• No results found

Not everyone can be narcissistic, so how is it possible that we say that we know so many? : a quantitative study of the multidimensionality of the NPI and the impact of extraversion, beyond the Big Five, on the relation

N/A
N/A
Protected

Academic year: 2021

Share "Not everyone can be narcissistic, so how is it possible that we say that we know so many? : a quantitative study of the multidimensionality of the NPI and the impact of extraversion, beyond the Big Five, on the relation"

Copied!
48
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

MASTER THESIS

Not everyone can be narcissistic,

so how is it possible that we say that

we know so many?

A quantitative study of the multidimensionality of the NPI and the impact of Extraversion, beyond the Big Five, on the relationship between narcissism and OCB

University of Amsterdam

Faculty of Economics and Business

Track: MSc Business Studies - Leadership and Management Student name: Ronald J. Schep

Student number: 10564314

Student e-mail: ronaldschep@outlook.com First supervisor: Prof. D. Den Hartog Second supervisor: Dr. A. de Hoogh

(2)

ii Table of Contents List of tables ... iv List of figures ... v Acknowledgements ... vi Abstract ... vii Introduction ... 1 Literature review ... 4 Narcissism ... 4

Narcissism and its dimensional structures. ... 5

Adaptive and Maladaptive narcissism. ... 6

Narcissism as a one-dimensional construct. ... 7

Narcissism in organisational contexts. ... 7

Organisational citizenship behaviour ... 8

Leadership behaviour as antecedent. ... 9

Narcissism and OCB. ... 9

Self- and other-rated OCB. ... 10

Narcissism’s facet structure and OCB. ... 11

Flavours of OCB. ... 11

Dominance complementarity theory ... 13

Leadership trait theory. ... 13

Interpersonal complementarity theory. ... 14

Narcissistic dominance. ... 15

Methods... 17

Sample and procedure ... 17

Measures... 18

Narcissism ... 18

Organisational Citizenship Behaviour. ... 18

Big Five. ... 19

Data preparation ... 19

Confirmatory factor analysis ... 20

Confirmatory Factor Analysis of the Three-Factor Solution. ... 20

Confirmatory Factor Analysis of the Two-Factor Solution. ... 20

(3)

iii

Results ... 22

Discussion ... 24

Implications for theory and practice ... 27

Limitations ... 28

Directions for future research ... 29

Conclusion ... 30

References ... 32

Appendices ... 40

(4)

iv List of tables

Table 1. CFA Fit Indexes of NPI Solutions……….20 Table 2. Means, standard deviations, and intercorrelations of the measures………...22 Table 3. Summary of Hierarchical Regression Analysis Leadership/Authority………..39 Table 4. Summary of Hierarchical Regression Analysis Exhibitionism/Entitlement………..40

(5)

v List of figures

Figure 1. Conceptual model ………16 Figure 2. Results of structural equation model ………...23

(6)

vi Acknowledgements

I would like to especially thank prof. dr. Deanne den Hartog for supervising this master thesis project. I have worked with great pleasure on this study thanks to her positive and enthusiastic attitude. See was always of great support when I got stuck or just needed professional advice. See gave me insights which made this thesis much better.

I would also like to offer special thanks to Dr. Annebel de Hoogh for being the second reader and expert in our thesis group sessions.

Finally, I would like to give special thanks to all the participants who made this study possible, to my fellow students with whom I collected the data and to those people who have read this master thesis and provided me with valuable feedback.

(7)

vii Abstract

Purpose: In the present study the effects of narcissism in an organisational context

are investigated. Past empirical studies have reported conflicting results: both negative and positive effects of narcissism in an organisational context have been reported. The present study investigated the effect of supervisor narcissism on pro-organisational behaviours, known as Organisational Citizenship Behaviour (OCB). In addition, the multidimensionality of the narcissism construct was investigated, in which both the effects of adaptive (more normal leadership behaviours) and maladaptive (unsocial, dominating and manipulative behaviours) dimensions were explored. Elaborating on dominance complementarity theory, it was also investigated whether extraversion forms a moderator of the relationship between leader narcissism and employee challenging and affiliative OCB. Method: A sample of 322 respondents filled in an online or paper based questionnaire, from which 161 leader subordinate dyads could be matched. Results: The present study showed that narcissism, as expected, is a multidimensional construct in which a distinction must be made between adaptive and maladaptive narcissism. The leadership/authority facet (adaptive) had a positive relation with OCB whereas the exhibitionism/entitlement facet (maladaptive) showed a negative relation with OCB. These opposite effects are strong evidence for the multidimensionality of the narcissism construct. Furthermore, the study found no moderating effect of extraversion in the narcissism-OCB relationship. The degree of employee dominance did not interfere with the supervisor employee relationship. Conclusion: Narcissism must be treated as a multidimensional concept which is composed of two opposite dimensions. Further research should report both the score and effect size for global narcissism as the individual scores and effect sizes for adaptive and maladaptive narcissism.

(8)

1

Introduction

If you tell your friends, family or acquaintances that this research is about the effects of narcissistic leadership then there is always this sign, mostly a smile, of recognition. It appears, if you ask any employee if they have ever had a narcissistic boss you would be surprised how many people, apparently, have come in contact with narcissistic leadership. Most stories end with devastating leadership judgments and the deteriorating effects of it on the organisation or team performance. In addition to its public attention, narcissism is also a subject of interest for many researchers in a variety of disciplines (e.g. Ackerman et al., 2011; Miller & Campbell, 2008). A quick glance at the published work about narcissism resulted in more than 100.000 (Google scholar) publications and 6.800 (PsycINFO) publications from various research fields. Increasing research attention has turned to the determination of the effects of narcissism in organisations (Grijalva, Harms, Newman, Gaddis, & Fraley, 2013).

Narcissism can clearly be seen, from a leadership perspective, as a blessing and a curse (De Hoogh, Den Hartog, & Nevicka, 2013; Hochwarter & Thompson, 2012). In addition, there is still considerable uncertainty about the effects of narcissism in organisations (Meurs, Fox, Kessler, & Spector, 2013; Rosenthal & Pittinsky, 2006). Narcissism, which has a negative connotation, is associated with a desire for authority and power, egocentrism, the need to be admired and a lack of empathy. It describes someone who does not tolerate contradiction, who gets quickly frustrated and has an inflated self-view (Campbell, Hoffman, Campbell, & Marchisio, 2011; Hoffman et al., 2013). On the other hand narcissists are confident, often charismatic and optimistic, which are seen as positive leadership traits (Rosenthal & Pittinsky, 2006). Therefore narcissism is positively linked to leader emergence (Brunell et al., 2008; Nevicka, De Hoogh, Van Vianen, Beersma, & McIlwain, 2011), confidence (e.g. Emmons, 1984), extraversion (e.g. Grijalva et al., 2013) and narcissists are perceived as more intelligent (Paulhus, 1998). Campbell and colleagues (2011) refer to these

(9)

2 seemingly contradictory sides of narcissism as its ‘bright’ and ‘dark’ sides. This terminology has been adopted by other researchers and is sometimes referred to as adaptive and maladaptive narcissism (Ackerman et al., 2011).

The past decade can be characterized as the quest for answering the question whether the effects of narcissism are positive or negative. Many recent studies tried to find an answer and several insightful contributions were made. Some of those studies report predominantly positive findings in the relationship between narcissism and effective leadership, such as the positive relation of narcissism and peer perception of leadership ratings (Judge, LePine, & Rich, 2006). Galvin, Waldman and Balthazard (2010) have found a positive link too between charismatic leadership and narcissism through the visionary boldness aspect. Nevertheless most of the studies on narcissism explored its more negative implications. Resick, Whitman, Weingarden and Hiller (2009), for example found evidence that manager turnover was higher in firms which were led by narcissistic CEO’s. Another negative effect established was on team performance through reduced group-level information exchange (Nevicka et al., 2011).

On closer examination it turns out that the published effects of narcissism are rather scattered, both positive or negative and significant versus non-significant. One possible explanation for this conclusion comes from Miller and Campbell (2008) who state that a lack of consensus on how to conceptualise narcissism underlies these differences. Another possible explanation is proposed by Ackerman and colleagues (2011), elaborating on the research of Miller and Campbell. They state that, in the social-personality research, narcissism turned out to be both adaptive (positive) and maladaptive (negative). As such, recent debates focus on whether the Narcissistic Personality Inventory (NPI) is a valid measure to investigate non-clinical forms of narcissism (e.g. Ackerman et al., 2011; Miller & Campbell, 2008; Rosenthal, Matthew Montoya, Ridings, Rieck, & Hooley, 2011). Scholars argued that narcissism should be seen as a multidimensional concept in which it is important

(10)

3 to make a distinction between global and facet levels of narcissism (Grijalva & Newman, 2014), in which it is possible to make a distinction between mere adaptive facets and maladaptive facets of narcissism. Ackerman and colleagues (2011) state that it might be undesirable to use a total narcissism (global) score in which the various adaptive and maladaptive personality dimensions might blend into a composition and rule each other out. The present study addresses this issue, among other things, by investigating the multidimensionality of the narcissism construct.

As described earlier, social-personality research is predominantly interested in the effects of narcissism on, for example, workplace outcomes. A recent study on the effects of narcissism and workplace behaviour, which used the same facet structure as developed by Ackerman and colleagues (2011), investigated the effects of narcissism on counterproductive work behaviour (CWB) (Grijalva & Newman, 2014). They found a predominantly negative relationship between narcissism and CWB. Interestingly they also found evidence for the multidimensionality of the narcissism construct. The maladaptive facet was positively linked to CWB and the opposite was true for the adaptive facet, which was negatively linked to CWB. Nevertheless Grijalva and Newman (2014) continued to focus on the mere negative (maladaptive) effects of narcissism without elaborating on some of the positive (adaptive) narcissistic traits such as stability, extraversion, charisma and leader emergence.

Because of this limited focus, the present study aims to contribute to the multidimensionality discussion, following the study of Grijalva and Newman (2014), but by focussing on the effects of narcissistic leadership on positive, instead of negative, organisational behaviours. The present study will start with a general literature review about narcissism, positive workplace behaviours and the relationship between both. Secondly, empirical research will be presented to answer the question whether narcissism is related to positive workplace behaviours and the question how the previously mentioned sub-facets of

(11)

4 narcissism can be used in order to determine if the narcissism sub-facets are differently related to possible positive workplace behaviour. Grijalva et al. (2013) have concluded in recent research that it remains unclear whether certain types of employees, for example confident or passive employees, experience and judge the relationship with a narcissistic leader differently. Therefore, Grijalva and colleagues suggest to focus on employee type in the research on the effects of narcissism in organisations. For this reason, the influence of employee type on the relationship between narcissism and positive workplace behaviours will be examined too.

In summary, the present study (1) seeks to expand the existing literature in three ways; through the investigation of the possible link between narcissism and positive workplace behaviours; through the exploration of the question if the narcissism sub-facets are differently related to possible positive workplace behaviour; and through the determination of a possible influence of employee type on the narcissism positive workplace behaviours relationship. (2) This study uses an empirical approach to seek for construct clarification and the expected effects of narcissism on positive workplace behaviours and it aims to contribute to practice (3) in terms of how to distinguish between adaptive and maladaptive narcissism.

Literature review Narcissism

Narcissism is defined and used by both social-personality and clinical psychology professionals (Campbell et al., 2011). Narcissism officially emerged as a mental disorder in the DSM III. The DSM is a diagnostic manual used by practitioners to asses mental disorders, and the DSM the Narcissistic Personality Disorder (NPD) is found in the B cluster (Miller & Campbell, 2008). NPD is characterized by nine diagnostic criteria (e.g. has a lack of empathy; tends to be exploitative, manipulative, and arrogant; DSM-IV-TR; APA, (2000)

(12)

5 p.717). A person must have five out of the nine traits before NPD can be diagnosed. The NPD diagnosis is quite strict one and cannot simply be assigned to everyone with narcissistic traits. In addition to complying with the DSM criteria, one must also suffer from broken relationships and poor work performance (Campbell et al., 2011). Because of this situation, a discussion of the narcissism terminology used in the social-personality research field has started. High scores on the measure of the trait of narcissism do not necessarily mean suffering from this disorder (Grijalva et al., 2013). Social-personality researchers are not necessarily interested in the clinical or pathological form of narcissism but in any case, use the NPI, which is based on the definition of narcissism as formulated in the DSM (Ackerman et al., 2011; Corry, Merritt, Mrug, & Pamp, 2008; Miller & Campbell, 2008; Rosenthal et al., 2011). This has resulted in a definition of narcissism based on subclinical form of narcissism which is used in social-personality research. This distinction is important because the present study uses a non-clinical sample to examine the effects of narcissism in organisations.

Narcissism and its dimensional structures. A well-known measure in the

social-personality research of narcissism is the Narcissism Personality Inventory (NPI) (O’Boyle, Forsyth, Banks, & McDaniel, 2012). The NPI consist of 40 paired statements where a subject is forced to choose between one of the statements (e.g. “I like to be the center of attention.” and “everybody likes to hear my stories.”) (Raskin & Terry, 1988). Subjects are forced to choose between one of the two options from which one option is more likely to be chosen by narcissistic individuals. Despite the fact that the NPI is widely used to measure narcissism a recent scholarly debate raised concerns about (1) the consistency of the operationalization; (2) the ability to capture the core personality attributes; and (3) the reliability and interpretability of the factor structure (Ackerman, Donnellan, & Robins, 2012).

The latter concern about the different factor solution, which is addressed in the above mentioned study of Ackerman et al. (2012), could be an underlying reason for the positive

(13)

6 and negative reported effects of leader narcissism on perceived effectiveness. The problem with the NPI lies in the fact that researchers try to distinguish between pathological and non-pathological narcissism while the scale finds its origin in the psychology for clinical diagnostics (O’Boyle et al., 2012). In order to be able to make this distinction, researchers have created sub scales (Ackerman et al., 2012). Raskin and Terry (1988) proposed a factor structure with seven sub-dimensions; authority, self-sufficiency, superiority, vanity, exhibitionism, entitlement, and exploitativeness. Due to low internal consistency of the seven sub-dimensions, Corry et al. (2008) established a two factor model with the same 40 item of the NPI measure. The first dimension encompasses the Leadership/Authority facet and the second dimension covers the Exhibitionism/Entitlement facet. It is interesting, in the light of the present study, that this two factor solution distinguishes between adaptive and maladaptive narcissism (Ackerman et al., 2012). Ackerman et al. (2011) proposed a three factor solution where the sub-dimensions consist of Leadership/Authority (adaptive), Grandiose Exhibitionism (maladaptive) and Entitlement/Exploitativeness (maladaptive). The first dimension, as a whole, corresponds to first dimension of Corry et al. (2008) and the second dimension of Corry and colleagues has been divided into two separate scales, which is in line with the theoretical claims to separate grandiosity and entitlement (Ackerman et al., 2011).

Adaptive and Maladaptive narcissism. Scholars have made a theoretical distinction

between adaptive and maladaptive narcissism (Ackerman et al., 2011; Fox & Freeman, 2011). The Leadership/Authority dimension, as in Corry et al. (2008) and Ackerman et al. (2011), represents the adaptive aspects of narcissism and the other two dimensions form maladaptive aspects of narcissism. Individuals who score high on the Leadership/Authority dimension, which represents adaptive narcissism, tend to show more normal leadership behaviour like assertiveness, self-confidence and the need for achievement (e.g. Barry, Frick,

(14)

7 & Killian, 2003; Raskin & Terry, 1988). People who score high on the Leadership/Authority dimension like to take the lead, and they like to take the responsibility for their decisions (Raskin & Terry, 1988). The two other dimensions represent the maladaptive aspects of narcissism. Individuals who score high on the Exploitativeness/Entitlement or the Grandiose/Exhibitionism dimension tend to show more uncooperative and unsocial dominating behaviour (e.g. Barry et al., 2003; Meurs et al., 2013; Raskin & Terry, 1988)). They easily manipulate people and have a desire to view themselves as more important than others (Emmons, 1984; Raskin & Terry, 1988).

Narcissism as a one-dimensional construct. In contrast to these above mentioned

proposed multidimensional models, most of the published work uses mean scale scores for overall narcissism in their models (Miller & Campbell, 2008). This approach can be problematic in a way that different personality dimensions could conflate, such as the positive self-attitude and the more negative need to exploit (Ackerman et al., 2011). As the NPI seems to capture both adaptive and maladaptive aspects of narcissism, which possibly rule each other out (Barry et al., 2003). A total score on the NPI measures both leadership ability as well as the willingness to exploit others (Meurs et al., 2013). Therefore, the meta-analysis of Grijalva et al. (2013) acknowledge the necessity to perform future empirical research in order to find evidence for the effects of individual sub-dimension on the narcissism-leadership relationship. To contribute to this appeal, the present study will attempt to create a model in which the sub-dimensions will be individually tested to measure their effects.

Narcissism in organisational contexts. The link between narcissism and leadership

has been established by many scholars (e.g. Brunell et al., 2008; Campbell et al., 2011; Rosenthal & Pittinsky, 2006). Nevertheless, an essential question remains unanswered, how should one approach narcissism (Campbell et al., 2011)? Is narcissism only destructive for organisations or can it have positive effects too? Campbell and Campbell (2009), use the

(15)

8 contextual reinforcement model in explaining the reported differences in the effect of narcissism in organisations. They conclude that narcissism can be adaptive or maladaptive depending on the context. The contextual reinforcement model describes in which areas narcissism will be effective and ineffective. According to the model narcissism is beneficial in unstable and chaotic situations in which leader emergence plays a vital role and narcissism is harmful in stable and calm situations in which leader stability and enduring relationships plays a vital role (Campbell et al., 2011). Campbell and colleagues still acknowledge the fact that the reported results of narcissism effectiveness are rather scattered. Campbell et al. (2011) state, because of the opposing results, that more explanatory research of the influence of narcissism effectiveness is needed. Therefore, the present study will not use the contextual reinforcement model, but focus on the multidimensionality of the narcissism construct.

Organisational citizenship behaviour

To be able to contribute to the exploration of the multidimensionality of narcissism in organisations it is necessary to focus on the effects of narcissistic leadership on pro-organisational behaviours. The focus on pro-pro-organisational behaviours will help in the clarification of the multidimensionality of the narcissism construct, because it is generally assumed that narcissism leads to negative organisational outcomes (Nevicka et al., 2011). So, the question which this present study aims to answer is if narcissism can also lead to pro-organisational behaviours.

Organisational citizenship behaviour (OCB) is commonly known as behaviour that can be characterized as activities which indirectly contribute to the organisation (Van Dyne & LePine, 1998). It is generally defined as a set of pro-organisational behaviours which go beyond, and are not described in, a employee’s job description (Podsakoff, Whiting, Podsakoff, & Blume, 2009). OCB is conceptualized in the book written by Organ (1988). He defined this extra-role behaviour as “individual behaviour that is discretionary, not directly or

(16)

9 explicitly recognized by the formal reward system, and that in the aggregate promotes the effective functioning of the organization” (p. 4). For example, this pro-organisational behaviour includes refilling the paper of the printer, helping colleagues to lessen their workload and contributing to a healthy work-related atmosphere.

Leadership behaviour as antecedent. As OCB is defined as pro-organisational

behaviour, the following interesting question comes to mind; how do organisations influence or promote this discretional behaviour? According to MacKenzie et al. (2011) a lot of research has been devoted to the antecedents of OCB, antecedents as commitment (e.g. LePine, Erez, & Johnson, 2002), personality traits (e.g. Chiaburu, Oh, Berry, Li, & Gardner, 2011) and leadership related factors (Podsakoff, MacKenzie, Paine, & Bachrach, 2000). Following the conclusions of Podsakoff et al. (2000), leaders have an important role in establishing and maintaining citizenship behaviours. Especially transformational leadership is correlated to organisational OCB, because of the nature of transformational leadership which stimulates performance above and beyond expectations (Podsakoff et al., 2000).

Narcissism and OCB. A lot of research (e.g. O’Boyle et al., 2012) dealt with the link

between narcissism and negative work place behaviours or negative organisational outcomes, like bullying, aggression and frustration. Those studies are mainly focused on the narcissistic employee and his or her behavioural outcomes. The present study tries to broaden this scope of influence and explores the effects of narcissistic leadership. It would be inappropriate to assume that narcissistic leadership is the only factor that influences employee’s OCB, but leadership style forms an explanatory mechanism. The recent study of Grijalva and Newman (2014) investigated the relationship between narcissism and counterproductive work behaviour (CWB). They found that narcissism is a strong and significant positive predictor of CWB. CWB can be defined as voluntary behaviour that does not fit the organisational standards and sometimes even threatens the welfare of colleagues and or the organization as a

(17)

10 whole (Bennett & Robinson, 2000; Sackett, Berry, Wiemann, & Laczo, 2006). Narcissistic individuals tend to exhibit more CWB in the organization (Spector & Fox, 2010). This raises the question how to relate these findings to pro-organizational behaviour? Both OCB and CWB consist of active actions, carried out by individuals, beyond a task or job description (Judge et al., 2006). Both behaviours are associated with constructive (OCB) or destructive (CWB) behaviours and are focused towards the organisation or individuals (Spector & Fox, 2010). Most studies that explored the relationship between OCB and CWB have found negative correlations between the two constructs (e.g. Judge et al., 2006; Sackett et al., 2006). According to Sackett and colleagues (2006) this serves as evidence for a theoretical distinction between the two constructs.

Self- and other-rated OCB. To what extent is a leader capable to observe

organisational citizenship behaviours of his or her employee? According to Carpenter, Berry and Houston (2014), there is no significant difference between self and other ratings of OCB. In particular, the ratings of employee and supervisor were generally consistent, on multiple levels, in their meta-analysis. They based their findings on analysis which showed that other-ratings of OCB did not contribute to the already explained unique information by self-rated OCB (Carpenter et al., 2014). In conclusion, according to Carpenter and colleagues, supervisors are just as capable in rating OCB as employees themselves.

Taking into account the theoretically opposite relation between OCB and CWB (e.g. Judge et al., 2006; Sackett et al., 2006; Spector & Fox, 2010) and the conclusion of Grijalva and Newman (2014) that narcissism is positively related to CWB, it is expected that narcissism will have a negative relation to OCB.

Hypothesis 1a: Narcissism is negatively related to organisational citizenship behaviour.

(18)

11

Narcissism’s facet structure and OCB. A striking result in the study of Grijalva and

Newman (2014) has to do with the distinction between the sub-facets of the narcissism concept as identified by Ackerman and colleagues (2011). Grijalva and Newman (2014), found that the adaptive (Leadership/Authority) facet had a negative relation with CWB. In other words, individuals who scores high on the adaptive narcissism facet tend to show less disruptive organisational behaviour. It is thus expected, taking into account the opposite relation between OCB and CWB, that adaptive narcissism has a positive relation with OCB.

On the other hand, people with high scores on the maladaptive (Grandiose/Exhibitionism) facet showed a positive relationship with the level of CWB, in other words those people tend to exhibit more negative organisational behaviour (Grijalva & Newman, 2014). Thus it is hypnotized that the maladaptive narcissism facet will be negatively related to OCB.

In the present study the design of the study of Grijalva and Newman (2014) will be followed in the search for the effects of narcissism on perceived employee pro-organisational behaviour (OCB). The assumption is that this would probably highlight the negative consequences of narcissistic leadership, or will serve as a promising start for further exploration of the narcissism sub-facets.

Flavours of OCB. In the present study a distinction is made between two dimension

of OCB, challenging and affiliative OCB (Van Dyne & LePine, 1998). Challenging OCB can be characterized as change oriented proactive behaviour, or employee voice (Kim, Van Dyne,

Hypothesis 1b: The ‘adaptive’ facet is positively related to organisational citizenship behaviour.

Hypothesis 1c: The ‘maladaptive’ facet is negatively related to organisational citizenship behaviour.

(19)

12 Kamdar, & Johnson, 2013). Employees tend to speak out their desire to question the organisational status quo, by providing constructive criticism (MacKenzie et al., 2011). A narcissistic supervisor might be extra sensitive for proactive, change-related behaviour, as will be explained in the dominance complementarity section (Grant, Gino, & Hofmann, 2011). It is expected that a distinction between the OCB dimensions has a dampening effect on the relationship between narcissism and challenging OCB. The reason for this assumption is that narcissism and extraversion are highly correlated and that opposite dominance dimensions lead to more favourable relationships and similar dominance dimensions lead to more hampering relationships (Grant et al., 2011; Grijalva & Harms, 2014; Sadler, Ethier, Gunn, Duong, & Woody, 2009).

Affiliative behaviours, or affiliative OCB, are focused on maintaining and establishing good interpersonal relations with others. Helping, which is also labelled as altruism, is one of the most studied forms of this behaviour (e.g. Organ, 1988; Van Dyne & LePine, 1998). The expectation is that affiliative OCB has no dampening effect on the relationship with narcissism, given that opposite dominance dimensions (extraversion and altruism) lead to more favourable relationships (Sadler et al., 2009).

Hypothesis 2a: The ‘adaptive’ facet is positively related to challenging organisational citizenship behaviour.

Hypothesis 2b: The ‘maladaptive’ facet is negatively related to challenging organisational citizenship behaviour.

Hypothesis 2c: The ‘adaptive’ facet is positively related to affiliative organisational citizenship behaviour.

Hypothesis 2d: The ‘maladaptive’ facet is negatively related to affiliative organisational citizenship behaviour.

(20)

13 A theoretical distinction between the two OCB dimensions is, according to Van Dyne and LePine (1998), legitimate. They conclude in their study that the two OCB dimensions explained enough unique variance to be able to make a distinction between affiliative and challenging OCB.

Dominance complementarity theory

After the exploration of the multidimensionality of the narcissism construct, the question remains as to whether circumstances exist that affect the effects of narcissistic leadership in organisations. An encouraging recommendation comes from the meta-analysis of Grijalva and colleagues (2013), in which they recommend that it would be interesting to investigate whether different types of employees (e.g. confident vs. passive) influence the effects of narcissistic leadership in organisations. As leadership is not just about the behaviour of the leader itself but also about the followers, the dynamic interaction between the characteristics of the narcissistic leader and the characteristics of the followers will be examined.

Leadership trait theory. In the long history of leadership research, trait theories have

been extensively studied for the understanding of leadership effectiveness, wherein leadership performances generally are attributed to personality traits such as intelligence and dominance (Mumford, Zaccaro, Connelly, & Marks, 2000). In the context of leadership traits, the big five personality traits are often related to leader effectiveness. Of those personality traits, extraversion is considered to have the strongest link with leader emergence and leader effectiveness (Judge, Bono, Ilies, & Gerhardt, 2002; Judge & Bono, 2000). Judge et al. (2002), found that employees who scored high on the extraversion measure were more likely

(21)

14 to be perceived as effective and to emerge as leaders. To summarize, research has shown that extroverts have a clear advantage in leadership roles (Grant et al., 2011). But, according to the same study by Grant and colleagues (2011), it is possible that, while extraversion is a good predictor of leadership effectiveness and emergence, extraversion does not always automatically lead to a positive organisational contributions.

Interpersonal complementarity theory. According to Grant and colleagues (2011),

this may be due to the dynamic interplay between leader and followers, as leadership is a two-way interventional process. This insight, proposed by Leary in the late 1950s, led to a different way of examining the leadership-effectiveness relationship (Grijalva & Harms, 2014). This field of research (complementarity theory) acknowledged that it is not desirable to solely focus on the characteristics of an effective leader, but to explore those effective leadership traits in combination with the characteristics of the followers (Grijalva & Harms, 2014). Research on complementarity theory states that complementarity requires similar levels of affiliation (friendliness invites friendliness) and opposing levels of dominance (dominance fits with submission) (Grijalva & Harms, 2014). Where high levels of complementarity result in more satisfying and stable relationships (Grant et al., 2011; Grijalva & Harms, 2014; Sadler et al., 2009). Besides satisfying relationships, it also appears that complementarity in relationships enhances objective performance effects (Grant et al., 2011).

In addition, the dominance complementarity theory has a special focus for the degree of dominance of leader and follower. The theory suggests that opposite dominance levels, in which dominance corresponds with submissiveness, leads to better organisational and interpersonal outcomes (Grant et al., 2011). The application of this theory might have consequences for the proposed effects of narcissistic leadership on OCB, because narcissists

(22)

15 are known as highly extraverted and dominant people (e.g. Emmons, 1984; Grijalva & Harms, 2014; Raskin & Terry, 1988).

Narcissistic dominance. Much research attention has been paid to the mere negative

effects of narcissism. The few reported positive effects of narcissism could be, in most cases, attributed to the same characteristics which are attributed to good and effective leadership (Judge et al., 2002). Good examples of these characteristics are dominance and extraversion, typical leadership traits, which to some extent explain the positive reported effects of narcissism in organisations (Grijalva et al., 2013). Nevicka and colleagues (2011) found a positive relation between extraversion and leader emergence and Grijalva and colleagues (2013) found an even stronger effect and conclude that extraversion fully mediates the relationship between narcissism and leadership emergence. Extraversion is, in a narcissistic context, best understood as behaviour in which individuals like to be in the centre of attention, act dominant and are assertive (Grant et al., 2011).

Grant and colleagues (2011) showed that group performance improved when extraverted leaders collaborated with passive employees. They supported their findings with a possible explanation that extraverted leaders may feel threatened by proactive employees. This effect may be even stronger for narcissistic leaders because they are hypersensitive to negative information (e.g. questioning leadership decisions) (Penney & Spector, 2002), infringement of positive self-image (Campbell et al., 2011) and ego-threats (Hochwarter & Thompson, 2012). It is therefore expected, and this is supported by the findings of (Podsakoff, Podsakoff, MacKenzie, Maynes, & Spoelma, 2014), that a narcissistic supervisor feels threatened by an extraverted employee which has a negative impact on the leader’s rating of perceived OCB. Based on the above arguments, it is hypothesized that:

Hypothesis 3a: Extraversion moderates the relationship between ‘adaptive’ narcissism and challenging OCB such that adaptive

(23)

16 A conceptual research model has been developed that links extraversion to adaptive and maladaptive narcissism. See figure one for the theoretical framework that guides the present study.

narcissism will have a weaker positive relationship with challenging OCB.

Hypothesis 3b: Extraversion moderates the relationship between ‘adaptive’ narcissism and affiliative OCB such that adaptive narcissism will have a weaker positive relationship with affiliative OCB

Hypothesis 3c: Extraversion moderates the relationship between ‘maladaptive’ narcissism and challenging OCB such that maladaptive narcissism will have a stronger negative relationship with challenging OCB.

Hypothesis 3d: Extraversion moderates the relationship between ‘maladaptive’ narcissism and affiliative OCB such that maladaptive narcissism will have a stronger negative relationship with affiliative OCB.

(24)

17

Methods

The research design used in this study is cross-sectional in nature and will be assessed using a self-report survey.

Sample and procedure

The present study is part of a broader study about the effects of leadership and behaviour in organisations. Therefore, there is also data collected that will not be used in this study. The present study investigates whether specific (mal)adaptive leadership behaviours affect organisational aimed employee behaviours. Which makes all Dutch employed people part of the general population. In order to say something about the effects of leadership behaviour, dyads are examined. The sampling frame of this study consists of supervisor/employee dyads working in different organisations. The respondents were approached, using own contacts according to the convenience sampling method, asking if they wanted to participate in the study of leadership and behaviour. This sampling method was used due to limited resources available to conduct for example probability sampling but also because of the fact that potential subjects were readily available. Respondents who were willing to participate in the study received a web-link for the online survey or a survey package containing the paper based questionnaire, depending on preference. Each survey was accompanied with an informative introduction letter, stressing the voluntary character and confidential processing of the questionnaire. The paper-based surveys was also provided with a freepost number envelope.

A total of 322 individuals completed the questionnaire, which represents a response rate of 69%. Of those completed questionnaires 161 leader subordinate dyads could be matched. The leaders’ average organisational tenure was 9.28 years (SD = 7.84) and 36% of those leaders were female. The average leaders’ age was 40.80 years (SD = 10.36) and almost 93% of the leaders worked full time. More than 87% of the leaders has indicated to have a

(25)

18 higher education (HBO or WO) diploma. The subordinates (62% were female) averaged 30.79 in age (SD = 10.21) and 65% worked full time. The subordinates’ average organisational tenure was 4.23 years (SD = 4.85) and 78% obtained a higher education diploma.

Measures

Narcissism. The narcissism construct is measured using the Narcissistic Personality

Inventory (NPI) which is developed by Raskin and Terry (1988). An example item is: ‘I know that I am good because everyone keeps telling me so’. The NPI, as used in the study of Raskin and Terry, is a 40-item forced choice measure in which respondents must choose between two divergent statements. This original dichotomous forced-choice format is for example used by Grijalva and Newman (2014) in their recent study about narcissism and counterproductive work behaviour, reporting an Cronbach’s alpha of 0.85. Other recent studies investigated the NPI with a Likert scale response format, in order to be able to examine the NPI facet structure which requires more complex calculations (O’Boyle et al., 2012). Barelds and Dijkstra (2010), used a five point Likert scale, for their Dutch translated NPI questionnaire, in order to investigate the narcissism facet structure and they reported a Cronbach’s alpha of 0.93. Instead of the five point Likert scale format of the NPI, the present study uses a 7 point Likert scale for the 40 narcissistic expressed items in order to provide consistency in the formats and reduce confusion among respondents.

Organisational Citizenship Behaviour. The general OCB concept is measured as a

composite of two OCB dimensions. To measure the OCB concept a 12 items measure is used, from which five items are based on the OCB voice measure created by LePine and Van Dyne (1998). An example item is: ‘The employee speaks up and encourages others in this group to get involved in issues that affect the group’. They reported high internal validity with an Cronbach’s alpha of 0.95 (LePine & Dyne, 1998). The other seven items are based on the

(26)

19 research from MacKenzie, Podsakoff, and Fetter (1991) and tend to measure OCB (affiliative). The first three items measure the sub-dimension altruism and the other four measure courtesy (MacKenzie et al., 1991). The OCB measurements are based on reports from the supervisors and use a 7 point Likert scale response format.

Big Five. The subordinates personality traits were measured using the shortened, 25

items mini-IPIP, based on the study of Donnellan, Oswald, Baird and Lucas (2006). The shortened measures is used to evaluate the five different personality dimensions as extraversion agreeableness, conscientiousness, neuroticism, and openness (or intellect/imagination) (Donnellan et al., 2006). For all 25 items a 7 point Liker response scale is used. Prior studies demonstrated alpha reliabilities in the range of .79 to .84 (Donnellan et al., 2006).

Data preparation

In order to conduct the preliminary analyses an assessment whether the data set met the requirements to conduct confirmatory factor analysis and linear regression is made. Of the three scales used in the present paper the OCB-scale yield no missing values, the Big Five has 3 missing values and the NPI holds 11 missing values. A MCAR test is conducted in order to establish if the data is missing randomly. The test resulted in MCAR p > .05, thus data is missing randomly. In the present study ‘HOT DECK’ is used to replace the missing values. Of the initial 161 cases, 159 cases turn out to be eligible for analysis, after the HOT DECK procedure. The data set was also checked for normality with the Kolmogorov-Smirnov test. The NPI scale was significant (p > .05), normality is assumed. The other scales were not significant, thus normality could not be assumed. The data did not yield any outliers. All the variables scored well below 2.5 times the Standard Deviation (SD). An analysis for heteroscedasticity or the residuals is conducted. No indications of heteroscedasticity were found. Finally no items are eligible for recoding.

(27)

20

Confirmatory factor analysis

In order to establish whether a three or two factor solution should be used in this study a confirmatory factor analysis (CFA) is conducted to test the two models. Data was analysed using the AMOS statistical program version 22. The first model is a 25 items three-factor model proposed by Ackerman and colleagues (2011), representing the following sub-facets; leadership/authority, grandiose/exhibitionism and entitlement/exploitativeness. The second model is a 23 items two-factor model, identified by Corry and colleagues (2008), representing the domains of leadership/authority and exhibitionism/entitlement.

Confirmatory Factor Analysis of the Three-Factor Solution. The initial model fit

of the three-factor solution was not acceptable; χ2 (d.f. 272, n = 325) = 702,047 p < .001, CFI = .719, TLI = .690, root mean square error of approximation (RMSEA) = .100. However, after allowing items 27 and 34, 40 and 34, 15 and 26, 19 and 29 to correlate the model was more suitable, see table 1 for the model indexes.

Confirmatory Factor Analysis of the Two-Factor Solution. The proposed second

model incorporated 23 of the 40 NPI items. The initial model fit was also insufficient; χ2 (d.f. 229, n = 276) = 510,736 p < .001, CFI = .773, TLI = .749, RMSEA = .088. However, after allowing items 27 and 10, 11 and 27, 7 and 30, 16 and 24 to correlate, the indices revealed a viable model fit, see table 1 for the model indexes.

Table 1 CFA Fit Indexes of NPI Solutions

Model χ2 (df) CFI RMSEA TLI

Three-factor 591,736 (268) ,788 ,087 ,763

Two-factor 365,912 (222) ,881 ,064 ,852

Note: CFA = Confirmatory Factor Analysis; NPI = Narcissistic Personality Inventory; CFI = Comparative Fit

Index; RMSEA = root mean square error of approximation; TLI = Tucker–Lewis index; ML = Maximum Likelihood.

(28)

21 Most common used cut-off scores are reported in the research of Hu and Bentler (1999). They suggested the following guidelines as an alternative to the, relatively responsive to sample size, chi-square. A acceptable model holds a minimum score of >,95 for the comparative fit index (CFI) and the Tucker-Lewis index (TLI) measure, and a value of <,06 for the RMSEA measure. The second model with the of leadership/authority and exhibitionism/entitlement domains will be used in the present paper, since the indicators for model fit are higher for two-factor solution. However, the results should be interpreted with caution because not all the, by Hu and Bentler (1999), reported indicators for good model fit are met. In order to compute the leadership/authority and exhibitionism/entitlement scales, the factors scores are used instead of the less accurate scale averages. By this procedure, the scales take into account the relative weight of the contribution of an item to the latent factor.

Common Method Bias. Common method bias could be a problem in this study,

because of the research design chosen with self-report scales and the socially sensitive questions. Ideally, the narcissism and perceived OCB measures should be collected from different sources. In the present study they are both measured in the supervisors questionnaire. Given the nature of the type of questions, it is possible that socially desirable answers are given. Moreover, a questionnaire completed by a narcissistic supervisor might also contain a lower rate of perceived OCB, due to the nature of narcissism, in comparison to a non-narcissistic leader. Therefore the model is checked for common method bias, according to the procedure suggest by Podsakoff, MacKenzie, Lee and Podsakoff (2003). The procedure to follow is a comparison of the standardized regression weights in a model with and without the common latent factor. The analysis of the standardized regression weights resulted in signs of common method bias for the EE scale. Therefore common method bias adjusted factors are used in order to control for common method bias for the EE scale.

(29)

22

Results

Means, standard deviations and correlations among the variables in the present study are reported in table 2. The correlation between extraversion and agreeableness can be explained by the fact that, according to Podsakoff, and colleagues (2000), the dimensions of extraversion consists of friendliness and gregariousness.

Table 2 Means, standard deviations, and intercorrelations of the measures

M SD 1 2 3 4 5 6 7 8 9 10 11 1. Narcissism 4,27 ,73 (.94) 2. Leadership/Authority 4,75 ,80 ,876** (.83) 3. Exhibitionism/Entitlement 3,97 ,82 ,963** ,834** (.85) 4. OCB 5,43 ,73 ,020** ,093** -,020** (.88) 5. OCB-CH 5,30 ,85 ,082** ,172** ,042** ,923** (.82) 6. OCB-AF 5,52 ,80 -,033** ,042** -,061** ,959** ,788** (.86) 7. Openness 4,74 ,83 ,108** ,005** ,106** -,056** -,019** -,078** (.62) 8. Neuroticism 3,35 1,11 ,106** ,067** ,099** -,094** -,108** -,093** ,140** (.82) 9. Conscientiousness 5.22 ,91 -,018** -,022** -,037** ,080** ,085** ,079** -,143** -,204** (.78) 10. Agreeableness 5,38 ,75 -,056** -,034** -,097** ,068** ,033** ,081** ,191** -,014** ,164** (.76) 11. Extraversion 4,69 ,95 ,042** -,039** ,033** -,007** ,046** -,034** ,460** -,122** ,165** ,473** (.75)

Note: N = 159. * p < .05/ ** p < .01. Cronbach’s α reliability coefficients are displayed in the diagonal.

Multiple regression analysis was used to test the first two hypotheses. Hypothesis 1a assumed that higher levels of supervisor narcissism are negatively related to OCB. Narcissism is regressed on OCB. Leader narcissism did not have a significant effect on OCB (β = .02, p = .79), therefore there is no support for Hypothesis 1a. The second part of the first hypothesis examines the effects of the L/A and E/E facets on OCB. The proposed model significantly fits the data, (F = 3.297, p < .05, R2 = .04). In support of Hypothesis 1b, the L/A facet predicts OCB positively (β = .36, p < .05) and Hypothesis 1c is supported too because the E/E facet predicts overall OCB negatively (β = -.32, p < .05).

To test the second series of hypotheses Structural Equation Modelling is used. The second Hypothesis addresses the facet level relationships with the sub-dimensions of OCB.

(30)

23 L/A (β = .35, p < .05) and E/E (β = -.25, p = NS) were related to Challenging OCB and together explained 4% of the variance in OCB-CH, supporting Hypothesis 2a and giving no support for Hypothesis 2b. The L/A (β = .31, p < .05) and E/E (β = -.31, p < .05) facet were significantly related to Affiliative OCB and collectively explained 3% of the variance in OCB-AF, which is in support of Hypotheses 2c and 2d (see figure 2).

Hierarchical moderated multiple regression is used in order to test the third series of Hypothesis which involves the employee type (confident vs. passive) moderation of the narcissism-OCB relationship, questioned by Grijalva et al. (2013). This analysis is performed on the basis of the procedure mentioned by Aiken and West (1991).

In the Hierarchical Regression Analysis the control variables, the other four dimensions of the Big Five, were entered in step one. The main effects were added in the second step and the third step contained the interaction term. The interactions are separately tested for the two dimensions of OCB, as show in table 3 for L/A and table 4 for E/E (appendix A). No support was found in favour for the moderating effect of extraversion on the narcissism-OCB relationship. The interaction was not found to be statistically significant, thus rejecting Hypothesis 3a, b, c and d.

(31)

24

Discussion

The purpose of the present study was to examine the relationship between narcissistic leadership and organisational citizenship behaviours by focussing on the multidimensionality of the narcissism construct. The findings of the present study suggest three main conclusions. First, global narcissism has no significant relationship with perceived OCB. That is, no statistical relationship could be established between the degree of supervisor narcissism and the level of perceived OCB. Interestingly the results contradict the findings of Judge et al. (2006), who were able to establish a significant negative relation between supervisor narcissism and OCB. A possible explanation comes from the fact that OCB, in the Judge et al. (2006) study, was measured on the same single person source as narcissism. Which was not the case in the present study, were perceived OCB was measured on a supervisor level. It is noteworthy that this initial result served as the encouragement to explore the possibility of the multidimensionality of the narcissism construct. This is in line with the reasoning in previous studies (e.g. Ackerman et al., 2011), stating that the usage of global narcissism might be undesirable because the different narcissism dimensions could rule each other out.

Second, the present study confirms the narcissism facet structure proposed by Corry et al. (2008) and used the leadership/authority and exhibitionism/entitlement dimensions for further analyses. The results showed that even though these dimensions were highly correlated and effects are not strong the leadership/authority dimension had a positive relation with perceived OCB and the exhibitionism/entitlement dimension a negative relation with perceived OCB. These findings are in line with the results presented in the study of Grijalva and Newman (2014) and as expected in opposite direction (Sackett et al., 2006; Spector & Fox, 2010).

The separation into two OCB dimensions, known as challenging and affiliative OCB, did not significantly change the narcissism-OCB relationship. On the basis of the literature

(32)

25 study, it was expected that, by distinguishing between challenging and affiliative OCB, the narcissism-OCB relationship would change. It was hypothesized that narcissistic supervisors might be extra sensitive to more assertive, change-related behaviours, which could have a dampening effect on the relationship between narcissism and challenging OCB (Grant et al., 2011). Unfortunately, this effect was not found in the present study. The L/A narcissism facet showed a positive relation with challenging and affiliative OCB and the E/E narcissism facet a negative relation with challenging and affiliative OCB. A possible explanation comes from the fact that the present study measured perceived OCB, from a supervisor perspective, instead of actual employee OCB. According to Carpenter and colleagues (2014), supervisors are just as capable to rate the level of OCB as employees themselves. But they also note that employees have a better sense of the direction and purpose of behaviours enacted than others. This might explain the difficulty, for a supervisor, in distinguishing between the two OCB concepts.

Third, the present study did not find a moderating effect of extraversion on the narcissism-OCB relationship. Thus, the dominance level of an employee did not affect the OCB perception of a narcissistic supervisor. The results contradicted the expectations as presented in the dominance complementarity theory because, it was expected that a narcissistic supervisor would felt threatened by their extraverted employee and that this effect would be stronger for someone who challenged the status quo (Podsakoff et al., 2014).

Potential explanations for the non-significant results are discussed below. First, what is actually measured with extraversion? Is it the tendency to be sociable and talkative or the degree of dominance and self-confidence (Lucas, Diener, Grob, Suh, & Shao, 2000). Of the Big Five traits extraversion has one of the broadest conceptualizations. It represents sociability, friendliness up to self-confidence and assertiveness (Judge et al., 2002). In the present study extraversion is measured with the mini-IPIP, a shortened Big Five measurement

(33)

26 scale (Donnellan et al., 2006). The mini-IPIP extraversion factor tends to overestimate the presence of the sociability facet. The study of Donnellan and colleagues (2006) showed that extraversion was to a large extent covered by friendliness and gregariousness and only a relatively small part was captured by assertiveness. This may partially explain the high correlation between agreeableness and extraversion (see Table 2). This finding could explain why the present study failed to establish a significant interaction between narcissism and extraversion. Given that the dominance complementarity theory requires similar dominance levels for a effect on the relationship between supervisor and employee. Phillips and Bedeian (1994) even found that extraverted employees developed better a relationship with their supervisor because they seek for interaction. This is totally counterintuitive to the dominance complementarity theory, but reasonable if one considers the presence of the sociability facet of extraversion. A strong focus on the sociability facet of extraversion could serve as an explanation for the fact that no moderating relationship was found. This could be a very interesting subject for future research, it is encouraged to further explore the effects of dominant extraversion in the narcissism employee relationship.

Second, an alternative explanation for the relationship between narcissism and perceived OCB may found in the inflated self-view of a narcissist. The inflated self-view overestimates own behaviour and disqualifies other behaviour (Grijalva & Harms, 2014). The disqualification of other behaviour could be a reasonable explanation for the negative link between the E/E narcissism facet and OCB. A supervisor who scored high on the maladaptive E/E narcissism scale could had have troubles in admitting the quality of others and giving credit to others.

Arguing the other way around, following the reasoning of Campbell and colleagues (2011), the systematic tendency to overestimate own performance could serve as a possible

(34)

27 explanation for the positive relation between LA and perceived OCB: “I’m a good supervisor so my employee must be performing OCB”.

The present study used common method bias adjusted factors in order to carry out the planned analysis. The reasoning of Podsakoff et al. (2003), who state that common method variance often could be a problem and that researchers need to control for it, is followed. The degree of variance is due to the design of the study (cross-sectional and self-report survey) or due to the nature of the type of questions (socially sensitive questions) instead of actual measured differences, thus it might that common method influenced the results. It should be noted that another stream of research warns against the use of common method factors (CMF). Conway and Lance (2010) concluded that it is a misconception that self-reported data may lead to bias due to its design. They did not find evidence for common method bias in self-reported data. Spector (2006) concludes that even social desirability is unlikely to cause inflation in the results. It was expected that participants would underreport social desirable items, but social desirability only modestly inflated the results (Spector, 2006). Given the high sensitivity of the questions asked, with a topic as narcissism, the choice was still made for a CMF in the present study.

The present study also identified a potential outlier issue. Technically, none of the data points were designated as outliers, given the procedure used to determine if a data point is eligible to be eliminated (2.5 times the SD, and none were). However, analysis showed that the results may still be sensitive to several less extreme outliers. While the choice was made not to remove any data points, it is important to note this.

Implications for theory and practice

The present study has some implications for theory and practice. First, much research has focussed on the exploration of the effects and behaviours of narcissists. The indirect effects of narcissism at a supervisor level and behavioural effect at the employee level are

(35)

28 relatively underexplored. The present study theoretically contributed to a deeper meaning of the effects of narcissism in a broader ‘organisational’ context, especially regarding the narcissism facet structure. For example, some recruitment policies of organisations might be aimed to screen out narcissists in an early stage. Those policies are tuned to a predominantly negative view of narcissism. This might be too short-sighted, if the results of the present study are incorporated. A high score on global narcissism, does not automatically mean bad results for the organisation. Organisations need to reconsider their conclusions based on a global NPI narcissism score and increasingly recognize the multidimensionality of the narcissism construct. Narcissism must be treated as a multidimensional concept in which the different dimensions could have different effects.

Second, with regard to the narcissism facet structure the presented findings are particularly relevant for science as initial steps towards a better understanding of the narcissism concept. In the present study L/A narcissism relates positively to perceived OCB, whereas E/E or maladaptive narcissism relates negatively to perceived OCB. These results are direct extensions to the proposed multidimensionality of the narcissism construct and are similar to the results presented in the study of Grijalva and Newman (2014), but in the context of OCB or pro-organisational behaviours. It is therefore recommended that further research reports both the score and effect size for global narcissism as the individual scores and effect sizes for adaptive and maladaptive narcissism.

Limitations

The present study was carried out with care but still has potential limitations. One of the major shortcomings of the present study was the use of a cross-sectional research design. Within a cross-sectional design, it is not possible to distinguish between cause and effect. For example, it is not possible to test whether narcissistic leadership causes employees to engage more or less pro-organisational behaviour. It was also not possible to rule out alternative

(36)

29 explanations for the observed effect, with the chosen study design. It is possible that narcissistic leaders give themselves the credit by assigning a higher score to employee OCB.

Another limitation of the present study concerns the characteristics and the method of sampling. The chosen convenience sampling method restricts the generalizability of the results. The results from the present study are applicable to this particular sample only and unfortunately not generalizable to the entire Dutch working population. Additionally, in the sample almost 35% of the employees worked part-time which could have affected the levels of perceived OCB (Bolino, Klotz, Turnley, & Harvey, 2013).

The present study was carried out in the context of a broader study on leadership and organisational behaviours. As a result, multiple concepts were measured using a single +100 items questionnaire which might reasonably discouraged the likelihood of participation. The duration of completing the online questionnaire was approximately between 9 and 32 minutes. There are indications that some participants rushed to complete the questionnaire, which is not always beneficial for the reliability of the results.

Directions for future research

As indicated previously, it may be promising to repeat the present study, with a few minor modifications. Firstly, OCB should be replaced as actual organisational citizenship behaviours and need to be measured also on the employee level, instead of perceived OCB only measured on a supervisor level. Secondly it might be promising to explore the effects of a less sociable extraversion and OCB measurement. A suggestion is an OCB scale which measures more assertive change oriented behaviours, like taking charge and for extraversion a scale including more explicitly a dominance facet and maybe even arrogance. Where possibly the dominance complementarity theory becomes an important influential factor in the relationship between narcissism and OCB.

(37)

30 The present study used the two factor solution proposed by Corry et al. (2008) to explore the multidimensionality effects of narcissism in organisations. But still more research is needed in order to establish a definitive consensus about the narcissism facet level structure. Initial narcissism facet model structures need to be further confirmed in multiple languages and cultures, for instance like the reliability confirmation of the NPI measure in a variety of different studies and languages (Barelds & Dijkstra, 2010).

Finally, it is recommended to explore the possibilities of isolating narcissism as a leadership style. Since previous research has shown that leadership styles, like transformational and transactional, have a clear influence on followers and organisational performances (Krasikova, Green, & LeBreton, 2013; Resick et al., 2009). It is therefore meaningful to control for the effects of those earlier mentioned leadership styles to be able to fully isolate the effects that narcissism has on followers and organisational performance. For example, future research could control for transformational, transactional and charismatic leadership style, while further exploring the narcissism-OCB relationship.

Conclusion

Not everyone can be narcissistic, so how is it possible that we say that we know so many? Is narcissism so highly prevalent or is it more complicated? In society, the word narcissism has acquired a severe negative connotation and confusion about its effects are thus easily made. Within organisations, there is nowadays still much misperception about the concept of narcissism. Why do so many employees think that their supervisor is narcissistic, or at least has some narcissistic traits? A possible explanation for this phenomenon is that many narcissism traits go hand in hand with typical leadership traits, such as for example extraversion and self-confidence. However, does that imply that every supervisor is narcissistic and is that a bad thing? The present study concludes that narcissism is a complex trait or style in which both positive and negative leadership traits are present: narcissism is

(38)

31 not necessarily dreadful; it depends on the individual dimensional narcissism scores. It is time that both adaptive and maladaptive narcissism are separately measured as the latter may be far more problematic. In summary, the present study hopes to serve as a starting point for further research to get a better grip on the effects of narcissistic leadership in organisations. Moreover, it draws attention to the fact that narcissism must be seen as a composition of two opposing dimensions in which valuable insights are made into the differentiated dimensional effects. Taken together, there is - and perhaps should be - a little bit of narcissism in all of us. Each person, supervisor or employee, needs to be somewhat narcissistic. Yet hopefully we can learn to lessen the maladaptive and extreme forms, while taking advantage of the adaptive element of narcissism.

(39)

32 References

Ackerman, R. A., Donnellan, M. B., & Robins, R. W. (2012). An Item Response Theory Analysis of the Narcissistic Personality Inventory. Journal of Personality Assessment,

94(2), 141–155.

Ackerman, R. A., Witt, E. A., Donnellan, M. B., Trzesniewski, K. H., Robins, R. W., & Kashy, D. A. (2011). What Does the Narcissistic Personality Inventory Really Measure? Assessment, 18(1), 67–87.

Aiken, L., & West, S. (1991). Multiple regression: Testing and Interpreting Interactions. Newbury Park, CA and London: Sage Publications.

APA (American Psychiatric Association). (2000). Diagnostic and Statistical Manual of

Mental Disorders (4th ed., text rev.). Washington, DC: Author.

Barelds, D. P. H., & Dijkstra, P. (2010). Narcissistic Personality Inventory: Structure of the adapted Dutch version. Scandinavian Journal of Psychology, 51(2), 132–138. doi:10.1111/j.1467-9450.2009.00737.x

Barry, C. T., Frick, P. J., & Killian, A. L. (2003). The Relation of Narcissism and Self-Esteem to Conduct Problems in Children: A Preliminary Investigation. Journal of

Clinical Child &amp; Adolescent Psychology, 32(1), 139–152.

doi:10.1207/S15374424JCCP3201_13

Bennett, R. J., & Robinson, S. L. (2000). Development of a measure of workplace deviance.

Journal of Applied Psychology, 85(3), 349.

Bolino, M. C., Klotz, A. C., Turnley, W. H., & Harvey, J. (2013). Exploring the dark side of organizational citizenship behavior. Journal of Organizational Behavior, 34(4), 542– 559. doi:10.1002/job.1847

Brunell, A. B., Gentry, W. A., Campbell, W. K., Hoffman, B. J., Kuhnert, K. W., & DeMarree, K. G. (2008). Leader Emergence: The Case of the Narcissistic Leader.

(40)

33 Personality and Social Psychology Bulletin, 34(12), 1663–1676.

doi:10.1177/0146167208324101

Campbell, W. K., & Campbell, S. M. (2009). On the Self-regulatory Dynamics Created by the Peculiar Benefits and Costs of Narcissism: A Contextual Reinforcement Model and Examination of Leadership. Self and Identity, 8(2-3), 214–232.

doi:10.1080/15298860802505129

Campbell, W. K., Hoffman, B. J., Campbell, S. M., & Marchisio, G. (2011). Narcissism in organizational contexts. Human Resource Management Review, 21(4), 268–284. Carpenter, N. C., Berry, C. M., & Houston, L. (2014). A meta-analytic comparison of

self-reported and other-self-reported organizational citizenship behavior: SELF-REPORTED AND OTHER-REPORTED OCB. Journal of Organizational Behavior, 35(4), 547– 574. doi:10.1002/job.1909

Chiaburu, D. S., Oh, I.-S., Berry, C. M., Li, N., & Gardner, R. G. (2011). The five-factor model of personality traits and organizational citizenship behaviors: A meta-analysis.

Journal of Applied Psychology, 96(6), 1140–1166. doi:10.1037/a0024004

Conway, J. M., & Lance, C. E. (2010). What Reviewers Should Expect from Authors Regarding Common Method Bias in Organizational Research. Journal of Business

and Psychology, 25(3), 325–334.

Corry, N., Merritt, R. D., Mrug, S., & Pamp, B. (2008). The Factor Structure of the Narcissistic Personality Inventory. Journal of Personality Assessment, 90(6), 593– 600. doi:10.1080/00223890802388590

De Hoogh, A. H. B., Den Hartog, D. N., & Nevicka, B. (2013). Gender Differences in the Perceived Effectiveness of Narcissistic Leaders. Applied Psychology, n/a–n/a.

Referenties

GERELATEERDE DOCUMENTEN

3.2.2 The Indefinite [Cl`N] and Why Cantonese and Mandarin Differ From the above data on indefinite noun phrases in Mandarin and Cantonese, it is clear that indefinite bare nouns

For claw-free graphs and chordal graphs, it is shown that the problem can be solved in polynomial time, and that shortest rerouting sequences have linear length.. For these classes,

military intervention in the Middle East in the search for terrorists (Chomsky 2003, 107). Even though both countries were subjected to U.S. domination, which should have

The focus of this research is to empirically test the effect of industry structure variables on the degree of small firm internationalization in the context of a transition

Assuming that the effect of surprise on consumption level and compensatory consumption through ethnocentric preference is due to a nonconscious threat response,

Most similarities between the RiHG and the three foreign tools can be found in the first and second moment of decision about the perpetrator and the violent incident

The classical version is one in which the narrative component is supposed to be largely dominant, sustained through periodic moments when the emphasis shifts towards

Table 6 Regression results of the moderation effects of the extraversion trait on the relationship between happiness (subjective well-being, happiness and life satisfaction) and