespersons and real estate salespersons (Aziz et al., 2002; Aziz, 2004; Aziz, 2005). However, in these researches performance was measured with self-reported
performance scores. Moreover, the results may not be high in external validity due to the sales sector they were all conducted in. One might argue that, since Machiavellianists have a high tendency to manipulate, sales increase because the durance of the relationship between a salesperson and a client is usually short. Whereas the negative consequences of manipulation perhaps have more pervasive outcomes in relationships which last longer. The results of these studies may not be interpreted as valid for different situations. Moreover, because
performance was operationalized with self-reported performance scores, Machiavellianists might be more prone to manipulate these answers as well. Note that it is called the
manipulative personality for a reason (Paulhus & Williams, 2002). Zettler and Solga (2013) controlled for this influence by measuring task performance at the colleague level. In contrast to Aziz et al. (2002), Aziz (2004), and Aziz (2005), they found a negative relationship
between Machiavellianism and task performance. Their sample can also be deemed more external valid, due to the variety of participants they recruited through a social network site.
7
In contrast to the earlier mentioned researches, the sample did not only contain participants in the sale sector. The relationship between Machiavellianism and performance outcomes which influence organizations, such as sales and task performance, seems to remain somewhat unclear. A contingency approach might be best, with respect to different sectors and different task performance outcomes. Perhaps the influence of Machiavellianism on other performance outcomes is more direct and less dependent on a contingency approach. For example,
behaviors which are outside of one’s job description are more voluntary, and therefore may be more prone to change when related to Machiavellianism.
Machiavellianism and Extra-role Behaviors
Ricks and Fraedrich (1998) found that employees with high scores on
Machiavellianism had higher sales but received poor management reviews. This in some way corroborates the findings earlier mentioned on sales (Aziz et al., 2002; Aziz, 2004; Aziz, 2005), but also sheds light on the possible negative consequences of Machiavellianism (poor management reviews). Performance can be measured, as mentioned earlier, by simply looking at sales or task performance. However, performance can also be measured in several other ways. For example, with an employee’s so called extra-role behavior. Extra-role behavior can be described as behavior that does not match ones job description, but does contribute to the way an employee works and is treated in an organization (Blader & Tyler, 2009). Extra-role behavior is often divided into two major kinds of behavior, organizational citizenship
behavior and counterproductive work behavior. Organizational citizenship behavior (OCB) consists of behaviors which are not in ones job description, but benefit the organization (Sacket, et al., 2006). OCB can further be divided into two sub dimensions, namely OCB towards the organization and interpersonal OCB. In the latter case, OCB is directed towards others, such as coworkers. Examples of OCB are helping colleagues with tasks or making
8
coffee for the department.
The other kind of extra-role behavior is called counterproductive work behavior (CWB). CWB is behavior that is not reflected in ones job description, but can take place and can harm the organization (Sacket et al., 2006). Just as with OCB, CWB can be divided into two sub dimensions. Again these dimensions reflect CWB towards the organization and interpersonal CWB. Examples of CWB are theft and abuse on the work floor, or coming in late.
The relationship between Machiavellianism and extra-role behaviors seems to be an interesting one. Concerning OCB, Zettler and Solga (2013) found an inverted U-shape
relation with the two dimensions of OCB and Machiavellianism. They found that, in moderate amounts, Machiavellianism does not have a negative effect on OCB. This finding might help explain the positive effects of Machiavellianism on sales performance found earlier (Aziz et al., 2002; Aziz, 2004; Aziz, 2005; Ricks & Fraedrich, 1998) . However, the inverted U-shaped relationship implies that high amounts of Machiavellianism lead to substantially less OCB. Even more so, the inverted U-shape found by Zettler and Solga (2013) is skewed to the left. This means that when the amount of Machiavellianism increases just a little, after a certain point OCB goes down fast. Moreover, the overall effect of Machiavellianism on OCB was found to be negative (Zettler & Solga, 2013).
When it comes to CWB, the implications of these findings might not be directly clear. Intuitively one can argue that OCB and CWB are opposites of one another, resulting in a U-shaped relationship between Machiavellianism and CWB. However, OCB and CWB cannot be placed on the opposite ends of a single continuum (Sacket et al., 2006). They are separate concepts, but do however negatively correlate with each other. This in the sense that when OCB is high, CWB is likely to be lower (Sacket et al., 2006). Dahler et al. (2009), just as Zagenczyk et al. (2014) also claim in their articles that research on Machiavellianism
9
consistently indicates that people who score high on Machiavellianism are more likely to take part in behaviors which can be associated with or described as CWB. A positive relationship between employee Machiavellianism and CWB is expected because OCB correlates
negatively with CWB (Sacket et al., 2006), and Machiavellianism leads to an overall
decrease in OCB (Zettler & Solga, 2013). Furthermore, in the research of Dahler et al. (2009) and Zagenczyk et al. (2014) the researchers already stated that Machiavellianism and CWB are positively associated.
A mechanism through which this relationship might work can possibly be found in the strong goal focus Machiavellianists have (Christie & Geis, 1970). Because of this strong goal focus, Machiavellianists might think that everything is justified to achieve their goal,
including CWB. Literally nothing stands in their way to achieve their goal. CWB might be seen by them as simply a justified means to achieve that goal.
Another possible mechanism through which Machiavellianism might lead to CWB is that Machiavellinists do not even recognize some of their CWB behaviors as bad.
Machiavellianists tend to think cold (Christie & Geis, 1970) and therefore might not even recognize when they display CWB towards another individual. Because they think cold, they might themselves not recognize that what they do to their colleague is hurtful, while it actually qualifies as CWB. An example of such a form of CWB can be making fun of someone else (Sacket et al., 2006).
A third possible mechanism that might explain why Machiavellianism can lead to CWB can be found in the earlier mentioned motives associated with Machiavellianism. Machiavellianism correlates negatively with the motives for organizational concern and prosocial behavior (Becker & O’Hair, 2007). Because these two motives would lead to a decrease of negative behaviors towards the organization and colleagues (CWB), the absence of them will lead to no decrease, or even an increase of CWB. When someone has low
10
organizational concern and prosocial behavior, as is the case with Machiavellianists, that person is more likely to engage in CWB. Note that when OCB is low, CWB is likely to be higher, although they are not opposite ends of a single continuum (Sacket et al., 2006). On the bases of these three proposed mechanisms, along with the findings of earlier research, the following hypothesis is proposed.
Hypothesis 1: There is a positive relationship between employee Machiavellianism and CWB
When looking at this relationship, a somewhat pragmatic view may arise. When this relationship is found, does this mean that people who score high on Machiavellianism should be avoided in organizations because they will engage more in CWB? As in most relationships, there are probably moderators influencing this relationship. A moderator is a third variable that influences the strength of the relationship between two other variables (Kenny, 2014). This means that a moderator on the relationship under investigation might suppress the effect of Machiavellianism on CWB, leading to people who score high on Machiavellianism, to show not as much CWB. Hereby toning down the negative consequences of
Machiavellianism, while trying to keep the positive outcomes, such as sales. Perhaps
leadership plays such a moderating on the relationship between Machiavellianism and CWB.
Leadership, CWB and Machiavellianism
When it comes to moderations with the constructs currently under investigation, Den Hartog and Belschak (2012) already found that ethical leadership has a negative relationship with CWB, and that Machiavellianism moderates this relationship. When leaders are high Machiavellianists, the positive effects of ethical leadership are subdued. Although this
11
similar constructs. It remains unclear however if leadership can also moderate the relationship between Machiavellianism and CWB. Since Machiavellianism is a personality trait, it cannot be as easily manipulated as leadership, which is a set of behaviors performed by a leader. Therefore, in order to come to practically useful results, it is wise to look at leadership as a moderator in the relationship between Machiavellianism and CWB. This way, leaders may overcome the negative consequences associated with Machiavellianism by altering their leadership behaviors. Earlier research already shows that transformational leadership is capable of moderating the relationship between Machiavellianism and OCB (Belschak, Den Hartog & Kalshoven, 2013). As noted earlier, although OCB and CWB are not two opposites of the same continuum, they do negatively correlate to a moderate extent (Sacket et al., 2006). Therefore, it is expected that leadership will be able to moderate the relationship between Machiavellianism and CWB as well. There is a lack of research in this area in the literature. In the following sections transformational and transactional leadership will be introduced as possible moderating leadership styles on the relationship between Machiavellianism and CWB.
Transformational leadership and Machiavellianism
According to Bass (1999), transformational leadership is a form of leadership that lifts up followers’ motivation and morale, and moves them beyond the leader’s direct
self-interests. Transformational leadership emphasizes on emotions and values (Yukl, 1999) and can be best described through four different elements (Bass, 1999). These elements are
charisma or idealized influence, inspiration, intellectual stimulation and individualized
consideration. Charisma or idealized influence and inspiration refer to the leader creating a
vision for the future accompanied by setting examples of the way to get there, as well as showing determination and confidence to get there, and setting a high standard for
12
performance. Intellectual stimulation refers to the leader’s stimulation of individuals, and in this way promoting their creativity and innovation. The fourth dimension, individualized
consideration refers to the leader paying attention to the development of the employees,
hereby addressing employees’ growth (Bass,1999).
Transformational leadership, when compared to other leadership styles, is more strongly related to outcomes associated with OCB (Judge & Piccolo, 2004). A possible explanation for this relationship might emanate from a resulting commitment to the organization due to transformational leadership. In his research, Koh (1990, cited in Bass, 1996) found that commitment to the organization and related citizenship behaviors were higher for transformational leaders. More recently, Dunn, Dastoor and Sims (2012) corroborated these findings and found the same relationship between transformational leadership and affective commitment. One might argue that because transformational leadership leads to more affective commitment, whereas affective commitment is known to lead to less CWB (Gill, Meyer, Lee, Shin & Yoon, 2011), transformational leadership leads to less CWB. When also taking Machiavellianism into account, it is expected that
transformational leadership subdues the relationship between Machiavellianism and CWB. Earlier research already shows that transformational leadership is capable of moderating the relationship between Machiavellianism and OCB (Belschak et al., 2013). Although OCB and CWB are not two opposites of the same continuum, they do negatively correlate to a moderate extent (Sacket et al., 2006). Therefore, it is expected that transformational leadership will be able to moderate the relationship between Machiavellianism and CWB as well. When Machiavellianists receive high transformational leadership, this is expected to lead to a decrease in their CWB. On the other hand, Machiavellianists who receive low
transformational leadership, are not expected to show a decrease in CWB.
13
relationship between Machiavellianism and CWB is through individualized consideration. Note that individualized consideration refers to the leader paying attention to the development of the employees, hereby addressing employees’ growth, and that it is one of the four
elements associated with transformational leadership (Bass, 1999). Machiavellianists on the other hand, are considered to have a high self-interest and can be called selfish (Dalton & Radtke, 2012; Belschak et al., 2013). If the transformational leader gives much attention to the self-interests of the Machiavellianist through individualized consideration, this might lead the Machiavellianist to display more positive and less negative emotions towards the leader and organization. These positive and negative emotions respectively correlate negatively and positively with CWB (Fox et al., 2001). It can be expected that transformational leadership moderates the relationship between Machiavellianism and CWB, through the element of
individualized consideration. Machiavellianists have high self-interests (Dalton and Radtke,
2012; Belschak et al., 2013), to which transformational leaders pay attention (Bass, 1999) and in turn elicit more positive and less negative emotions, leading to less CWB (Fox et al., 2001). Building on the findings of earlier research, and taking the two proposed mechanisms into account, the second hypothesis studied in this research is stated as the following.
Hypothesis 2: Transformational leadership moderates the relationship between employee
Machiavellianism and CWB. High transformational leadership leads to a
weaker relationship between Machiavellianism and CWB, whereas low
transformational leadership leads to no weakening of the relationship.
The moderation model reflecting the second hypothesis is depicted in Figure 1 below. In the next section of this theoretical framework another leadership style that possibly moderates the
14
relationship between Machiavellianism and CWB will be discussed, namely transactional leadership.
Figure 1. The proposed moderation model.
Transactional leadership and Machiavellianism
Transactional leadership can best be described as a form of leadership that focuses on the exchange relationship between a leader and follower. Self interests of both leader and follower are central in this leadership style (Bass,1999). Just as with transformational leadership, transactional leadership can be divided into several sub dimensions. The three dimensions of transactional leadership are management by exception – active, management by
exception – passive and contingent reward (Judge & Piccolo, 2004). Contingent reward
refers to the degree in which a leader clarifies what rewards subordinates can expect when they meet certain expectations. Management by exception generally refers to the degree in which a leader takes corrective measures on the basis of outcomes of transactions between a leader and follower. More specifically, the management by exception-active dimension refers to when the leader takes this corrective measure before a behavior leads to serious problems.
Management by exception-passive however, refers to when a leader takes this corrective
measure after the behavior has already lead to a problem (Judge & Piccolo, 2004).
15
followers on leaders (Bass, 1998, cited in Judge & Piccolo, 2004). Although it seems like this leadership style might be fundamentally opposite to transformational leadership,
transformational and transactional leadership are not on opposite ends of a continuum, they are not mutually exclusive (Bass, 1985). This is an important notion for the current research, because else both leadership styles would not be able to moderate the same relationship in a similar way.
People who score high on Machiavellianism, tend to think cold, pragmatic, immoral and cynical, they tend to have agentic motivations, a strategic long-term planning and use deceit and exploitation (Christie & Geis, 1970). Since they are often described as people who do not give much attention to emotions, they might benefit more from clear direction. The emotions that go along with other forms of leadership may then be interpreted as less important. The structure that transactional leadership offers might lead a person who scores high on Machiavellianism to display more behavior aimed at achieving for the organization as a result. Dalton and Radtke (2012) argue that because high Machiavellianists are concerned with the economic outcomes of their decisions and have a high self-interest, they will be more prone to rewards and sanctions. Transactional leadership offers clarity as to what rewards or sanctions one gets for certain desired behaviors. Moreover, high Machiavellianists even showed to prefer forming transactional psychological contracts with their leader (Zagenczyk et al., 2014). These transactional psychological contracts also mediated the relationship between Machiavellianism and CWB (Zagenczyk et al., 2014). In this sense, people who score high on Machiavellianism are expected to show less CWB when transactional
leadership is exerted, due to the provision of structure and clarity of rewards and sanctions. People who score high on Machiavellianism tend to care more about structure than about emotions (Christie & Geis, 1970; Paulhus & Williams, 2002).
16
relationship between Machiavellianism and CWB can be found in trust. As noted earlier, transactional leadership can lead to trust of followers in leaders (Bass, 1998, cited in Judge & Piccolo, 2004). Furthermore, trust is negatively associated with CWB (Demir, 2011). In this way, a negative relationship between transactional leadership and CWB can be expected through trust. For Machiavellianists, transactional leadership is expected to also lead to a heightening of trust in the leader, since no difference in initial trust display exists between high and low Machiavellianists (Gunnthorsdottir, McCabe & Smith, 2002). Trust, in turn, results in less CWB (Demir, 2011). With earlier research and the two proposed mechanisms in mind, the third hypothesis is proposed.
Hypothesis 3: Transactional leadership moderates the relationship between employee
Machiavellianism and CWB. High transactional leadership will lead to a
weaker relationship between Machiavellianism and CWB, whereas low
transactional leadership will not lead to a weakening of this relationship.
Following from the third hypothesis is the moderation model depicted below in Figure 2. In the next part of this master thesis the proposed hypotheses will be tested on the basis of empirical research.
17 Method
Sample and Procedure
In the current research, data was collected through survey research. Participants were people who knew the researcher either directly or indirectly. These were all Dutch speaking natives, so the survey was in Dutch as well. Potential respondents either received an email with a link to Qualtrics survey software or a paper form of the survey from the researcher. They were asked to participate in a master research project concerning leadership and behavior at work. They were then asked, in the same message, to let a colleague fill in the same questionnaire as well and fill in the same unique code (provided by the researcher) in order to connect their results. The survey was set up in a way that respondents answered on items about themselves and their supervisor at first, and ended with CWB items about a certain colleague of them. People tend not to be very honest in answering about their own counterproductive work behaviors (CWB), therefore this construct was measured at the colleague level. Note that especially high Machiavellianists might not be truthful about their own CWB. That same colleague filled in the whole questionnaire as well, and used the initial colleague to base their answers on the CWB items on. An emphasis was made on the
anonymity of the results, as well as on the importance of the obtainment of the colleague’s results. Because of the sensitivity of the results for colleagues, the participants were asked to fill in the questionnaires separated from each other. Emails with a link to Qualtrics were sent to a total of 122 respondents, of which 72 filled in the questionnaire (response rate = 59%). Additionally, 42 surveys were printed, of which 40 were returned to the researcher (response rate = 95%; overall response rate = 68%). There were six respondents of which the code was filled in only once, so no data of their CWB existed. Therefore these cases were deleted from the dataset, resulting in a total of 106 respondents. Of these 106 respondents, 66 filled in the questionnaire online through Qualtrics, and 40 filled in a printed form of the questionnaire.
18
Of the 106 participating respondents, there were 56 males and 50 females. Their average age was 31.58 years (SD = 12.97), and the average tenure of the professional relationship between the colleagues that filled in each other’s CWB items was 55.59 months (SD = 49.73)
Measures
Machiavellianism was measured with an adapted form of the Mach-IV scale
developed by Christie and Geis (1970). The Mach-IV scale was translated to the Dutch language. Of the 10 items on the scale, 2 were contra-indicative and 8 indicative. The scale consisted of 10 statements on which respondents had to indicate how much they agreed with those statements. They answered on a 7-point Likert-scale ranging from ‘strongly disagree’ to ‘strongly agree’. A sample item is: ‘The best way to deal with people is to tell them what they want to hear’. A high score on the items resembles high Machiavellianism. The reliability of the adapted form of the Mach-IV scale was α = 0.63, which is not very high. This reliability would be slightly higher if one item was deleted (α = 0.65). This item did not seem to be of high conceptual importance for Machiavellianism and was therefore deleted (you should only do something, if it is morally justifiable; contra-indicative). The reliability of this scale was now α = 0.65. Since α approaches the theoretical cutoff score of .7 and the deletion of more items did not improve α, no further action was taken.
Transformational leadership was measured with the Charismatic Leadership in
Organizations questionnaire (CLIO) (De Hoogh, Den Hartog & Koopman, 2004). The CLIO consisted of 11 indicative items on which respondents answered with a 7-point Likert-scale. The items consisted of statements about the respondents’ leaders on which respondents answered on a scale from ‘strongly disagree’ to ‘strongly agree’ with the statements. A sample item is: ‘My leader talks with subordinates about what is important to them’. A high
19
score on these Likert-scales indicates a high transformational leader. The reliability of this scale was α = 0.90, which is very high. Since this is already high, and the deletion of items did not increase this value, no action was taken.
Transactional leadership was measured with the Multifactor Leadership Questionnaire
(MLQ), developed by Bass and Avolio (1989). The MLQ consisted of 8 indicative items, which were translated to the Dutch language. Respondents answered on a 7-point Likert-scale to what degree they agreed with statements made about their leader. This scale ranged from ‘strongly disagree’ to ‘strongly agree’. A sample item is: ‘my leader makes clear what’s in it for me when I do my job’. A high score on these scales indicates a high transactional leader. This scale also had a high reliability with α = 0.87. Reliability would be slightly higher if items were deleted, but has not been done since the difference was not substantial and the relevant items seemed to be of conceptual importance.
CWB was measured with Fox and Spector’s (1999) model of work frustration
aggression. The 16 indicative items used were translated to the Dutch language and were to be filled in by colleagues, because people tend not to be honest about their own shortcomings (perhaps especially Machiavellianists). The scale consisted of 16 behavioral examples on which respondents had to indicate how much this example resembled their colleague’s behavior. They again answered on a 7-point Likert-scale ranging from ‘strongly disagree’ to ‘strongly agree’. A sample item is: ‘my colleague uses working time for private matters’. A high score on the 16 items resembles high CWB. The reliability of this scale was again very high, with α = 0.90. Since the deletion of an item would only result in a slightly higher reliability, no item was deleted.
Demographics. Alongside the used constructs, some demographics were also asked.
These included age and gender, but also the tenure of the relationship between the two colleagues. Perhaps the time these colleagues worked together influences the relationships
20
between the different constructs.
Analytical Strategy
Data cleaning and missing values
A new variable was created, in which the answers on the CWB items were matched to the right respondents (other respondent with the same unique code). Due to the option to request (not demand) an answer on every question before going to the next page, there were no missing values in the cases that came in through Qualtrics. In the 40 questionnaires filled in on paper, there were only two missing values, both in the CWB questionnaire. In order to use Hotdeck, correlations of these two items were addressed with four items that measured autonomy, and six items that measured organizational commitment. Only two of these ten items significantly correlated with only one of the two CWB items that had a missing value. Therefore one Hotdeck item was created, for the seventh CWB item, resulting in eventually only one value missing in the total data set. Pairwise deletion was used in the remainder of the research due to this small number.
Throughout the whole questionnaire, only two items where counter indicative. These were the seventh and ninth item of an adapted form of the Mach-IV scale developed by Christie and Geis (1970). These items were recoded. This way, when participants scored high on a counter-indicative item, their score for the construct was depicted low, as intended.
Results
In the following section the results from this study are presented. First of all, an overview of the descriptives and different correlations between the variables used will be given, after which the three hypotheses, which were proposed in the theoretical framework, will be tested. This testing will be done with regression analyses.
21
Descriptives and Correlations
In Table 1 the means, standard deviations and correlations of the different theoretical
constructs measured are depicted, as well as age, gender and the tenure of the professional relationship between the colleagues that filled in the CWB items about each other. The
reliabilities are also displayed on the diagonal. A few of the correlations are significant. There is a negative significant correlation between Machiavellianism and age (r = -.24, p <.05) and between transformational leadership and relation tenure (r = -.21, p <.05). There is a positive significant correlation between transformational leadership and gender (r = .22, p <.05), which means that women perceive more transformational leadership than man do (gender is coded as 1 for men and 2 for women). Some significant correlations at the p <.01 level were also found. Relation tenure correlated significantly with age (r = .47, p <.01) and transactional leadership correlated significantly with transformational leadership (r = .49, p <.01). Four negative correlations, also significant at the p <.01 level were found. There was a negative correlation between Machiavellianism and gender (r = -.43, p <.01), meaning that men scored higher on Machiavellianism. The second negative correlation was the one between
Machiavellianism and transformational leadership (r = -.27, p <.01), the third between transactional leadership and age (r = -.33, p <.01). The fourth and last negative correlation was the one between CWB and gender (r = -.28, p <.01), which shows that men display more CWB.
For the analyses that follow in the next paragraph, it is important to know that since gender, age and relationship tenure correlate with some of the constructs under investigation, they should be controlled for. The implications of and possible explanations for the different correlations found will be discussed in further depth in the discussion.
22 Table 1: Means, Standard Deviations, Correlations
Variables M SD 1 2 3 4 5 6 7 1 . Age 31.58 12.97 - 2 . Gender 1.47 0.50 .18 - 3 . Relation Tenure (months) 55.59 49.73 .47** -.04 - 4 . Machiavellianism 2.92 0.69 -.24* -.43** .16 (.65) 5 . Transformational Leadership 4.79 1.01 -.07 .22* -.21* -.27** (.90) 6 . Transactional Leadership 4.14 1.03 -.33** -.06 -.13 .06 .49** (.87) 7 . Counterproductive Work Behaviors 1.81 0.71 -.19 -.28** .02 .17 -.17 .05 (.90) **. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).
Hypotheses testing
According to the first hypothesis, Machiavellianism predicts CWB. In order to test this
first hypothesis, a linear regression analysis was conducted. This linear regression analysis controlled for age, gender and relationship tenure. The analysis showed that Machiavellianism does not significantly predict CWB (β = .02, p = .896). The first hypothesis could therefore not be supported.
The second hypothesis stated that transformational leadership moderated the relationship between Machiavellianism and CWB. In order to test this hypothesis the
PROCESS plug-in to SPSS by Hayes (2012) was used. PROCESS showed that the coefficient with which the effect of Machiavellianism on CWB changed, when transformational
leadership changes, was β =0.18, with a significance level of p = .053. The proportion of the total variance of CWB uniquely attributable to the interaction of Machiavellianism with transformational leadership is 3 per cent and marginally significant, p = .053. PROCESS further showed that for the very low (lowest 10%), low (lowest 25%) and moderate (50%)
23
levels of transformational leadership, there is no significant effect of Machiavellianism on CWB (p = .475; p = .785; p = .079). However, for high (75% and higher) and very high (90% and higher) levels of transformational leadership, there is a significant positive effect of Machiavellianism on CWB (β =.30, p = .022; β =.38, p = .012). This means that although transformational leadership moderates the relationship, high levels of transformational leadership lead to a stronger positive relationship between Machiavellianism and CWB, instead of a weaker relationship. A visual representation of this interaction effect is displayed in Figure 3. Transformational leadership did moderate the relationship between
Machiavellianism and CWB. However, since the effect was in the opposite direction, the second hypothesis could not be supported.
Figure 3. The interaction effect of transformational leadership and Machiavellianism for
CWB.
The third hypothesis stated that transactional leadership moderated the relationship between Machiavellianism and CWB. PROCESS showed that the coefficient with which the effect of Machiavellianism on CWB changed, when transactional leadership changes, was
β =0.13, with a significance level of p = .206. The proportion of the total variance of CWB 3 3,5 4 4,5 5 5,5 6
Low Machiavellianism High Machiavellianism
C
WB
Low Transformational Leadership High Transformational leadership24
uniquely attributable to the interaction of Machiavellianism with transformational leadership is 2 per cent and not significant (p = .206). PROCESS further showed that for the very low (lowest 10%), low (lowest 25%) and moderate (50%) levels of transactional leadership, there is no significant effect of Machiavellianism on CWB (p = .724; p = .404; p = .056).
However, for high (75% and higher) and very high (90% and higher) levels of transactional leadership, there is a significant positive effect of Machiavellianism on CWB (β =.29, p = .032; β =.37, p = .045). Just as with transformational leadership, high and very high levels of transactional leadership lead to a strengthening of the relationship between Machiavellianism and CWB. This means that although transactional leadership moderates the relationship between Machiavellianism and CWB, this moderation works in the opposite direction of what was expected. Figure 4 shows a visual representation of the interaction effect found.
Transactional leadership did moderate the relationship between Machiavellianism and CWB. However, since the effect was in the opposite direction, the third hypothesis could also not be supported.
Figure 4. The interaction effect of transactional leadership and Machiavellianism for CWB. 1,5 2 2,5 3 3,5 4
Low Machiavellianism High Machiavellianism
C
WB
Low Transactional Leadership High Transactional Leadership25 Discussion
This master thesis investigated the relationship between Machiavellianism and CWB, and the possible moderation effects of transformational and transactional leadership on this relationship. Dahler et al. (2009) as well as Zagenczyk et al. (1014) found that
Machiavellianism and CWB were positively associated. Furthermore, because OCB correlates negatively with CWB (Sacket et al., 2006), and Machiavellianism leads to an overall decrease in OCB (Zettler & Solga, 2013), the first hypothesis stated that
Machiavellianism predicts CWB. Three mechanisms were proposed through which this relationship works, goal focus, motives and the recognition of CWB. However, the findings could not be corroborated in this research and therefore no support for the first hypothesis was found.
Because others found that Machiavellianism and CWB were positively associated (Dahler et al., 2009; Zagenczyk et al., 2014), a possible explanation as to why there was no effect in the current research can perhaps be found in the methodology. In this research the choice was made to measure CWB at the colleague level. Participants filled in 17 items about their colleague and indicated to what extent their colleague engaged in CWB. Although the motivation for this method was to enhance honesty of answers on these questions, some additional problems may have come into play. One of these problems is the manipulative nature of Machiavellianists. Because Machiavellianists have a manipulative nature, they might be especially good in hiding their CWB from other colleagues or in mitigating what they are doing when they do get caught. This quality to hide ones CWB might result in lower CWB scores filled in by colleagues, simply because they do not see those behaviors. If this was indeed the case, results will definitely have been influenced by this. When colleagues would have seen these behaviors, they would have been more likely to report them.
26
for reprisal. Although during the survey anonymity was stressed multiple times, and
instructions were given to honestly answer about colleagues, some anxiety might have taken place at this point. Sometimes colleagues were probably present in the same room while filling out the questionnaire. In the case of filling out the questions about a colleague scoring high on Machiavellianism, this anxiety might be higher than when a colleague scores low on Machiavellianism. People who score high on Machiavellianism engage more in negative reciprocity than people who score low on Machiavellianism (Meyer, 1992). This means that people who score high on Machiavellianism are more prone to taking revenge when
something negative has been done to them. In this case, reporting about CWB might be seen as offensive, resulting in a mitigation of the relationship between Machiavellianism and CWB.
Perhaps publication bias plays a role in the availability of articles offering support against the first hypothesis. It is more likely that articles will be published if they find positive results instead of negative ones. Maybe there is no methodological explanation for why no relationship between Machiavellianism and CWB was found in this master thesis. Perhaps the studies which reported the relationship were flukes and no actual relationship between
Machiavellianism and CWB exists. Although this possibility argues against the whole line of thought presented in the theoretical framework, it needs to be mentioned to offer a complete image. A possible theoretical explanation as to why no relationship was found can be sought in the fact that Machiavellianists are calculating and use strategic long-term planning in their actions (Christie & Geis, 1970; Dalton & Radtke, 2012). The use of CWB simply does not really fit in this part of the definition of Machiavellianism. Although more of the
characteristics of Machiavellianism seem to stroke with the expectation of CWB, these do not. For someone who scores high on Machiavellianism, engaging in CWB is detrimental for his or her strategic long-term planning, because when caught, this will reflect badly. Although a
27
methodological imperfection is more likely to be responsible for the absence of a found relationship between Machiavellianism and CWB, theoretical ones like these have to be considered as well.
The second hypothesis stated that transformational leadership moderated the relationship between Machiavellianism and CWB. Earlier research already showed that transformational leadership was capable of moderating the relationship between
Machiavellianism and OCB (Belschak, Den Hartog & Kalshoven, 2013). Although OCB and CWB are not two opposites of the same continuum, they do negatively correlate to a moderate extent (Sacket et al., 2006). Mechanisms concerning commitment and individualized
consideration were introduced which were both able to explain the possible moderation
effect. Transformational leadership was expected to moderate the relationship between Machiavellianism and CWB. This research however, was unable to find support for the second hypothesis. Transformational leadership did moderate the relationship between Machiavellianism and CWB. However, the moderation was not in the expected direction. When transformational leadership was high or very high, the relationship between
Machiavellianism and CWB became stronger, instead of weaker as expected.
The results found concerning the second hypothesis, and the inability to support it are fundamentally different from the first hypothesis. It is not the case that no significant
relationship has been found. The effect that was found was significant, but in the opposite direction of what was expected. This finding implies that methodological imperfections might not have played the role they did concerning the first hypothesis, but that there is a theoretical fundamental difference between the second hypothesis and what the results show.
Possible explanations as to why the second hypothesis could not be supported must therefore be sought in a theoretical or hypothetical manner. A possible explanation for why transformational leadership leads to a stronger relationship between Machiavellianism and
28
CWB can possibly be found in the focus on emotions (Yukl, 1999). Whereas transformational leaders mainly focus on emotions, Machiavellianists are more easily described as emotionally cold (Paulhus & Williams, 2002). Because of this difference in communication and handling between leadership style and employee, problems may arise. The emotional coldness can explain why high transformational leaders lead Machiavellianists to display more CWB. Not recognizing the emotions might also lead to cynical thinking because people who score high on Machiavellianism tend to think cynical (Christie & Geis, 1970). Because of the emotional coldness and cynical thinking, Machiavellianists might not recognize the positive concern of the leader, and interpreted it as meddling or interfering. In turn, they might react on this meddling or interfering from the leader by showing more CWB.
The third hypothesis stated that next to transformational leadership, transactional leadership would also be able to moderate the relationship between Machiavellianism and CWB. Dalton and Radtke (2012) argued that because high Machiavellianists are concerned with the economic outcomes of their decisions and have a high self-interest, they are more prone to rewards and sanctions. Transactional leadership offers clarity about what rewards or sanctions one receives for certain desired behaviors. High Machiavellianists also prefer to form transactional psychological contracts with their leader (Zagenczyk et al., 2014). The third hypothesis, however, could also not be supported by the current research. Whereas, just like transformational leadership, transactional leadership did moderate the relationship between Machiavellianism and CWB, this moderation again was not in the expected direction. For the high and very high levels of transactional leadership, the relationship between Machiavellianism and CWB became stronger, instead of weaker as was expected. Again, because a significant effect is found in the opposite direction, an explanation should be found in a hypothetical or theoretical manner, as opposed to a methodological one. A possible explanation for the moderation effect found is that in the used sample, not all
29
CWB behaviors measured are seen as necessarily bad by leaders. If this is the case, and Machiavellianists are more prone to forming transactional psychological contracts with their leader (Zagenczyk et al., 2014), they would know that they were allowed or able to perform certain aspects of CWB. If the leader displays transactional leadership it will become clear for the Machiavellianist that he or she is allowed or able to display these behaviors. Transactional leadership would then lead to a strengthening of the relationship between Machiavellianism and CWB, as the results show. In future research it might be beneficial to look if the different CWB items measured are actually perceived as bad, and if so, how bad. In doing this,
researchers can be sure all the CWB scales are seen as a negative thing.
Another possible explanation can perhaps account for the results found concerning both the second as the third hypothesis. The amount of attention a leader gives to people that score high on Machiavellianism might explain the found moderation effects. Christie and Geis (1970) noted earlier that by close monitoring of employees by leaders, the employees’ job autonomy goes down. In turn, the number of behavioral tactics a Machiavellianist can use to achieve their goals goes down as well, resulting in lower goal attainment. Since the
Machiavellianist has a strong goal focus (Christie & Geis, 1970), and the attention interferes with the attainment of that goal, the Machiavellianist is likely to have a negative reaction to the attention. CWB can be such a negative reaction.
Another explanation that might account for why both leadership styles lead to a strengthening of the relationship between Machiavellianism and CWB has to do with the strong goal focus leadership brings as well. When leadership increases, whether it is
transformational or transactional, goals of the leader and the organization will become more clear. Note that a strong goal focus is something high Machiavellianists have (Christie & Geis, 1970). When leaders elicit goals through their leadership, this might elicit and feed the goal focus of the Machiavellianist. Moreover, the Machiavellianist has the idea that he or she
30
has to accomplish the goal set. Machiavellianists might think that everything is justified to achieve their goal, including CWB. Literally nothing stands in their way to achieve their goal. CWB might be seen by them as simply a justified means to achieve that goal.
An overall conclusion as to the moderation effects found in this master thesis might be that intensive leadership, whether it is transformational or transactional, will not work for people who score high on Machiavellianism when it comes to their display of CWB.
Although several explanations are addressed concerning interfering with the goal focus, or justifying CWB through the goal focus, autonomy seems to have a positive influence on Machiavellianists. When it comes to Machiavellianists, Belschak et al. (2013), already found that job autonomy could partially explain the relationship between transformational leadership and OCB. Furthermore, earlier research already showed that autonomy correlates negatively with CWB (Fox et al., 2001). These findings advocate for the implementation of more autonomy for Machiavellianists. Before turning to the implications however, the causality problem and the significant correlations found in the results section as well as some exploratory factor analyses will be discussed.
The research done in this master thesis was done cross-sectional. This means that no causality can be derived from the results. Based on this research, it is uncertain whether Machiavellianism leads to more CWB, when moderated by leadership, or if CWB actually leads to Machiavellianism. When the latter is the case, the same results would have been found. Although the literature points in the direction of the current research, this point of causality needs to be made. It helps offering a veracious image of the results. In future research, experimental research could offer a causal image of the relationship. Although a personality trait as Machiavellianism seems hard to investigate in an experimental manner, certain paradigms exist to investigate such constructs (Eysenck, 1997).
31
Table 1 in the results section shows the correlations between the different constructs measured in this research. Some of these correlations found can be expected, while others are more surprising. That relationship tenure is longer when people are older (r = .47, p > .01) can be expected. It seems logical that older people work longer with each other than their younger counterparts, simply because they have a longer (working) life. Relationship tenure also correlates with transformational leadership (r = -.21, p < .05). This negative correlation indicates that when tenure of the relationship between two colleagues is longer,
transformational leadership is perceived as lower. Since no prior research is done in this direction, only a speculative explanation for this correlation can be given. Perhaps behaviors of a transformational leader are relied on for certain emotional support, that can later be taken over by colleagues, when their professional relationship tenure gets longer. Future research can be done in this direction, however the direct practical relevance of this specific research area is not very clear. More interesting would be to combine this relationship tenure with organizational tenure and leader tenure. This way, the proposed explanation can be tested, without the problem of changing colleagues and leaders. In such a research, age has to be taken into account as a control variable, since it obviously influences the other variables, simply by enabling higher values.
Another interesting but in some way predictable correlation is the one between transformational and transactional leadership (r = .49, p < .01). Bass (1985), as mentioned earlier in this master thesis, already said that both leadership styles are not mutually exclusive. This correlation corroborates that notion. An explanation for this correlation can possibly be found in the amount of attention a leader spends on his or her employees. More attention might lead to more recognized aspects of both leadership styles by respondents who filled in the survey. Most of the prior research focuses on the differences between these two leadership styles, or the different effects they have on other variables. Therefore, in future research it
32
might be interesting to look at the similarities, especially since the correlation found here is fairly high.
Other correlations might be more interesting at first, but were also found in earlier research. For example, men score higher on Machiavellianism and display more CWB when compared to women (r = -.43, p < .01 and r = -.28, p < .05, respectively). This replicates the findings of Brown and Guy (1983) who already found a main effect of gender on
Machiavellianism in their research. A relationship between CWB and gender was also found in earlier research by Hyde (1984). Hyde (1984) concluded from his research that men were more prone to the exhibition of CWB than women.
Another interesting correlation is the one between age and Machiavellianism (r = -.24,
p < .05). This correlation implies that younger people are more Machiavellian, and thus more
prone to think cold, pragmatic, immoral and cynical, they tend to have more agentic motivations, strategic long-term planning and use deceit and exploitation (Christie & Geis, 1970). This correlation seems to be somewhat shocking. However, in earlier research, this relationship was already found (Christie & Geis, 1968; cited in Murray & Okanes, 1980). Age does not only correlate significantly with just Machiavellianism, but also correlates negatively with transactional leadership (r = -.33, p <.01). This means that older people experience less transactional leadership. This correlation can be looked at in two ways. On the one hand this correlation can be expected, in the sense that older people already know what is expected from them and what they expect from their leader. The leadership behaviors are deemed unnecessary and therefore not perceived. On the other hand, this correlation can be interpreted as more striking. Although older people might look for, and want transactional leadership, they receive it less because the leader perhaps thinks it is redundant. With a survey, this potential problem can be investigated. Next to a scale that measures transactional leadership, a required or wanted transactional leadership scale can be developed. If this
33
second scale also correlates negatively with age, and roughly resembles the other scale, there is no problem whatsoever. In fact, the negative correlation found in this master thesis would work beneficial for both the leader and employee. If these scales do not roughly resemble each other, but differ substantially, awareness should be created to let leaders know older employees need transactional leadership and clarity as well.
Another correlation that can be found in the matrix in Table 1 is the one between transformational leadership and gender (r = .22, p < .05). This correlation implies that women receive or perceive more transformational leadership than men. An explanation for this correlation can be found in the research of Kark, Waismel-Manor and Shamir (2012). They found that transformational leadership has more ‘feminine’ qualities and therefore female employees or subordinates are more likely to identify and recognize these behaviors. These findings can possibly partially or fully explain the found correlation between transformational leadership and gender.
When keeping the proposed second hypothesis in mind, that transformational leadership moderates the relationship between Machiavellianism and CWB, the correlation between transformational leadership and Machiavellianism (r = -.27, p < .01) might subdue the effect that can be found. This correlation implies that Machiavellianists receive less transformational leadership. Because of this, the moderating effect of transformational leadership might not be depicted as strong as it is, simply because people who score high on Machiavellianism receive less transformational leadership. However, this correlation might also show that there is another process working. Because Machiavellianists are often
described as emotionally cold (Paulhus & Williams, 2002), they might simply not recognize the emotional behaviors a transformational leader shows. If this is the case, they will perceive less transformational leadership, resulting in a negative correlation between Machiavellianism and transformational leadership.
34
Next to the earlier reported analyses used to test the different hypotheses, some exploratory factor analyses have been done as well. One exploratory factor analysis was conducted to see whether the four different dimensions of transformational leadership could be detracted from the used questionnaire. Note that the definition of transformational
leadership consists of the four elements named charisma or idealized influence, inspiration,
intellectual stimulation and individualized consideration (Bass, 1999). A principal component
factor analysis was conducted to test whether the different items loaded on four components which resembled these elements. The factor analysis could not replicate this fourfold of the transformational leadership construct. Transformational leadership was not the only construct used that had different elements in its definition.
Transactional leadership can be divided into three sub dimensions, management by
exception – active, management by exception – passive and contingent reward (Judge &
Piccolo, 2004). A principal component factor analysis was conducted to test whether the different items loaded on these three components. The factor analysis could not replicate these three components, however it did extract two components, which roughly resembled
contingent reward and management by exception in general. A two-fold in the analysis
showed the focus of one component on errors, mistakes, and achieving norms (management
by exception), while the other component showed clear transactions (contingent reward).
CWB, as noted earlier, can be divided into two sub dimensions. These dimensions reflect CWB towards the organization and interpersonal CWB (Sacket et al., 2006). To test whether these two dimensions could be found in the current research, an exploratory factor analysis was conducted. This principal component analysis lead to some interesting results. Eigenvalues, items and component loadings of this factor analysis are depicted in Table 2 below. All items loaded high on the first component. Therefore, component loadings which differed not much from their loading on the first component, are highlighted on their second
35
highest component loading. Consequently, 8 items loaded on the first component, 5 on the second, 3 on the third, and only 1 on the fourth component.
Table 2. Component loadings for 17 survey items*
Components 1 2 3 4
Trying to hide mistakes ,829
Deliberately work badly ,808
Knowingly work slow, while busy at work ,803 -,306
Exaggerate/ falsifying work results ,800 -,311
Deliberately work slow/ sloppy ,779 -,304
Not informing a colleague about a work related thing ,735 -,406 Unfairly blame someone else for a mistake ,686 -,546 Not helping a colleague where possible ,641 -,527 Deliberately interrupt someone who is working ,633
Call in sick, without being actually sick ,623
Deliberately ignore a leader ,621 ,330 -,426
Deliberately come back late from lunch/ break ,597 ,474 ,389 Deliberately spill organizational resources ,583
Tell people outside their job, how annoying it is ,570 ,418 Spreading a potential harmful rumor ,495 -,408 ,456
Use working time for private matters ,452 ,481 ,369
Kidding someone at work ,407 ,427 -,373
Eigenvalues 7,462 1,753 1,344 1,106
Percentage of total variance 43,894 10,311 7,908 6,505
Number of test measures 8 5 3 1
*Loadings => . 30
When interpreting the results of this factor analysis a few things stand out. First of all, since every item loads high on the first component, perhaps this component just stands for CWB. However, when looking at the items that load the highest on this component, a pattern can be seen. These items seem to reflect a component that can be associated with failing in
36
task performance. In terms of the two dimensions proposed by Sacket et al. (2006) they most readily reflect CWB towards the organization. Failing in task performance clearly is bad for the organization, without necessarily being negative for other individuals. Note that Sacket et al. (2006) divided CWB in CWB towards the organization and towards individuals. In light of this definition, component two and four can also be explained. Although component four only has a high factor loading of a single item, this item stands out among the others. The eleventh item is the only item concerning a leader, therefore this component reflects CWB towards the leader. The items loading high on the second component all have a social factor in them as well. This component seems to reflect CWB towards colleagues in particular. Component 2 and component 4 therefore can be interpreted as the second dimension from Sacket et al. (2006), CWB towards individuals. The only distinction here, is the focus towards colleagues or the leader. The third component stands out among the others, simply because it does not seem to fit in the definition by Sacket et al. (2006). When looking at the items which load high on this component, this component can be interpreted as one associated with gossip and rumor. One can argue that this component as well might be part of CWB towards individuals because rumor and gossip intuitively seem to be directed at individuals. However, as one of the relevant items states, it also entails negative expressions about ones job and therefore harms the organization. This notion makes the third component stand out among the others and advocates for a new theoretical dimension of CWB. This dimension cannot be explained by the two dimensions found in the research of Sacket et al. (2006).
Implications
Although the proposed hypotheses could not be supported by the research done in this thesis, the resulting implication is useful at least. The main finding of this thesis is that less leadership might lead to less CWB for Machiavellianists. It seems as if employees who score
37
high on Machiavellianism benefit from less attention from the leader. Benefit here is used in the sense of displaying less CWB. When it comes to organizations, the implication of this finding is that when a leader has to deal with one or more employees scoring high on
Machiavellianism, perhaps the leader should consider giving these employees less attention, and perhaps more autonomy. To know which employee is a high Machiavellianist, and which one is not, is difficult for a leader. Machiavellianism obviously cannot be seen from the looks of a person. Furthermore, measuring Machiavellianism before hiring someone might raise some ethical concerns. However, when someone is an employee of an organization, sometimes questionnaires are being collected throughout a working period. Some of these might be to measure performance, or to ask for input to nourish innovation. A small scale for Machiavellianism could be implemented here, with the goal to foster a better understanding between leader and employee, in order to achieve better results. In this way, the results found here can directly be translated into practice. As mentioned in the introduction, leadership styles can lead to the avoidance of some of the negative consequences of having an employee with high Machiavellianism, while keeping the positive consequences. Perhaps no one real leadership style is able to do this, but certain aspects or elements like giving autonomy can. Other implications of the results in this thesis can be found in the large number of possibilities it opens for future research, of which some are already mentioned throughout this discussion and others will be discussed in the next section.
Limitations and Future Research
A few suggestions for future research have already been made throughout the previous part of the discussion, the first suggestion here is of a methodological kind. Perhaps it is better to be seen as a tip, something to take into account when performing a similar research as the one done in this master thesis. Subsequently from this first point, a suggestion for future
