Are narcissistic leaders looking for similar employees?
The influence of leader narcissism on the relationship between personality
traits of the employee and their OCB as perceived by the leader.
Thesis MSc. in Business Studies
Leadership and Management track
Author: Chris van den Hout
Student number: 10514937
Supervisor: mw. prof. dr. Deanne den Hartog
Date: August 2014
2
Table of Contents
Abstract ... 3 Introduction ... 4 Literature Review ... 7 Narcissistic leadership ... 7Personality traits and perceived employee OCB ... 9
Positing hypotheses ... 11
Controlling for mediation ... 17
Methods ... 18
Participants and Design ... 18
Measures ... 20 Analysis ... 22 Additional analysis ... 24 Results ... 25 Additional results ... 28 Discussion ... 31 Personality traits ... 31 Narcissism ... 33 Perceived similarity ... 35 Implications ... 36 Limitations ... 38 Future research ... 39 Conclusion ... 40 Reference List ... 41 Appendices ... 47
List of Figures
Figure 1: Conceptual model Openness ... 14Figure 2: Conceptual model Extraversion ... 15
Figure 3: Conceptual model Neuroticism ... 16
List of Tables
Table 1: Descriptive statistics sample ... 19Table 2: Means, Standard Deviations, Correlations and Reliabilities ... 23
Table 3: Correlation matrix ... 26
3
Abstract
This thesis focuses on the potential moderating effect of leader narcissism in the relationship between on one side the employee’s personality traits of Openness, Extraversion and Neuroticism and on the other side the leader’s perception of the employee’s organizational citizenship behavior (OCB). This study examines whether narcissistic leaders perceive employees with certain personality traits as more positive when they have to judge them on the extent to which they display OCB. It is argued that narcissistic leaders generally score high on the personality traits of Openness, Extraversion and low on Neuroticism. Furthermore, based on the high arrogance and self-admiration of narcissists, it is expected that narcissistic leaders would evaluate employees with similar traits higher in their displayed OCB than employees with dissimilar personality traits.
Contrary to expectations, no significant relationships are found between the personality traits of the employee and the perceived OCB construct. Supervisor narcissism does not affect these findings in any way. Additional analyses show that perceived OCB was related to perceived similarity, in such a way that supervisors who perceived their employee to be similar to themselves rated them higher on their OCB. This relationship is also not affected by supervisor narcissism in any significant way. The implications and limitations of this study are discussed and directions for future research are suggested.
4
Introduction
Last year you were passed over for a promotion even though you felt you earned it more than your colleague who did get promoted. The last 12 months you have been going above and beyond what is expected of you, delivering excellent work throughout the year. Only to find out you are again being passed over for a promotion by your supervisor. Perhaps the problem is not with the work you have delivered or your actual behavior, but with the supervisor that you are working for and the favoritism your supervisor shows towards certain types of employees. What if your supervisor’s narcissistic traits influence the way you are being evaluated on your organizational citizenship behavior (OCB) based on your personality?
Today’s world is full of different leaders such as political, religious and business leaders. Besides the field in which these leaders are active, one can also categorize leaders by the type of leadership they practice. Recent research has focused on different types such as ethical, motivational, inspirational, charismatic, transformational and narcissistic leadership. Studies have shown that people with narcissistic traits are more likely to emerge as leaders in today’s society (Brunell, Gentry, Campbell, Hoffman, Kuhnert & De Marree, 2008; Nevicka, De Hoogh, Van Vianen, Beersma, & McIlwain, 2011). Although the results of their practiced narcissistic leadership style on operational performance are mixed (Campbell, Hoffman, Campbell, & Marchisio, 2011; Stein, 2013), their tendency to emerge as leaders cannot be overlooked.
Studies on narcissistic leaders have uncovered that they are confident, tend to make bold risky decisions and can be charming if they want to be (Rosenthal & Pittinsky, 2006). Seeing as narcissistic persons tend to emerge as leaders, it is likely that many people work for a narcissistic leader. This means that in many organizations and on different levels within
5 these organizations a narcissistic leader decides who gets bonuses or even who will be promoted and who will not. Therefore it is important to know what kind of employee is favored by narcissistic leaders. Our knowledge of narcissistic leaders has grown significantly over the last decades, yet we still know very little about what kind of personality traits they might favor when it comes to their employees.
From the perspective of the employee it is interesting to know whether the fact that you are being passed for promotion by your supervisor might have anything to do with the fact that your supervisor does not like your personality and therefore does not evaluate your work objectively. Your supervisor is usually the person that evaluates you, distributes bonuses and possibly decides who will be put forward for promotion. Therefore knowing whether your narcissistic supervisor is likely to have certain preferences is of importance.
From the perspective of recruitment agencies it is interesting to know what personality traits of an employee may help create a positive work evaluation by a narcissistic leader. This way recruitment agencies will be more likely to satisfy their client, the narcissistic leader, by selecting employees that possess certain personality traits above those who do not. It is also useful for organizations themselves to know whether narcissistic leaders favor certain employees. When an employee receives mixed evaluations from different supervisors, a reason could be that the supervisors differ in narcissism. It could also be beneficial for organizations to know whether some employees are overrated or underrated because their supervisor is a narcissist. This thesis focuses on the potential moderating effect of leader narcissism in the relationship between on one side the employee’s personality traits of Openness, Extraversion and Neuroticism and on the other side the leader’s perception of the employee’s OCB. This study examines whether narcissistic
6 leaders perceive employees with certain personality traits as more positive when they have to judge them on the extent to which they display OCB.
In the next chapter, the literature review, the most prominent articles and research in the fields of narcissism, personality traits, OCB and similarity are discussed. The research question and its related hypotheses are posited near the end of the literature review. In the following methods chapter the sample, design and measures of the study are described in detail. A clear description of the data analyses concludes the methods chapter. The results of the analyses are presented in the next chapter, which is followed by the discussion chapter. In the discussion chapter the findings of this study are discussed in relation to the literature. The limitations and implications are addressed and directions for future research are provided. This thesis ends with a conclusion chapter in which the aim and findings of this research are recited.
Literature Review
Narcissistic leadership
Narcissism in leaders and the narcissistic leadership style have been the subject of research for decades now. Narcissism can be defined as the relatively stable individual difference consisting of grandiosity, self-love and inflated self-views (Campbell et al., 2011). Narcissists are typically overconfident, dominant and focused on themselves (Brunell et al., 2008; Nevicka et al., 2011). An important contribution to the research on narcissism was the development of the Narcissistic Personality Inventory (NPI). Raskin and Hall (1979) constructed a questionnaire that measured to what extent people differ on the trait narcissism. Emmons (1984; 1987) showed the validity of the NPI in multiple studies and found four sub-factors of narcissism: leadership/authority, self-absorption/self-admiration, superiority/arrogance and exploitativeness/entitlement. Raskin and Terry (1988) confirmed the validity of the NPI and found evidence for the existence of seven first-order components: authority, exhibitionism, superiority, vanity, exploitativeness, entitlement and self-sufficiency. The sub-factors found by Emmons (1984; 1987) resemble the first-order components found by Raskin and Terry (1988).
The research on narcissism and narcissistic leadership has progressed a lot since the creation of the NPI back in the late 1980’s. A few different perspectives in the research on narcissistic leadership have gotten more attention than others. One of the phenomena’s that has grasped the attention of various researchers has been the emergence of narcissistic leaders. Brunell and colleagues (2008) found that people high on narcissism are more likely to emerge as group leaders within leaderless groups. They also discovered that the power factor of narcissism is a good predictor of leader emergence and that the attention that
8 comes with it is not a primary motivator. They reasoned that high Extraversion which is usually associated with narcissism might be a factor in helping narcissists to become leaders. Nevicka and colleagues (2011) also found that narcissistic people tend to emerge as leaders. They also found that the context in terms of team’s high versus low reward interdependence does not affect the emergence of the narcissistic leader. Ouimet (2010) researched the emergence of narcissistic leaders by reflecting back on existing theories and concluded that there are four factors that help trigger the manifestation of narcissistic leadership: idiosyncratic, cultural, environmental and structural factors. These studies on emergent leadership make clear that narcissistic persons will emerge as leaders in many situations. This makes it all the more relevant to understand what personality traits among employees are evaluated positively by these kind of leaders.
Another important aspect of the research on narcissistic leadership is the effect narcissistic leaders have on the performance of an organization. Up until Rosenthal and Pittinsky (2006) argued for a more functional definition of narcissistic leadership, researchers debated whether this kind of leadership was either good or bad. Researchers now tend to agree that narcissistic leadership can have both positive and negative effects on performance. According to Stein (2013) one individual can exhibit both constructive and reactive forms of narcissistic leadership. Constructive narcissism exists of hubris, omnipotence and omniscience and is considered to have a positive effect on organizational performance. Reactive narcissism also exhibits either or both contemptuousness or/and vengefulness and has a more negative effect on organizational performance (Stein, 2013). Relating back to leader emergence, Paunonen and colleagues (2006) researched if these two sides of narcissistic leadership affect leader emergence as well. They found that the positive or the so-called bright side of narcissistic leadership, high levels of esteem and
self-9 confidence, positively correlates with emergent leadership and that the dark side, high levels of manipulativeness and impression management did not show any correlation. Higgs (2009) concluded that narcissistic leadership can be both positive and negative. He did however emphasize that the negative effects are to be found within the organization itself. The long-term performance outcomes can suffer because of the negative effect narcissistic leaders tend to have on the internal climate (Higgs, 2009).
Nevicka and colleagues (2011) showed that the context wherein narcissistic leaders work also affects their performance. They studied the performance of narcissistic leaders in high and low team interdependence reward contexts and found that narcissists perform better when reward interdependency is high. Campbell and colleagues (2011) reviewed the literature on narcissistic leadership in organizational contexts and concluded that its positive effects are primarily seen in situations which are characterized as chaotic or novel. Its negative effects on the other hand are more obvious in stable and long-term situations. It is interesting to see that Campbell and his colleagues (2011) also note that the most prominent negative effects of narcissistic leadership are in damaging relationships on the long run. This supports the findings of Higgs (2009) that narcissistic leadership has a negative effect on the internal climate.
Personality traits and perceived employee OCB
For a long time efforts have been made to describe and measure different personalities. The most agreed upon model is the Five-Factor Model (FFM), also known as the Big Five Personality model (Digman & Inouye, 1986; Digman, 1990; Goldberg, 1990; NORMAN, 1963). The terms used for the five factors have been adapted over the years but studies show that the terms Openness to experience, Conscientiousness, Extraversion, Agreeableness and Neuroticism are adequate for the taxonomy of personality (McCrae & Costa, 1985; McCrae
10 & Costa, 1987). The personality trait Openness to experience is sometimes also termed Intellect or Imagination. For readability and clarity reasons, from here on out this thesis will use the word Openness when discussing this particular personality trait.
People high on Openness are imaginative, daring, independent and prefer variety over structure. They are willing to think outside of the box and are open-minded when it comes to handling problems. Someone high on Conscientiousness is most likely neat, punctual, self-disciplined and reliable. The adjectives sociable, talkative and outgoing characterize someone high on Extraversion. When high on Agreeableness a person is more likely to cooperate with others, be sympathetic and lenient. Traits associated with Neuroticism are anxious, angry, depressed and emotional (McCrae & Costa, 1985; McCrae & Costa, 1987; Barrick & Mount, 1991; Organ, 1994).
Many studies have been undertaken linking personality to a wide variety of factors such as job performance (Barrick & Mount, 1991), coping behavior (Allen, Greenlees, & Jones, 2011) and academic performance (Poropat, 2009). Researchers are also interested in investigating the link between personality and OCB (Abu, 2010; Chiaburu, Oh, Berry, Li, & Gardner, 2011; IJ Hetty & Euwema, 2007; Organ, 1994). OCB is a complex phenomenon that involves voluntary action to help colleagues without the promise of any rewards for that help (Organ, 1988). Studies show that OCB of the employee leads to positive evaluations of the supervisor and reward allocations (N. P. Podsakoff, Whiting, Podsakoff, & Blume, 2009; P. M. Podsakoff, MacKenzie, Paine, & Bachrach, 2000b). A meta-analysis of research on the five personality traits and OCB shows that the three personality traits of Openness, Extraversion and Neuroticism/Emotional Stability all have incremental validity for predicting citizenship (Chiaburu et al., 2011). Chiaburu and colleagues (2011) found that these three traits are especially useful for predicting change-oriented citizenship. People high on
11 Openness show greater adaptability (LePine, Colquitt, & Erez, 2000) and a better understanding of the context surrounding a task, which can have a positive impact on OCB displayed (Chiaburu et al., 2011). The more relaxed and undemanding attitude of people low on Neuroticism makes them more likely to exhibit OCB (Small & Diefendorff, 2006). The more flexible behaviors showed by someone high on Extraversion could make them more likely to display OCB (Kumar, Bakhshi, & Rani, 2009).
Positing hypotheses
The literature on narcissism and leaders with high levels of this trait dates back decades. At first research focused on identifying and measuring narcissism and its components. From there the focus of research on narcissistic leadership shifted to different perspectives. The positive and negative effects on performance have been studied and so has the emergence as leaders of people with narcissistic traits been. When taking the results of these studies in consideration it is obvious narcissistic leaders are likely to emerge in a wide variety of situations and they can have both a positive as well as a negative influence on the performance and social climate of an organization.
Studies have shown that the negative effects of narcissistic leadership are mostly found in the relationship area and for the internal climate within an organization (Campbell et al., 2011; Higgs, 2009). Seeing as narcissistic leaders are likely to emerge and hold positions of power, they are also likely to be supervising any number of employees. Taking this into account it is interesting to discover whether these negative effects on the relationships within an organization might be the result of perceived mistreatment of employees that work under the supervision of narcissistic leaders. This perceived mistreatment can be the result of being passed for promotion or getting smaller bonuses than they feel entitled to. Reasoning that these rewards are assigned based on the
12 perception of displayed OCB of the employee by the supervisor, it is interesting to know if narcissistic leaders evaluate certain personality traits within employees more positively than others.
Judge and colleagues (Judge, Bono, Ilies, & Gerhardt, 2002) conducted a meta-analysis to investigate if the FFM could be used as an organizing framework in determining dispositional predictors for leadership. They acknowledged that leadership can be divided into leader emergence and leader effectiveness. Their research shows that when leadership is taken as the combined construct of leader emergence and effectiveness, the personality traits of Extraversion, Conscientiousness, Openness and Neuroticism are good predictors of leadership. Leaders tend to score somewhat higher on Extraversion, Conscientiousness and Openness than people who are not in a leading role. Leaders are also more likely to score somewhat lower on Neuroticism than people who are not leaders (Judge et al., 2002). Extraversion was the strongest predictor of leader emergence, followed by Conscientiousness. Openness also showed a consistent correlation with leader emergence across the included studies (Judge et al., 2002).
Following that Extraversion is the strongest predictor of leader emergence, it is not surprising that narcissists tend to emerge as leaders (Brunell et al., 2008). Research shows that narcissism is strongly related to Extraversion (Egan & McCorkindale, 2007; Mathieu, 2013; Paulhus & Williams, 2002; Vernon, Villani, Vickers, & Harris, 2008). Narcissism is also found to positively correlate with Openness and negatively correlate with Agreeableness (Egan & McCorkindale, 2007; Mathieu, 2013; Paulhus & Williams, 2002; Vernon et al., 2008). Slightly less consistent are the findings on the relationship between Neuroticism and narcissism, but all significant correlations found are negative (Egan & McCorkindale, 2007; Mathieu, 2013).
13 It is reasonable to believe that employees have a wide range of personality traits. If narcissistic leaders perceive the OCB of employees with certain personality traits more favorable than others, this will have important implications for supervisors, employees and recruitment agencies.
In order to research if leader narcissism affects the relationship between personality traits of the employee and the OCB of the employee as perceived by the leader, this thesis posits three hypotheses. The three personality traits that have been found to predict OCB, Openness, Extraversion and Neuroticism (Chiaburu et al., 2011), are taken as the independent variables in each of the conceptual models. Research shows that narcissists tend to have high levels of self-admiration and arrogance (Emmons, 1984; Emmons, 1987; Raskin & Terry, 1988; Rosenthal & Pittinsky, 2006). Inherent to their self-admiration, narcissists are likely to be very happy with their own personality. Because of their arrogance it is also likely that narcissists will see their own type of personality as the most desirable one. This is expected to lead to narcissists favoring employees with similar personality traits. Following this reasoning the expectations are that narcissistic leaders will evaluate employees with similar personality traits to themselves higher in OCB. In other words, the level of narcissism of the supervisor is expected to moderate the relationship between the three personality traits and the perceived OCB.
Considering that leaders generally score somewhat higher on Openness and Extraversion and somewhat lower on Neuroticism (Judge et al., 2002) and that narcissists have high levels of self-admiration and arrogance (Emmons, 1984; Emmons, 1987; Raskin & Terry, 1988) this thesis argues that narcissistic leaders will evaluate employees high on Openness and Extraversion and low on Neuroticism more positively than leaders that are not narcissists. This results in the following hypotheses and associated conceptual models.
14 An employee that has higher levels of the Openness trait is more likely to be adaptive when it comes to new approaches and creative in coming up with new ideas for dealing with problems. That employee will also have a better understanding of abstract ideas and more imagination than someone with lower levels of the Openness trait. When a supervisor recognizes these qualities in an employee it is likely this will positively affect the perceived OCB of the employee by that supervisor. Research shows that narcissists tend to have high Openness (Mathieu, 2013; Paulhus & Williams, 2002; Vernon et al., 2008) and leaders in general also tend to score higher on Openness (Judge et al., 2002). It is therefore likely that narcissistic leaders have high levels of the Openness trait and, as a result of their self-admiration and arrogance (Emmons, 1984; Emmons, 1987; Raskin & Terry, 1988), will appreciate it all the more when an employee has this as well. Following from the above argumentation, we hypothesize that:
Hypothesis 1: There is a positive relationship between the level of Openness of the employee and the level of perceived OCB of that particular employee by the supervisor, and this relationship is moderated by the narcissism of the supervisor, such that this relationship is stronger for higher values of narcissism of the supervisor.
Fig. 1: Conceptual model Openness
Openness trait of employee Narcissism of supervisor OCB of employee as perceived by supervisor
15 An employee with high levels of the Extraversion trait will be more outgoing and more vocal than his or her colleagues with lower levels of Extraversion. This employee will therefore be more likely to attract the attention of the supervisor. When the extraverted employee displays some form of OCB it will be picked up on more easily by the supervisor because extraverted employees have the tendency to attract the attention of their supervisors (Fuller & Marler, 2009). Therefore it is expected that higher levels of Extraversion of the employee will lead to higher ratings of perceived OCB by the supervisor. Considering that leaders and narcissists tend to have high levels of Extraversion (Graziano & Tobin, 2001; Judge, LePine, & Rich, 2006; Judge et al., 2002), a narcissistic leader is likely to have high levels of the Extraversion trait as well and will recognize and appreciate someone with the same kind of Extraversion. This suggests a narcissistic supervisor will evaluate an extroverted employee even better when rating their OCB. Following from the above argumentation, we hypothesize that:
Hypothesis 2: There is a positive relationship between the level of Extraversion of the employee and the level of perceived OCB of that particular employee by the supervisor, and this relationship is moderated by the narcissism of the supervisor, such that this relationship is stronger for higher values of narcissism of the supervisor.
Fig. 2: Conceptual model Extraversion
Extraversion trait of employee OCB of employee as perceived by supervisor Narcissism of supervisor
16 Employees with high levels of Neuroticism are more likely to have mood swings and be stressed out. They also worry more and are more easily upset than their colleagues with lower levels of Neuroticism. When supervisors recognize that an employee is upset or worried often, it is likely that they will focus on these qualities and not on the OCB displayed by this employee when asked to rate the employee on their OCB. This will therefore result in lower perceived OCB for employees with high levels of Neuroticism. Seeing as narcissistic leaders are likely to have low levels of the Neuroticism trait, considering both leaders and narcissists tend to have low levels of Neuroticism (Graziano & Tobin, 2001; Judge et al., 2006; Judge et al., 2002), it is argued that they will not be very tolerant towards employees that do have high levels of it. Therefore it is expected that narcissistic leaders will rate employees that have a tendency to be neurotic lower on their OCB. Following from the above argumentation, we hypothesize that:
Hypothesis 3: There is a negative relationship between the level of Neuroticism of the employee and the level of perceived OCB of that particular employee by the supervisor, and this relationship is moderated by the narcissism of the supervisor, such that this relationship is stronger for higher values of narcissism of the supervisor.
Fig. 3: Conceptual model Neuroticism
Neuroticism trait of employee OCB of employee as perceived by supervisor Narcissism of supervisor
17 Determining whether these hypotheses are supported or not will lead to the answer of the main research question of this thesis:
“Is the relationship between employee’s personality traits and the leader’s perception of the employee’s OCB moderated by the level of narcissism of the leader?”
Controlling for mediation
Supervisors might have a hard time determining which personality traits an employee possesses and to what extent, if they are even aware of the existence of these constructs. As a result of this supervisors might not be able to directly use these constructs in evaluating the OCB of employees. It is therefore possible that perceived similarity mediates the relationship between the personality traits of the employee and the perceived OCB.
Research shows that perceived similarity between supervisors and employees affects the performance ratings supervisors give their employees. Perceived similarity and performance ratings are strongly related in such a way that perceived similarity is likely to lead to better performance ratings (Turban & Jones, 1988; Wayne, Liden, Graf, & Ferris, 1997). Another study found that relational similarity, the extent to which supervisor and employee actually share personality traits, does not significantly affect performance ratings (Strauss, Barrick, & Connerley, 2001). Straus and colleagues (2001) did find significant support for the influence of perceived similarity on performance ratings. For this thesis it implies that narcissistic leaders might not evaluate employees directly based on their actual personality traits. The extent to which the leaders perceive themselves to be similar to their employees could interfere and influence the evaluations. Following this line of reasoning it could therefore be deemed needed to check for the possible mediating role of perceived similarity.
18
Methods
Participants and Design
A total of 161 employee-supervisor dyads participated in this study. The employee participants consisted of 61 men and 100 women, with an average age of 30.81 years (SD=10.19). 124 of the employees had either finished a Bachelors or a Masters degree. Almost two thirds (65.2%) of employees had a full-time contract, the others were working part-time. Among the supervisors there were 103 men and 58 women, with an average age of 40.78 years (SD=10.39). The participants worked in a wide variety of industries, including retail, consultancy, education, government, finance and so on. The only criteria for inviting the participants was that their job had to be their main activity in life, therefore students with extra jobs for example were not able to participate. Because the focus of this study was on the relationship between certain characteristics of the employee and the supervisor no further restrictions, like industry in which the work takes place or tenure, were placed on the sample. The sample, in particular at least one person of the dyads, mainly consisted of acquaintances of the six students that gathered the data. In total 232 dyads were approached to fill in the questionnaires and 161 dyads completed both questionnaires, which results in a response rate of 69%. In table 1 on the next page is an overview of the descriptive statistics of the sample.
19 Table 1: Descriptive statistics sample
N=161 Frequency Percentage Frequency Percentage
Employee Supervisor Age < 25 years 50 31,1% 4 2,5% 25-35 years 76 47,2% 55 34,1% > 35 years 35 21,7% 102 63,4% Gender Male 61 37,9% 103 64,0% Female 100 62,1% 58 36,0% Education High school 9 5,6% 2 1,2% College 27 16,8% 13 8,1% Bachelor 60 37,3% 68 42,2% Master 64 39,8% 72 44,7% Other 1 0,6% 6 3,7% Nationality Dutch 135 83,9% 133 82,6% Non-Dutch 26 16,1% 28 17,4% Tenure < 2 years 57 35,4% 16 9,9% 2-5 years 62 38,5% 46 28,6% > 5 years 42 26,1% 99 61,5% Contract type Part-time 56 34,8% Full-time 105 65,2% Years leading < 1 year 28 17,40% 1-3 years 93 57,80% > 3 years 40 24,80%
The participants were asked to fill in a standardized questionnaire either through the online survey program, of which they would receive an email with a link directing them to it, or by pencil and paper. The participants only had to fill out the questionnaire once, therefore this study had a cross sectional design. Every employee-supervisor dyad was given an unique code to ensure that every employee questionnaire could be coupled to their supervisor questionnaire. The employee and the supervisor had different questionnaires to fill out and they were asked not to discuss the questions nor the answers with each other. There was no time limit for filling in the questionnaires and participants were allowed to stop at anytime if they wanted to finish at a later moment. The participants were first presented with a cover page that included instructions on how to complete the questionnaire, a reassurance of their
20 anonymity and a word of gratitude in advance. After which the participants were asked to rate items and statements on a 7-point format. At the end of the survey the participants were thanked for their time and effort in participating in this research.
Measures
The questionnaires measured a multitude of variables for both employee and supervisor. The relevant variables measured in the employee questionnaire for testing the hypotheses of this thesis are the personality traits of the employee, which were measured by using questions and statements that relate back to the big five personality traits of Openness, Conscientiousness, Extraversion, Agreeableness and Neuroticism. The relevant variables in the supervisor questionnaire used for testing the hypotheses are Leader Narcissism and Leader Perception Of Employee OCB. The variable Perceived Similarity, which entails to what extent supervisors believe the employee to be similar to themselves, was also measured for the purpose of doing additional analyses if deemed needed.
Leader narcissism. 40 items were used to determine the level of narcissism of the
supervisor (Raskin & Terry, 1988). Examples of sample items are ‘’I think I am a special person’’ and ‘’I always know what I am doing’’. The supervisors were asked to rate their level of agreement with the statements on a 7-point Likert-scale (1=strongly disagree, 7=strongly agree). Using a 7-point Likert-scale, as opposed to the more common used true or false ratings, provided overall consistency throughout the survey and offered the supervisors the ability to more precisely communicate their level of agreement with the statements.
Leader perception of employee OCB. Twelve items were used to determine the level
of OCB of the employee as perceived by the supervisor (Mackenzie, Podsakoff & Fetter, 1991; Van Dyne & Le Pine, 1998). An example of a sample item is ‘’The employee is always willing to help or to lend a helping hand to those around him/her’’. The supervisors were
21 asked to rate their level of agreement with the statements on a 7-point Likert-scale (1=strongly disagree, 7=strongly agree).
Leader’s perceived similarity. Three items were used to determine the leader’s
perception of similarity with the employee (Turban & Jones, 1988; Wayne et al., 1997). An example of a sample item is ‘’This employee and I are alike in a number of areas’’. The supervisors were asked to rate their level of agreement with the statements on a 7-point Likert-scale (1=strongly disagree, 7=strongly agree). This measure was put in the survey in order to be able to control for it and do additional analysis.
Big five personality. 25 items were used to determine the level of the big five
personality traits of the employees (Donnellan, Oswald, Baird & Lucas, 2006). Examples of sample items are ‘’I get chores done right away’’ and ‘’I am really interested in others’’. The employees were asked to rate their level of agreement with the statements on a 7-point Likert-scale (1=strongly disagree, 7=strongly agree).
Control variables. Some demographics were asked at the end of the survey (age,
22 Analysis
An initial screening of the data showed no abnormalities resulting from possible errors in data entry. A frequency test in the Statistical Program for the Social Sciences (SPSS) of the indicators used in this thesis showed that out of the total of 77 indicators, 12 indicators were missing one value and that one indicator was missing two values. In the employee questionnaire only two relevant indicators (bigf6 and bigf14) of the big five personality test had missing values. By establishing a significant correlation between the two indicators and two indicators of a different construct (bfass8 and bfass10), a deck for the use of the Hot Deck imputation (Myers, 2011) could be determined. After applying Hot Deck no missing values were found in relevant indicators of the employee questionnaire.
In the questionnaire for the supervisor no missing values were found for the perceived OCB construct, but 11 out of the 40 indicators of the narcissism construct were missing one value. No consistent significant correlation with indicators of different constructs could be found, therefore a correlation with complete indicators (narc12 and narc13) of the same construct was used to establish the deck variables. After applying Hot Deck no missing values were found in relevant indicators of the supervisor questionnaire. Considering the fact that the data consists of 161 dyads, the amount of missing values that was originally found is very low. Although the Hot Deck instruction does not encourage to use a deck consisting of theoretical important variables, it can be regarded as acceptable in the case of the narcissism values because of the small amount of missing values compared to the amount of total dyads.
The constructs measured in the questionnaire used in this thesis did not include any counter-indicative indicators, therefore no recoding of such indicators had to be done.
23 The reliability of the indicators of each construct was computed to check if the Cronbach’s alpha was greater than 0.7. Computing the reliability of the 12 indicators in the supervisor questionnaire measuring the perceived OCB gave an alpha of 0.89. The alpha of the 40 items measuring leader narcissism was 0.93. In the employee questionnaire three different relevant constructs are measured. Separate reliability tests showed that the alpha for Openness was 0.62, for Extraversion 0.75 and for Neuroticism 0.82. Although 0.62 is smaller than the 0.7 that was hoped for it is still considered acceptable.
Using SPSS the means for all five relevant scales were computed. For the supervisor data this resulted in two scale means, one for the 12 items of perceived OCB and one for the 40 items of narcissism. Three scale means were computed for the relevant employee data. Five items for each of the constructs of Openness, Extraversion and Neuroticism were used to computed the means. All the outcomes of the above calculations and the correlation between the relevant constructs are presented in table 2.
Table 2: Means, Standard Deviations, Correlations and Reliabilities
Mean SD 1 2 3 4 5 1 Openness 4,75 0,83 (,62) 2 Extraversion 4,69 0,95 ,24 (,75) 3 Neuroticism 3,35 1,12 ,14 -,10 (,82) 4 Narcissism 4,26 0,73 ,15 ,05 ,12 (,93) 5 Perceived OCB 5,43 0,73 -,11 -,03 -,10 ,02 (,89)
Note. Correlations greater than r = |.13| are significant for p<.05 (one-tailed)
No significant correlations are found between perceived OCB and the other constructs. This does not necessarily mean however that the moderating effect of the narcissism construct cannot still be significant on the relationship between the personality constructs and the perceived OCB construct.
24 Additional analysis
In order to check for mediation an additional analysis was performed using the construct of perceived similarity. Perceived similarity was measured using three indicators of which two indicators had one missing value and one indicator had two missing values. A deck for the Hot Deck imputation was determined through correlations with two indicators of the Perceived employees affect to the leader construct. After applying Hot Deck no missing values were found for the indicators of the perceived similarity construct. A reliability test of the three indicators gave an alpha of 0.71 and the mean of the three indicators was computed to provide a score for the construct of perceived similarity.
25
Results
The correlations between all the measured constructs and the included control variables are calculated. The most interesting correlations are mentioned here and will be addressed in the discussion, the full correlations table is presented on the next page in Table 3. Openness and Extraversion are significantly correlated at the 0.24 level. This means that a person high on Openness will also be likely to be high on Extraversion. Non-Dutch employees are higher on Neuroticism, whereas Non-Dutch supervisors are higher on narcissism. According to the negative correlations found between the employee age and the three personality traits, older employees are likely to be lower on Openness, Extraversion and Neuroticism. It is interesting to see that when the employee has a part-time contract, they are more likely to be perceived by their supervisor to display OCB, with a correlation of 0.20, and to be perceived similar to the supervisor, with a correlation of 0.22 (both significant at the 0.01 level 2-tailed). Perceived similarity and perceived OCB are significantly correlated at 0.39.
26
Table 3: Correlation matrix
E Neu Nar pOCB EA EG EN EC ET SA SG SN ST YL FoQ LoS ToS pSim Openness ,24** ,14 ,15 -,11 -,18 -,04 ,20* -,10 -,14 -,01 ,07 ,16* -,10 -,03 ,03 ,19* ,15 -,01 Extraversion (E ) -,10 ,05 -,03 -,19*
,06 -,10 -,01 -,07 -,07 ,10 -,07 -,10 ,03 -,13 -,11 ,06 ,07 Neuroticism (Neu) ,12 -,10 -,16* ,00 ,29** ,03 -,09 -,05 -,02 ,23** -,16* -,11 ,13 ,28** ,00 -,03 Narcissism (Nar) ,02 -,09 ,10 ,33** -,03 -,13 -,20* ,04 ,21** -,21** -,04 ,05 ,29** ,11 -,02 Perceived OCB (pOCB) ,14 -,07 -,11 -,20** ,12 ,19* ,03 -,16* ,10 ,19* ,01 -,10 -,18* ,39** Employee age (EA) ,18*
-,22** -,20* ,61** ,34** ,07 -,20* ,28** ,29** -,04 -,23** -,43** ,00 Employee gender 1=m/2=f
(EG) -,14 -,28** ,02 ,16* ,32** -,18* ,08 ,00 ,06 -,16* -,01 -,06 Employee nationality
1=Dutch/2=Non-dutch (EN) ,21** -,22** -,19* -,08 ,87** -,18* -,17* ,09 ,98** ,17* -,03 Employee contract
1=part-time/2=full-time (EC) -,06 -,05 -,16* ,27** ,08 -,12 -,28** ,22** -,09 -,22** Employee tenure (ET) ,27** ,00 -,19* ,33** ,37** -,06 -,22** -,37** ,11 Supervisor age (SA) -,07 -,13 ,57** ,36** -,01 -,16* -,15 -,12 Supervisor gender 1=m/2=f
(SG) -,07 -,08 ,06 ,15 -,09 -,16* ,12 Supervisor nationality
1=Dutch/2=non-Dutch (SN) -,15 -,15 ,04 ,89** ,18* -,04 Supervisor tenure (ST) ,41** -,01 -,17* -,26** -,05 Years leading (YL) ,05 -,17*
-,17* ,09 Frequency of contact 1=daily/2=weekly/3=monthly (FoQ) ,08 ,09 ,16 * Language of survey 0=Dutch/1=English (LoS) ,17* -,03 Type of survey 0=paper/1=online (ToS) -,02 Perceived similarity (pSim)
**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).
27 All three hypotheses were tested using the Process imputation as developed by Hayes (2013). All three hypotheses are concerned with the moderating effect of narcissism on the relationship between an independent and a dependent variable, therefore model 1 of the Process imputation has been used to run the tests. When using model 1 in the Process imputation (Hayes, 2013) the interaction term of the independent variable and the moderator is calculated for use in the regression analysis without the need to create a dummy variable.
To test the first hypothesis the Openness trait was used as the independent variable, perceived OCB as the dependent variable and narcissism as the moderating variable. The proportion of variance of the dependent variable that was explained by this solution was not even 2% (R-square=0.0174) with a probability of 0.4 (p=0.4292). The conditional effect of the narcissism trait was not significant for p=0.384 whereas it should be smaller than 0.05 in order for it to be significant (see Appendix A). The hypothesis that the Openness trait of the employee positively affects the perceived OCB of that employee by the supervisor, and that this relationship is moderated by the narcissism trait (H1) is therefore not supported by the data.
The independent variable was replaced by the Extraversion trait of the employee, with all other variables remaining the same, to test the second hypothesis. For this solution the variance explained was almost zero (R-square=0.0017) with a probability of 0.96 (p=0.9672). The moderating effect of the narcissism trait was again not significant for
p=0.9007 (see Appendix B). So no support was found for the second hypothesis (H2) that
narcissism moderates the relationship between the Extraversion trait and the perceived OCB by the supervisor.
28 The third hypothesis was tested by replacing the independent variable with the Neuroticism trait of the employee while again keeping the other two variables the same. Again almost no variance of the perceived OCB was explained by this model
(R-square=0.0177 and p=0.4217). The moderating effect was again not significant with a p of
0.3442 (see Appendix C). As with the previous two hypotheses, the third hypothesis (H3) stating that narcissism of the supervisor would moderate the relationship between the Neuroticism of the employee and the perceived OCB was also not supported by the data.
No significant correlations between perceived OCB and the three personality traits of Openness, Extraversion and Neuroticism were found.The first part of the three hypotheses stating that the personality traits of the employee affect the perceived OCB by the supervisor is therefore not supported for any of the three traits. The level of narcissism of the supervisor did not have a moderating effect on the relationships, so therefore the second part of all three hypotheses is also not supported. In sum, no support for any of the three hypotheses was found in this research. The main research question of this thesis “Is the relationship between employee’s personal traits and the leader’s perception of the employee’s OCB moderated by the level of narcissism of the leader?” can therefore be given a negative answer.
Additional results
Additional analyses were conducted in order to check if perceived similarity mediates the relationship between the personality traits and perceived OCB. In the Process imputation this is done by using model 4 with the independent variables being the constructs of Openness, Extraversion and Neuroticism, the dependent variable perceived OCB and the mediating variable perceived similarity. This test was run three times, once for each of the
29 independent variables, but none of the three tests showed any significant mediation by the perceived similarity construct (see Appendix D1-D3).
Perceived similarity and perceived OCB did correlate with each other at 0.388, significant at the 0.01 level (2-tailed). A hierarchical regression analysis with the dependent variable being the perceived OCB construct was done to examine the influence of the included control variables and personality constructs, see table 4 on the next page. The first model consists of only the included control variables. This model was significant (sig.
F-change=.014) and explained 8,6% of the variance (adjusted R2=0.086) of de dependent variable. In model 1 only the variables employee gender and employee contract type were significant predictors. Including the variables of Openness (model 2), Extraversion (model 3), Neuroticism (model 4) and Narcissism (model 5) only explained a minimal additional amount of variance in the dependent variable. None of the models 2-5 were significant, model 2 that included Openness came closest with a significant F-change of 0.99. Model 6, which added perceived similarity to the analysis, was significant (sig. F-change=.000) and explained over 23% of the variance of the dependent variable (adjusted R2=.0234). Perceived similarity explained the largest portion of this effect. Supervisor age (0.003) is also a significant predictor in this model and so is supervisor nationality (.049).
Another test was run to examine if the relationship between perceived similarity and perceived OCB was moderated by the narcissism trait. Perceived similarity was the independent variable, narcissism the moderating variable and perceived OCB the dependent variable. This solution explained 17% of the variance of the dependent variable (R2=0.1659), with a probability of p<0.001. The positive moderating effect of the narcissism trait was not significant however with a p of 0.0968 and only explaining 0.01 of the change (R2 -change=0.148, see Appendix E).
30
Table 4: Hierarchical Regression Analysis
Variable Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
B-value Sig. B-value Sig. B-value Sig. B-value Sig. B-value Sig. B-value Sig. Employee age -,012 ,911 -,036 ,735 -,033 ,765 -,052 ,642 -,059 ,597 -,017 ,868 Employee gender *1=m/2=f -,181 ,038 -,193 ,027 -,195 ,027 -,187 ,034 -,203 ,022 -,159 ,052 Employee nationality
*1=Dutch/2=Non-dutch -,225 ,540 -,184 ,616 -,184 ,617 -,160 ,663 -,259 ,489 -,292 ,397 Employee contract
*1=part-time/2=full-time -,227 ,012 -,252 ,006 -,252 ,006 -,252 ,006 -,251 ,006 -,157 ,066 Employee tenure -,019 ,852 -,019 ,851 -,020 ,846 -,012 ,904 -,006 ,953 -,072 ,451 Supervisor age ,185 ,062 ,204 ,039 ,205 ,040 ,214 ,032 ,228 ,023 ,283 ,003 Supervisor gender *1=m/2=f ,046 ,584 ,063 ,455 ,063 ,463 ,058 ,493 ,060 ,477 ,024 ,758 Supervisor nationality *1=Dutch/2=non-Dutch -,313 ,068 -,318 ,063 -,319 ,063 -,322 ,061 -,301 ,079 -,312 ,049 Supervisor tenure -,048 ,633 -,055 ,582 -,054 ,594 -,070 ,493 -,058 ,574 -,061 ,520 Years leading ,094 ,293 ,095 ,285 ,093 ,298 ,091 ,308 ,079 ,382 ,057 ,492 Frequency of contact *1=daily/2=weekly/3=monthly -,043 ,601 -,053 ,518 -,051 ,539 -,042 ,612 -,041 ,622 -,080 ,296 Type of survey *0=paper/1=online
-,164 ,070 -,156 ,083 -,156 ,084 -,170 ,063 -,173 ,058 -,163 ,053 Language of survey *0=Dutch/1=English ,486 ,224 ,476 ,231 ,480 ,229 ,479 ,230 ,528 ,187 ,550 ,136 Openness (O) -,133 ,099 -,137 ,100 -,128 ,125 -,137 ,101 -,115 ,136 Extraversion (E) ,015 ,853 ,005 ,955 ,001 ,991 -,031 ,682 Neuroticism (NEU) -,085 ,302 -,087 ,293 -,074 ,330 Narcissism (N) ,110 ,195 ,125 ,109 Perceived similarity (P) ,386 ,000 R-square ,160 ,176 ,176 ,182 ,192 ,321 Adjusted R-square ,086 ,097 ,091 ,091 ,096 ,234 R-square change ,160 ,016 ,000 ,006 ,010 ,129 Sig. F Change ,014 ,099 ,853 ,302 ,195 ,000
Note: a) Dependent variable = Perceived OCB.
31
Discussion
Personality traits
Many people have a supervisor that they have to report to and are evaluated by at some point in their working life. How supervisors perceive the work attitude of their employees and their displayed OCB is therefore an interesting research area. High levels of OCB of an employee will lead to more positive evaluations of their supervisor and reward allocations (N. P. Podsakoff et al., 2009; P. M. Podsakoff, MacKenzie, Paine, & Bachrach, 2000a). Research shows that the three personality traits of high Openness, high Extraversion and low Neuroticism all have incremental validity for predicting citizenship (Chiaburu et al., 2011). An employee with the combination of these three traits is more likely to display OCB, but this does not necessarily mean that the supervisor will also be aware of the fact that an employee is displaying OCB. Therefore the perceived OCB of the employee by the supervisor is not necessarily equal to the actual displayed OCB of the employee. Following the argumentation that high Openness, high Extraversion and low Neuroticism would lead to higher levels of OCB of the employee, this research examined the relationship between these three traits and the level of perceived OCB of the employee by the supervisor and the possible influence of the narcissism trait of the supervisor on these relationships.
Contrary to what was expected and the literature (Chiaburu et al., 2011), no significant relationships were found between employee Openness, Extraversion and Neuroticism on the one hand and the perceived OCB by supervisors on the other hand. Therefore no support was found for the first part of the hypotheses examined in this study that stated that the personality traits of Openness, Extraversion and Neuroticism would affect the perceived OCB of the employee by the supervisor. Openness and Extraversion did show a positive correlation with each other, meaning that in this study an employee with
32 high Openness was also likely to have high Extraversion, and vice versa. This relation is not uncommon (Mathieu, 2013), but is not found in every study (Organ, 1994).
The remarkably low correlation between perceived OCB and the personality constructs could have any number of reasons. Organ (1994) found that self-report ratings of personality traits by employees did not predict supervisor’s performance ratings as well as co-workers ratings of personality did. This study used self-report ratings to measure the personality traits of the employees, which could explain the lack of correlation between the personality traits and the perceived OCB (Organ, 1994).
Another possible reasons could be that some variables might be in place that influence the relationship between actual OCB and perceived OCB by the supervisor. The amount of time the supervisor monitors or has contact with the employee can have an important effect on this relationship. It can be argued that the more time the supervisor spends with the employee, the more likely the displayed OCB is also perceived by the supervisor. Although the supervisors were asked how many times a week they had contact with their employee, which did not significantly influence the perceived OCB, they were not asked to specify the type and duration of the contact.
Another explanation for the lack of correlation could be that the personality traits were not measured comprehensive enough in this study. Each trait was measured by asking the participating employees to rate five statements according to what extent they performed certain behavior. The combination of these five ratings gave an overall rating of their level of that particular personality trait. Asking the participants to rate more statements could enhance the quality of the measured constructs, which in turn could influence the relationships with the perceived OCB construct.
33 It is interesting that controlling for the influence of the included control variables, both gender of the employee and whether they worked part-time or full-time made a significant difference. Male employees were rated higher by their supervisors in displayed OCB than their female colleagues. This could mean that men display more OCB, but it could also just mean that they are better at making their supervisor aware of their OCB. This could be the result of certain behavior like showing off when their supervisor is present.
Employees that worked part-time also obtained higher perceived OCB ratings than full-time employees did. Again, this could be just because part-time employees simply display more OCB. Another possible explanation could be that supervisors do not expect the same from part-timers and full-timers. If the expectations for full-timers are higher than for part-timers, it would be easier for the later to be perceived as displaying OCB, whereas the same actions of the employee working full-time would be regarded as normal work behavior.
Narcissism
Narcissists tend to have higher levels of Openness and Extraversion, and lower levels of Neuroticism than people low on Narcissism (Fuller & Marler, 2009; Graziano & Tobin, 2001; Mathieu, 2013; Paulhus & Williams, 2002; Vernon et al., 2008). Leaders also tend to have these qualities (Judge et al., 2006; Judge et al., 2002), therefore it is argued that narcissistic leaders would be likely to have relatively high Openness and Extraversion and low Neuroticism. Reasoning that the arrogance and self-admiration of narcissistic leaders (Emmons, 1984; Emmons, 1987; Raskin & Terry, 1988) would lead them to evaluate employees with the same personality traits as themselves higher in OCB, it was expected that the level of narcissism of the supervisor would moderate the relationships between the three personality traits and the perceived OCB. Contrary to these expectations the level of
34 narcissism of the supervisor did not moderate the relationships of any of the three personality traits and perceived OCB in this study. Therefore the second parts of all three hypotheses were also not supported by the data in this research.
A possibility for this lack of moderating effect could be that the self-admiration and arrogance of narcissistic leaders does not directly translate to more appreciation and better evaluation of employees that have similar personality traits. Another possibility could be that the supervisors are not aware of the personality traits of their employees and therefore the narcissistic leaders can not favor those employees that share the same traits as they do. Yet another possible explanation could also be found in the design of the survey, perhaps the sequence of the questions influenced the manner in which participants rated certain statements or the way the data was collected could have had an influence. A final possible reason for not finding any significant relations could be that this study is too ambitious in analyzing the dataset. The distance between self-rated personality traits of the employee, perceived OCB of that employee by the supervisor and supervisor narcissism could have been too much to result in any significant effects.
Although the level of narcissism of the supervisor does not moderate the relationships between the personality traits of the employee and the perceived OCB, it does show significant correlation with other variables. Supervisors that were non-Dutch, which included mostly people of Asian descent in this study, are found to be more narcissistic than their Dutch counterparts. Employees that were non-Dutch on the other hand, showed to have higher levels of Neuroticism than Dutch employees. Considering that studies found that narcissism negatively correlates with Neuroticism (Egan & McCorkindale, 2007; Mathieu, 2013), it is interesting to see that there might be a substantial difference in personality traits and narcissism in Asian supervisors and Asian employees.
35 In this study narcissism negatively correlates with the tenure of the supervisor. This means that the longer supervisors are working for the same employer, the lower levels of narcissism they possess. The age of the supervisor also shows a negative but weaker correlation with narcissism, which is logical since supervisor age and tenure are strongly correlated. These findings suggest that older supervisors who have been working for the same company longer than the younger supervisors are less narcissistic. Perhaps the role of supervisor makes younger people more confident and narcissistic, because their age group might not consist of many supervisors yet. Another possible explanation could be that when supervisors are fairly new in a company, they are confident and happy with themselves because they acquired the position and this could result in higher narcissism scores.
Perceived similarity
As mentioned in the previous chapter, employees with part-time contracts are found to be rated higher in OCB by their supervisors than employees that worked full-time. Part-time employees also score higher in perceived similarity according to ratings from the supervisors, compared to full-time employees. A regression analysis revealed that the relationship between part-timers and higher ratings of perceived OCB, is largely explained by the perceived similarity construct. Another interesting result is that perceived similarity and perceived OCB were found to significantly correlate at .39 (at the 0.01 level 2-tailed), which confirms what previous research found (Glomb & Welsh, 2005; Waismel‐Manor, Tziner, Berger, & Dikstein, 2010). Supervisors are thus more likely to rate employees higher on their OCB if they perceive these employees to be similar to themselves.
Although it can be expected this relation would be stronger when the supervisor is a narcissist, because it could then be explained as an effect of the arrogance and self-admiration of the supervisor (Emmons, 1984; Emmons, 1987; Raskin & Terry, 1988), this
36 study found no significant moderating effect by the narcissism construct on this relationship. Perceived similarity does not have a mediating role between any of the three personality traits and perceived OCB. The age of the supervisor also explains some variance in the perceived OCB score, although less than perceived similarity. The results show that older supervisors are more likely to rate their employees high on perceived OCB. Possible explanations for this could be that older supervisors are less critical towards their employees or that they are better at observing OCB displayed by their employees.
Implications
This research has several noteworthy theoretical implications. This study adds to the existing literature on personality traits and their effects on OCB. Previous research has shown that the personality traits of Openness, Extraversion and Neuroticism have incremental validity for predicting OCB (Chiaburu et al., 2011). Contrary to the expectations and the literature this study did not find any significant relationships between these three traits and the perceived OCB construct. This could be caused by the use of self-rated personality traits of the employee and supervisor rated OCB of the employee, which can diminish the results of the effect of personality on OCB (Chiaburu et al., 2011; Organ, Podsakoff, & MacKenzie, 2006).
This study also contributes to the existing literature on narcissism by examining its influence on the relationships between the personality traits and OCB. Following that narcissists and leaders tend to be high in Openness and Extraversion and low in Neuroticism (Graziano & Tobin, 2001; Judge et al., 2006; Judge et al., 2002; Mathieu, 2013; Paulhus & Williams, 2002; Vernon et al., 2008) and narcissists are likely to be arrogant and admire themselves (Emmons, 1984; Emmons, 1987; Raskin & Terry, 1988), it was expected that narcissistic supervisors would rate employees who scored high on those same personality
37 traits, higher in perceived OCB. Against expectations narcissistic supervisors did not rate employees with similar or dissimilar personality traits significantly different from each other.
Another theoretical implication of this study is its extension of the literature on similarity and OCB. In line with previous research (Glomb & Welsh, 2005; Waismel‐Manor et al., 2010) this study found that supervisors rate employees higher on OCB if they feel that they are similar to them. Additionally this study adds another contribution to this field by having examined the role of supervisor narcissism in this relationship. This study found that leader narcissism does not influence the relationship between perceived similarity and perceived OCB.
Besides theoretical implications this study also has some practical implications that are worth mentioning. Previous research shows that employees who are considered to display many OCB’s are more positively evaluated and are allocated more rewards than those who are lower in OCB (N. P. Podsakoff et al., 2009; P. M. Podsakoff et al., 2000b). This study has found that employees are perceived to be higher on OCB when the supervisors perceive the employee as similar to themselves, which is in line with previous research (Glomb & Welsh, 2005; Waismel‐Manor et al., 2010), regardless of the personality traits possessed by the employee or the level of narcissism of the supervisor. This implies that employees benefit when working for a supervisor that is similar to them, or more specifically when the supervisor believes them to be similar. From the employees’ perspective it would therefore be wise to search for a job or function within a company where their supervisor is much like them. From the supervisors’ perspective it is important to be aware of this phenomenon, because it can affect their judgment. Companies as a whole should also be aware that some evaluations and reward allocations are influenced by the relationship between perceived similarity and OCB.
38 Another practical implication of the findings of this research is closely related to the previous one. The results of this study suggest that narcissistic supervisors do not favor a specific type of employee when rating them on their displayed OCB. This means that when employees feel they are mistreated by their supervisor it is unlikely the cause lies with the level of narcissism of that supervisor. Taking this into consideration combined with the previous point, it would be wise for employees to search for other reasons, such as dissimilarity with the supervisor, and not get hung up on the fact that their supervisor is a narcissist.
Limitations
This research is subject to a number of limitations. The first limitation is its cross-sectional design which does not allow for any conclusions to be drawn concerning causality. The sample selection is another limitation of this study, since in general one of every dyad was an acquaintance of one of the students that were involved in collecting the data. Knowing that someone familiar would be able to read the answers might have influenced the ratings given by the participants, even though strict confidentiality was of course emphasized. It is also possible that the participants felt the urge to give social desirable answers, which would also influence the outcomes of this study. The data was collected through an online survey and through a pencil and paper version. It is possible that one of the dyads had to hand over their paper survey to the other person, who then in turn would give them both to a member of the research team. This could also have influenced the answers given and thus the outcomes of this research.
A final possible limitation that has to be mentioned is the Cronbach’s Alpha of the Openness trait which was 0.62. Although not far off the 0.7 that is the standard, it can still
39 have caused some measurement errors and might have affected the overall results of the study that concerned the Openness trait.
Future research
Some interesting subjects for further research are derived from this study. Future research could focus more on the quality and nature of the time spend between supervisor and employee and how this affects the relationship between displayed OCB as reported by the employee and perceived OCB by the supervisor.
Another interesting subject that future research could study is the influence of supervisor rated personality traits of the employee, instead of employee rated, on perceived OCB. By having the supervisor rate the personality traits of the employee it can be researched whether narcissists favor those employees that they perceive to have similar personality traits to themselves (Organ, 1994).
The non-Dutch group in this study consisted primarily of people of Asian descent. The employees of this group score significantly higher on Neuroticism than their Dutch counterparts. It is therefore all the more interesting that the non-Dutch supervisors are significantly more narcissistic than the Dutch supervisors, considering that narcissism and Neuroticism are found to negatively correlate (Egan & McCorkindale, 2007; Mathieu, 2013). For future research it is therefore interesting to study the differences in personality traits and the level of narcissism between Asian employees and their supervisors.
A final suggestion for future research is to use more extensive survey items to measure the personality traits of the employees in order to get a more accurate and detailed result. This could positively influence the results of any future research on personality and its relationship with OCB.
