THE ROLE OF NARCISSISM IN ROMANTIC RELATIONSHIP INITIATION
By
Robert A. Ackerman

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Psychology
2011

ABSTRACT
THE ROLE OF NARCISSISM IN ROMANTIC RELATIONSHIP INITIATION
By
Robert A. Ackerman
Two studies were designed to clarify the roles of normal, grandiose, and vulnerable
narcissistic traits in the initiation of romantic relationships. Study 1 examined whether these
traits were differentially predictive of relationship initiation motives connected to selfenhancement (i.e., motives that concern increasing positive sentiments towards the self), selfprotection (i.e., motives that concern decreasing negative sentiments towards the self), and
intimacy. The question of what types of partners (i.e., admiring, desirable, and caring) are
considered attractive by people with differing levels of these narcissistic dimensions was also
investigated. Normal narcissism was connected to the diminished adoption of self-enhancement
and self-protection relationship initiation motives. The grandiose narcissistic traits, in contrast,
were generally related to the increased adoption of self-enhancement motives and an increased
attraction towards desirable romantic partners. Last, the vulnerable narcissistic traits were
connected to the increased adoption of both self-enhancement and self-protection relationship
initiation motives. Study 2 examined the roles of these traits in actual dyadic interactions using a
paradigm similar to speed-dating. Normal narcissism was associated with the expression of
dominant and confident behaviors. The grandiose narcissistic traits were generally linked to the
expression of dominance and gregariousness; moreover, participants with greater levels of the
grandiose traits were considered more attractive and desirable by their dates. Findings for the
vulnerable narcissistic traits were not as straightforward as there were few connections between
these traits and outcomes in the interpersonal interactions. Taken together, this dissertation

demonstrates that the different narcissism dimensions play unique and important roles in the
romantic relationship initiation process.

TO MY FAMILY

iv

ACKNOWLEDGEMENTS

First and foremost, I thank my advisors, Debby Kashy and Brent Donnellan. This
dissertation would not have been possible without them. I also thank my undergraduate research
assistants who assisted me in collecting these data—Jessie Trierweiler, Mollie Blanchard-Brown,
Alyssa Segal, Ashley Pratt, Jason Maise, Miles Armaly, Sarah Effner, Shannon Postel, and
Lauren Ruben. Finally, I thank my family.

v

TABLE OF CONTENTS

LIST OF TABLES

vii

LIST OF FIGURES

xi

CHAPTER 1: GENERAL INTRODUCTION

1

CHAPTER 2: INTRODUCTION TO STUDY 1

21

CHAPTER 3: METHOD FOR STUDY 1

23

CHAPTER 4: RESULTS FOR STUDY 1

34

CHAPTER 5: DISCUSSION FOR STUDY 1

54

CHAPTER 6: INTRODUCTION TO STUDY 2

62

CHAPTER 7: METHOD FOR STUDY 2

72

CHAPTER 8: RESULTS FOR STUDY 2

80

CHAPTER 9: DISCUSSION FOR STUDY 2

108

CHAPTER 10: GENERAL DISCUSSION

115

APPENDICIES

196

REFERENCES

221

vi

LIST OF TABLES

TABLE 1: Summary of Hypotheses for the Different Narcissism Dimensions
at Each Stage of Romantic Relationship Initiation

127

TABLE 2: Pattern Coefficients from an Exploratory Factor Analysis of the
Relationship Initiation Motives Inventory (RIM)

130

TABLE 3: Zero-order correlations between primary independent variables
in Study 1

133

TABLE 4: Zero-order correlations between RIM variables and
attachment-related avoidance in Study 1

134

TABLE 5: Zero-order correlations between primary independent variables and
dependent variables in Study 1

135

TABLE 6: Parameter Estimates from Path Analysis using the Narcissistic
Personality Inventory (NPI) Traits to Predict Relationship Initiation Motivations 136
TABLE 7: Parameter Estimates from an Unconstrained Multiple-Groups Analysis
for Men and Women using the Narcissistic Personality Inventory Subscales to
Predict Relationship Initiation Motivations
137
TABLE 8: Parameter Estimates from Path Analysis using the Pathological
Narcissism Inventory (PNI) Traits to Predict Relationship Initiation Motivations 138
TABLE 9: Parameter Estimates from an Unconstrained Multiple-Groups Analysis
for Men and Women using the Pathological Narcissism Inventory Scales to
Predict Relationship Initiation Motivations
139
TABLE 10: Means and Standard Deviations for Romantic Targets as a Function
of Target Presentation Order
140
TABLE 11: Zero-order correlations between primary dependent variables in
Study 1

141

TABLE 12: Zero-order correlations between primary independent variables and
dependent variables in Study 1

142

TABLE 13: Multiple Regression Analyses using the Narcissistic Personality
Inventory Subscales to Predict Attraction towards the Romantic Targets

143

TABLE 14: Multiple Regression Analyses using the Pathological Narcissism
vii

Inventory Scales to Predict Attraction towards the Romantic Targets

144

TABLE 15: Means and Standard Deviations of Attraction towards Romantic
Targets

145

TABLE 16: Predicting Attraction towards Admiring Target with Narcissistic
Personality Inventory Traits

146

TABLE 17: Predicting Attraction towards Desirable Target with Narcissistic
Personality Inventory Traits

148

TABLE 18: Simple Slopes for each Remote Associations Test (RAT)
Condition for the Relation between the Narcissistic Personality Inventory
Subscale of Entitlement/Exploitativeness and Attraction towards the Desirable
Target

150

TABLE 19: Predicting Attraction towards Caring Target with Narcissistic
Personality Inventory Traits

151

TABLE 20: Predicting Attraction towards Admiring Target with Pathological
Narcissism Inventory Traits

153

TABLE 21: Predicting Attraction towards Desirable Target with Pathological
Narcissism Inventory Traits

155

TABLE 22: Predicting Attraction towards Caring Target with Pathological
Narcissism Inventory Traits

157

TABLE 23: Simple Slopes for each Remote Associations Test (RAT)
Condition for the Relation between the Pathological Narcissism Inventory
Scale of Narcissistic Vulnerability and Attraction towards the Caring Target

159

TABLE 24: Illustration of Asymmetric Block Design

160

TABLE 25: Social Relations Model Variance Partitioning for Dating Record
Variables in Study 2

161

TABLE 26: Reciprocity Correlations for Dating Record Variables in Study 2

163

TABLE 27: Zero-order correlations between narcissism variables in Study 2

164

TABLE 28: Zero-order correlations between participants’ scores on the
Narcissistic Personality Inventory (NPI) subscales, the Pathological Narcissism
Inventory (PNI) scales, and their Social Relations Model (SRM) Target effects
for the Interpersonal Adjective Scale-Revised (IAS-R) Dating Record Variables

165

viii

TABLE 29: Predicting Dates’ Perceptions of Participants’ Interpersonal
Behavior from the Participants’ Narcissistic Personality Inventory (NPI) Traits

166

TABLE 30: Predicting Dates’ Perceptions of Participants’ Interpersonal
Behavior from the Participants’ Pathological Narcissism Inventory (PNI)
Traits

167

TABLE 31: Zero-order correlations between participants’ scores on the
Narcissistic Personality Inventory (NPI) subscales, the Pathological Narcissism
Inventory (PNI) scales, and their Social Relations Model (SRM) Target effects
for the Appeal Dating Record Variables

168

TABLE 32: Predicting Dates’ Perceptions of Participants’ Admiration and
Desirability, and Dates’ Attraction to Participants, from the Participants’
Narcissistic Personality Inventory (NPI) Traits

169

TABLE 33: Predicting Dates’ Perceptions of Participants’ Admiration and
Desirability, and Dates’ Attraction to Participants, from the Participants’
Pathological Narcissism Inventory (PNI) Traits

170

TABLE 34: Evaluating whether the relation between participants’ NPI traits
and their dates’ perceptions of them as Warm-Agreeable is Moderated by
participants’ perception of their dates’ desirability

171

TABLE 35: Evaluating whether the relation between participants’ NPI traits
and their dates’ perceptions of them as Admiring is Moderated by participants’
perception of their dates’ desirability

172

TABLE 36: Evaluating whether the relation between participants’ PNI traits
and their dates’ perceptions of them as Warm-Agreeable is Moderated by
participants’ perception of their dates’ desirability

173

TABLE 37: Evaluating whether the relation between participants’ PNI traits
and their dates’ perceptions of them as Admiring is Moderated by participants’
perception of their dates’ desirability

174

TABLE 38: Zero-order correlations between dates’ perceptions of the
participants’ gregarious-extraversion, assured-dominance, arrogance-calculation,
desirability, and attractiveness
175
TABLE 39: Zero-order correlations between participants’ scores on the
Narcissistic Personality Inventory (NPI) subscales, the Pathological Narcissism
Inventory (PNI) scales, and their Social Relations Model (SRM) Perceiver
effects for the Rapport Dating Record Variables
TABLE 40: Zero-order correlations between participants’ scores on the
ix

176

Narcissistic Personality Inventory (NPI) subscales, the Pathological Narcissism
Inventory (PNI) scales, and their Social Relations Model (SRM) Target effects
for the Rapport Dating Record Variables

177

TABLE 41: Predicting Participants’ Disclosure, Trust, and Connection to their
Dates in the Social Interactions from the Participants’ Narcissistic Personality
Inventory (NPI) Traits

178

TABLE 42: Predicting Dates’ Perceptions of Participants as Trusting and
Responsive in the Social Interactions from the Participants’ Narcissistic
Personality Inventory (NPI) Traits

179

TABLE 43: Predicting Participants’ Disclosure, Trust, and Connection to
their Dates in the Social Interactions from the Participants’ Pathological
Narcissism Inventory (PNI) Traits

180

TABLE 44: Predicting Dates’ Perceptions of Participants as Trusting and
Responsive in the Social Interactions from the Participants’ Pathological
Narcissism Inventory (PNI) Traits

181

TABLE 45: Predicting Participants’ Attraction towards Admiring Dates
with the Participants’ Narcissistic Personality Inventory (NPI) Traits

182

TABLE 46: Predicting Participants’ Attraction towards Admiring Dates
with the Participants’ Pathological Narcissism Inventory (PNI) Traits

183

TABLE 47: Predicting Participants’ Attraction towards Desirable Dates
with the Participants’ Narcissistic Personality Inventory (NPI) Traits

184

TABLE 48: Predicting Participants’ Attraction towards Desirable Dates
with the Participants’ Pathological Narcissism Inventory (PNI) Traits

185

TABLE 49: Predicting Participants’ Attraction towards Caring Dates
with the Participants’ Narcissistic Personality Inventory (NPI) Traits

186

TABLE 50: Predicting Participants’ Attraction towards Caring Dates
with the Participants’ Pathological Narcissism Inventory (PNI) Traits

187

x

LIST OF FIGURES

FIGURE 1: Higher order structure of Pathological Narcissism Inventory

188

FIGURE 2: Scree plot of eigenvalues from an exploratory factor analysis of
items from the relationship initiation motives inventory

189

FIGURE 3: Path analysis model for predicting the Relationship Initiation
Motives Inventory subscales from the Narcissistic Personality Inventory
subscales

190

FIGURE 4: Path analysis model for predicting the Relationship Initiation
Motives Inventory subscales from the Pathological Narcissism Inventory
subscales

191

FIGURE 5: The Interpersonal Circumplex

192

FIGURE 6: Diagram of a single-variable mediational model

193

FIGURE 7: Mediational Model for Grandiose Narcissism-Attraction Link

194

xi

CHAPTER 1: GENERAL INTRODUCTION
Narcissism corresponds to an investment in the self and is characterized by several
intrapersonal and interpersonal strategies aimed at self-esteem regulation (P. F. Kernberg, 1998).
Although usually discussed in contexts pertaining to pathology, theorists have noted that such
self-investment can occur for healthy individuals and that it can be connected to self-esteem
regulation strategies that are normal and adaptive (e.g., O. F. Kernberg, 1975; Stone, 1998).
Given that the social realm is one of the primary venues through which the self is regulated,
examination of the role of narcissism within close relationships represents an especially
important opportunity to understand the construct. It also offers an excellent opportunity to better
understand how basic relationship processes (e.g., relationship initiation, interdependence) are
impacted by individual differences involving an excessive self-focus.
Unfortunately, empirical work on the role of narcissism in romantic relationships has
1

been scarce. The research that has been conducted suggests that narcissists adopt a gameplaying approach to love (Campbell, Foster, & Finkel, 2002), are less committed in their
relationships (Campbell & Foster, 2002), and report an increased susceptibility to infidelity
(Buss & Shackelford, 1997). Although this research has helped to illuminate how narcissism
manifests itself in ongoing close relationships, we still know very little about how narcissists
initiate such relationships in the first place (for a notable exception, see Campbell, 1999). Such
research is important because before we can fully understand the processes that occur in ongoing
relationships, we must first gain insight into the motivations and behaviors that initially gave rise
to them. Moreover, the field’s overreliance on a single measure to assess the construct (i.e., the
Narcissistic Personality Inventory) has led to disparate research findings (e.g., narcissism being
1

When using the term “narcissist,” I am referring to someone with elevated scores on particular
continuous dimensions of personality rather than implying that narcissism is typological.
1

linked to both self-esteem and aggressive behavior) and the formulation of theoretical models
(e.g., the Agency Model of Narcissism; see Campbell, Brunell, & Finkel, 2006) that conflate
adaptive and maladaptive features of personality.
Accordingly, the present research will make use of the distinction between normal and
pathological narcissism (Pincus & Lukowitsky, 2010). It will also distinguish between the
pathological narcissistic expressions of grandiosity and vulnerability. As will be argued, each of
these traits should be associated with somewhat different strategies to regulate self-esteem.
Moreover, such strategies should be related to various aspects of the relationship initiation
process, such as the motivations that individuals possess for pursuing relationships, what
individuals find attractive, how the self is presented in initial encounters, and what individuals do
to ensure the possibility of a second encounter. The goal of this dissertation is therefore to
uncover the different strategies associated with these traits and their resulting consequences for
the initiation stages of a romantic relationship.
The Conceptualization and Measurement of Narcissism
The construct of narcissism has a long history in psychology and has acquired myriad
meanings over the years (see, e.g., Ellis, 1898; Freud, 1914). Nevertheless, most contemporary
conceptualizations of narcissism are indebted to the theoretical work of O. F. Kernberg, Kohut,
and Millon. According to these theorists, narcissists possess abnormal self-structures and exhibit
pathological self-esteem regulation strategies (O. F. Kernberg, 1998; Millon, 1996). In line with
this conceptualization, it has been posited that narcissists are grandiose and self-absorbed (O. F.
Kernberg, 1975; Kohut, 1966; Millon, 1996), exhibitionistic (O. F. Kernberg, 1975; Kohut,
1966; Millon, 1996), entitled (O. F. Kernberg, 1975; Millon, 1996), exploitative (O. F. Kernberg,
1975; Millon, 1996), incapable of experiencing empathy (O. F. Kernberg, 1975; Millon, 1996),

2

and that they fluctuate between feelings of superiority and inferiority (O. F. Kernberg, 1975;
Kohut, 1966). Though it has been historically conceptualized as a disorder and discrete category,
many contemporary researchers in both social/personality (e.g., Raskin & Terry, 1988) and
clinical psychology (e.g., Pincus, Ansell, Pimentel, Cain, Wright, & Levy, 2009) now
conceptualize narcissism as a continuous trait.
The Diagnostic and Statistical Manual (DSM) of psychiatric disorders first included
Narcissistic Personality Disorder (NPD) as a diagnosable entity in its third edition. Criteria for
the DSM-III primarily reflected the insights of O. F. Kernberg, Kohut, and Millon (Levy,
Reynoso, Wasserman, & Clarkin, 2009). The Narcissistic Personality Inventory (NPI; Raskin &
Hall, 1979)—the most frequently used measure of narcissism within social and personality
psychology (see Cain, Pincus, & Ansell, 2008)—was based on the anticipated criteria for NPD in
the DSM-III. As such, it placed a heavy emphasis on those features of narcissism that involved
grandiosity, exhibitionism, entitlement, and exploitativeness.
Although the content for the NPI was purportedly derived from these anticipated DSMIII criteria, conceptualizations of narcissism in social/personality psychology are potentially
different from those found in the neighboring disciplines of psychiatry and clinical psychology
given that social/personality psychologists have generally emphasized features that reflect more
resilience and extraversion (Miller & Campbell, 2008). It is therefore notable that Pincus and his
colleagues (e.g., Pincus & Lukowitsky, 2010) have suggested that a distinction be made between
normal narcissism and pathological narcissism.
Pincus and Lukowitsky (2010) assert that whereas normal narcissism involves selfregulatory strategies that are generally adaptive and stem from a healthy personality
organization, pathological narcissism is characterized by maladaptive self-regulatory strategies

3

that generally produce distress and impairment (Pincus & Lukowitksy, 2010). Accordingly, they
proposed that normal narcissism is manifested in behaviors such as asserting dominance, holding
positive self-illusions, and engaging in achievement striving (Pincus & Lukowitsky, 2010). In
contrast, pathological narcissism is manifested in vulnerable self-states and characteristics that
more closely reflect the criteria identified in the DSM for NPD.
After an exhaustive review of the literature, Cain et al. (2008) further identified two
primary ways in which pathological narcissism is expressed: Grandiosity and Vulnerability.
Narcissistic Grandiosity is characterized by several intrapersonal (e.g., engagement in grandiose
fantasies, feelings of superiority, entitled beliefs) and interpersonal (e.g., exploitive and
exhibitionistic behavior) strategies aimed at enhancing the self. Pincus et al. (2009) developed
several subscales to capture this expression of narcissistic dysfunction (see Figure 1). Among
these were Grandiose Fantasy (i.e., a tendency to fantasize about unlimited admiration and
success), Self-Sacrificing Self-Enhancement (i.e., a tendency to engage in ostensibly selfless acts
to reinforce the grandiose self), and Exploitativeness (i.e., a tendency to engage in interpersonal
manipulation). Pincus et al. (2009) found that the Exploitativeness facet was related to more
domineering interpersonal problems whereas the Grandiose Fantasy and Self-Sacrificing SelfEnhancement facets were associated with more intrusive and overly-nurturant interpersonal
problems, respectively.
Narcissistic Vulnerability reflects the possession of a fragile self-concept and the frequent
experience of emotional dysregulation. In their research, Pincus et al. (2009) developed several
subscales to assess the features of Narcissistic Vulnerability (see Figure 1). These included
Contingent Self-Esteem (i.e., the tendency to experience self-esteem as substantially labile and
dependent upon external sources), Hiding the Self (i.e., a reluctance to portray the self as

4

imperfect and incompletely self-sufficient), Devaluing (i.e., the inclination to devalue those
individuals who do not reinforce the grandiose self as well as the inclination to experience
disappointment/shame for needing this reinforcement), and Entitlement-Rage (i.e., a tendency to
experience anger over unmet expectations of special favors or treatment).
In contrast to the scales assessing Narcissistic Grandiosity, the Narcissistic Vulnerability
subscales were consistently associated with lower self-esteem scores (Pincus et al., 2009).
Moreover, they displayed the strongest positive connections with shame, primitive defenses,
impaired reality testing, and low moral values (Pincus et al., 2009). The vulnerability subscales
of Devaluing and Entitlement-Rage were most strongly positively related to aggression (Pincus
et al., 2009). When examining their relations to interpersonal problems, Pincus et al. (2009)
found that whereas the Devaluing and Entitlement-Rage subscales were connected to vindictive
problems, the Hiding the Self and Contingent Self-Esteem subscales were related to avoidant and
exploitable problems, respectively.
According to Pincus and Lukowitsky (2010), social/personality psychologists have
primarily studied normal narcissism. This focus is most likely due to social/personality
psychologists’ focus on the experiences of nonclinical populations (especially college student
samples) as well as their almost exclusive reliance on the Narcissistic Personality Inventory
(NPI; Raskin & Hall, 1979) to operationalize narcissism in their research—a measure that Pincus
and Lukowitsky (2010) believe assesses normal narcissism. Nevertheless, there is still some
content on the inventory that assesses some of the maladaptive features traditionally associated
with narcissism (e.g., Ackerman, Witt, Donnellan, Trzesniewski, Robins, & Kashy, 2011).
Indeed, researchers have argued that the NPI seems to assess both adaptive and maladaptive
features (e.g., Barry, Frick, & Killian, 2003; Cain et al., 2008; Watson & Morris, 1991), and

5

recent analyses exploring the dimensional structure of the NPI have corroborated this notion
(Ackerman et al., 2011; Corry, Merritt, Mrug, & Pamp, 2008).
Most recently, Ackerman et al. (2011) found that the NPI consists of three related but
distinct dimensions: Leadership/Authority, Grandiose Exhibitionism, and
Entitlement/Exploitativeness. Leadership/Authority captures content related to assertiveness,
dominance, and self-perceived leadership ability. It is therefore most similar in form to Pincus
and Lukowitsky’s (2010) conceptualization of normal narcissism (Ackerman et al., 2011).
Individuals with higher levels of this construct report higher levels of extraversion, self-esteem,
Behavioral Activation System (BAS) activation, and self-control (Ackerman et al., 2011; Corry
et al., 2008; Emmons, 1984; Rhodewalt & Morf, 1995; Watson & Biderman, 1993). These
individuals also report higher levels of self-monitoring, surgency, warmth, and dominance
(Emmons, 1984), as well as an increased tendency to attribute the cause of positive events to an
internal, global, and stable source (Rhodewalt & Morf, 1995). Combined with the fact that higher
levels of this construct are associated with lower levels of neuroticism, depression, anxiety,
Behavioral Inhibition System (BIS) activation, and contingent self-esteem (Ackerman et al.,
2011; Emmons, 1984; Rhodewalt & Morf, 1995; Watson & Biderman, 1993), the
Leadership/Authority trait is generally considered a rather adaptive feature of personality.
Grandiose Exhibitionism, on the other hand, represents a somewhat more maladaptive
feature of the NPI than Leadership/Authority. Conceptually speaking, this trait appears to capture
the expression of narcissistic grandiosity. The Grandiose Exhibitionism subscale is characterized
by a combination of thoughts and behaviors related to vanity, exhibitionism, and beliefs of
superiority. People with higher levels of this trait report greater levels of extraversion and BAS

6

activity (Ackerman et al., 2011). Moreover, they express anger over not getting their way and are
somewhat oblivious to the feelings and needs of others (Ackerman et al., 2011).
Entitlement/Exploitativeness represents the most maladaptive feature of the NPI among
the three dimensions highlighted by Ackerman et al. (2011). Entitlement/Exploitativeness refers
to thoughts and behaviors that reflect beliefs of special privilege and a willingness to use others
for personal gain. In terms of Pincus and Lukowitsky’s (2010) distinction between grandiosity
and vulnerability, Entitlement/Exploitativeness appears to better represent an expression of
narcissistic vulnerability. However, unlike the PNI subscales used to assess narcissistic
vulnerability, this NPI subscale is narrower in scope and seems to reflect a set of strategies used
to cope with the experience of vulnerable self- and affect-states. Of all the facets of personality
embedded within the NPI, Entitlement/Exploitativeness has been shown to have the most
pernicious outcomes. Research has shown that this facet is related to higher levels of
interpersonal manipulation, antisocial tendencies, devaluation of others, neuroticism,
disagreeableness, entitlement-rage, and contingent self-esteem (Ackerman et al., 2011).
Entitlement/Exploitativeness has also been linked to lower levels of relationship satisfaction and
college adjustment among roommates (Ackerman et al., 2011).
Despite their apparent differences, each of the aforementioned narcissistic dimensions is
fundamentally concerned with self-esteem regulation. What appear to differentiate these
narcissistic dimensions are the particular strategies that are used to regulate self-esteem. Hepper,
Gramzow, and Sedikides (2010) recently found that individuals with greater levels of trait
narcissism (operationalized using a short-form of the NPI) are likely to use both selfenhancement (i.e., positivity embracement, favorable construals, and self-affirming reflections)
and self-protection (i.e., defensiveness) strategies. Given the divergent patterns of empirical

7

associations found for the narcissistic expressions of grandiosity and vulnerability, it is probable
that each particular expression of narcissistic dysfunction may be connected to different selfesteem regulation strategies.
Wright, Lukowitsky, Pincus, and Conroy (2010) recently remarked that PNI Narcissistic
Grandiosity appears to reflect maladaptive forms of self-enhancement. Similarly, Tritt, Ryder,
Ring, and Pincus (2010) commented that PNI Narcissistic Grandiosity is characterized by
approach motivation. It is plausible that the strategies used by individuals with greater levels of
NPI Grandiose Exhibitionism also primarily reflect self-enhancement. Indeed, each of the
intrapersonal (e.g., grandiose self-views and preoccupation with physical appearance) and
interpersonal (e.g., attention-seeking behaviors) strategies associated with this trait appear to be
in the service of affirming a grandiose self. Moreover, Grandiose Exhibitionism has been found
to be connected with increased BAS activity (Ackerman et al., 2011).
Unlike these NPI and PNI operationalizations of narcissistic grandiosity,
Entitlement/Exploitativeness appears to reflect a collection of strategies that have as their goal
self-protection. Indeed, individuals with higher levels of this narcissistic trait have been found to
possess self-esteem that is somewhat depleted and more labile and contingent upon external
sources (Ackerman et al., 2011). It therefore appears that both entitled beliefs and exploitive
behavior seem to be enacted in the service of trying to protect an inferior self (for similar
arguments, see Bishop & Lane, 2002; Zitek, Jordan, Monin, & Leach, 2010). PNI Vulnerability
has also been argued to be characterized by avoidance motivation (Tritt et al., 2010). Combined
with the emphasis placed on a fragile self-concept and the inclusion of content that appears to
serve the goal of self-protection (e.g., Hiding the Self), I expect PNI Narcissistic Vulnerability to
function in a similar manner as Entitlement/Exploitativeness.

8

To sum up, past work has suggested a distinction between normal and pathological
narcissism as well as a distinction between the pathological narcissistic traits of grandiosity and
vulnerability. Moreover, the Narcissistic Personality Inventory and the Pathological Narcissism
Inventory appear to assess related but distinct expressions of some of these traits. The present
research uses the Narcissistic Personality Inventory to assess normal narcissism (i.e.,
Leadership/Authority) and relatively normative expressions of narcissistic grandiosity (i.e.,
Grandiose Exhibitionism) and vulnerability (i.e., Entitlement/Exploitativeness). It also uses the
recently developed Pathological Narcissism Inventory (PNI; Pincus et al., 2009) to assess more
clinically relevant expressions of narcissistic grandiosity and vulnerability (i.e., Narcissistic
Grandiosity and Narcissistic Vulnerability, respectively). To be clear, these five dimensions are
not meant to exhaustively cover the definition of narcissism. Nonetheless, I believe that they
capture broad phenotypic expressions of the construct that can provide a good foundation for
better understanding the relation between narcissism and relationships. Thus, I will focus on
these five dimensions when theorizing about connections between narcissism and romantic
relationship initiation.
The Role of Narcissism in Romantic Relationship Initiation
Bredow, Cate, and Huston (2008) propose a conceptual model of romantic relationship
initiation that involves four stages. In the first stage, Appraisal of Initial Attraction, individuals
discern whether they are attracted to a potential partner. Bredow et al. (2008) emphasize that
individuals’ motives for pursuing the relationship are particularly important in this stage. If there
is attraction in the first stage, individuals then decide whether to make an overture (i.e., the
second stage). Such a decision depends upon how confident the individual is that the potential
partner will accept her or his overture. Bredow et al. (2008) state that such confidence hinges

9

upon two factors: (1) whether the potential partner is perceived to be open to the individual’s
motives, and (2) whether the potential partner is perceived to be attracted to the individual. If
only one of these factors is true, then the individual will have moderate confidence; however, if
both of them are true, the individual is argued to have high confidence.
The third stage in Bredow et al.’s (2008) model is Strategic Self-Presentation. Whereas
moderate confidence from the second stage leads to strategic self-presentation (i.e., selective
presentation of particular aspects of the self), high confidence leads to self-expression (i.e.,
presentation of the true self). The final stage in Bredow et al.’s (2008) model is Build-Up of
Rapport. Bredow et al. (2008) suggest that a build-up of rapport will lead to a successful
encounter whereas the lack of such a build-up will lead to an encounter that is terminated. The
present research examines the role of narcissism within three of these stages in detail.
Specifically, it investigates the roles of normal and pathological narcissistic traits in the
Appraisal of Initial Attraction, Strategic Self-Presentation, and the Build-Up of Rapport.
The Appraisal of Initial Attraction
When two people meet for the first time, they each carry with them particular motivations
(e.g., to fall in love). These motivations in turn have an impact on what characteristics they will
find attractive in a potential partner (Bredow et al., 2008). To the extent that there is a match
between the individual’s motives for pursuing a relationship and the potential partner’s
characteristics, then there will be romantic attraction (Bredow et al., 2008). Therefore, a
satisfactory understanding of the role of narcissism in romantic relationship initiation will first
require an exploration of the reasons that narcissists have for entering such relationships.
Campbell et al. (2006) proposed that narcissists possess agentic goals (e.g., the goal to
gain status and self-esteem) that drive them to pursue romantic relationships. According to

10

Campbell et al. (2006), potential romantic partners that are perceived to be ideal and provide
admiration will be desired by narcissists because they raise self-esteem. Moreover, potential
romantic partners that are of high status (e.g., through being very physically attractive or wellknown) will be desired because they help to enhance status. Such proposals, however, are based
on a conceptualization of narcissism that predominantly reflects self-enhancing rather than
defensive tendencies. This conceptualization seems to derive from other research showing that
the NPI is positively correlated with extraversion and approach-related motivations (e.g.,
Emmons, 1984; Foster & Trimm IV, 2008; Sedikides & Luke, 2008).
Although previous research corroborates the notion that narcissists’ goals tend to be selffocused and individualistic (Emmons, 1984; Watson & Biderman, 1993), it does not appear that
all narcissists are excessively extraverted or fully driven by appetitive motivations (see
Ackerman et al., 2011). Indeed, the entitled and exploitative features of narcissism tend to be
either unrelated, or even negatively related, to constructs such as Extraversion and Behavioral
Activation System (BAS) activity (see Ackerman et al., 2011). These findings lend support to the
idea that the grandiose and vulnerable expressions of narcissism are differentially related to
processes related to self-enhancement and self-protection. Such differences should likely play a
role in the adoption of distinct relationship motives.
Critical to this point, Cooper, Shapiro, and Powers (1998) surmised that two predominant
distinctions seem to underlie most theories of motivation: approach/avoidance motivation (e.g.,
Carver & White, 1994; Gray, 1987) and internal/external focus (e.g., Bakan, 1966; Bowlby,
1988). By applying these two pairs of dimensions to the realm of sexual behavior, Cooper et al.
(2008) derived four general motivational categories that include Enhancement (i.e., self-focused
motivations that entail enhancing pleasure), Intimacy (i.e., other-focused motivations that entail

11

increasing intimacy), Coping (i.e., self-focused motivations that concern the regulation of
negative affect and self-esteem threats), and Approval (i.e., other-directed motives that involve
the avoidance of disapproval from others). Although used by Cooper et al. (1998) to explain the
motives behind sexual behavior, this model may be especially useful to better understand the
motivations behind narcissists’ pursuit of romantic relationships.
Thus, I expect that those narcissistic traits reflective of self-enhancement (i.e., Grandiose
Exhibitionism and Narcissistic Grandiosity) will be positively associated with Enhancement
motives for pursuing romantic relationships. In contrast, I expect those narcissistic traits
reflective of self-protection (i.e., Entitlement/Exploitativeness and Narcissistic Vulnerability) to
be positively associated with Coping motives for pursuing romantic relationships. Finally, given
the self-focused nature of narcissism, I expect that all five of the narcissistic traits measured in
this research will be negatively related to Intimacy relationship motives.
These narcissistic traits should also have an impact on what partner characteristics
individuals find attractive. Campbell (1999) initially proposed the Self-Orientation Model to
account for the role of narcissism in romantic attraction. The model stipulates that narcissists’
level of attraction to a partner depends upon the partner’s perceived ability to provide the
narcissist with self-enhancement. Whereas some partners (i.e., individuals who are admiring
and/or possess highly desirable characteristics) possess characteristics that represent
opportunities for self-enhancement and are perceived as attractive, other partners (i.e.,
individuals who are interested in emotional intimacy or possess expectations of caring) do not
represent opportunities for self-enhancement and are therefore perceived as less attractive.
Campbell (1999) proposed admiration (i.e., the potential romantic partner’s provision of
praise and other positive feedback about the self) as a possible self-enhancement strategy for

12

narcissists during romantic attraction. Given that individuals with higher levels of grandiose
narcissistic traits are expected to adopt enhancement relationship motives and to seek out such
admiration in an approach-oriented manner, I hypothesize that such individuals will be especially
attracted to targets who readily provide such attributes (i.e., targets who are perceived to be
admiring and interested). Further evidence for this hypothesis comes from Campbell’s (1999)
second study in which he found that the strongest correlations between the NPI and romantic
attraction to the admiring target occurred for the Superiority and Vanity subscales. Due to their
preoccupation with obtaining positive feedback about the self and enhancing their self-image,
however, I hypothesize that individuals with higher levels of grandiose traits will also
demonstrate an aversion to caring targets.
Campbell (1999) further proposed identification as another possible self-enhancement
strategy for narcissists during romantic attraction. Identification refers to the association of the
self with an ideal and high status other (i.e., a perfect other). Interestingly, O. F. Kernberg (1975)
discussed how narcissists sometimes seek an ideal other for the purposes of protection. It is
possible that individuals possessing higher levels of narcissistic vulnerability will be more
attracted to those individuals who display highly desirable characteristics. Indeed, such
individuals may be perceived as sources of protection for an acknowledged vulnerable self. In
support of this conjecture, Campbell (1999) found that the NPI Entitlement scale was strongly
correlated with ratings of attraction towards a perfect and uncaring target. I therefore predict that
individuals with higher levels of vulnerable narcissistic traits (i.e., Entitlement/Exploitativeness
and Narcissistic Vulnerability) will be more romantically attracted to perfect targets (called
“desirable” targets in the present research) and less romantically attracted to caring targets.

13

Taken together, several hypotheses concerning the role of narcissistic traits in initial
attraction can be proposed. First, I predict that all five narcissistic dimensions (i.e.,
Leadership/Authority, Grandiose Exhibitionism, Entitlement/Exploitativeness, Narcissistic
Grandiosity, and Narcissistic Vulnerability) will be associated with reduced intimacy
motivations for entering romantic relationships and a decreased attraction towards potential
romantic targets who are caring. Second, given the consistent associations found between
Leadership/Authority and positive psychological states (Ackerman et al., 2011), I predict that
persons with higher levels of this trait will report lower levels of coping motivations. Third, I
predict that individuals with higher levels of Grandiose Exhibitionism and Narcissistic
Grandiosity will possess greater enhancement motivations for entering relationships and that
they will report increased attraction towards potential romantic targets who admire them. Last, I
predict that individuals with higher levels of Entitlement/Exploitativeness and Narcissistic
Vulnerability will possess greater levels of coping motivations and that they will report increased
attraction towards potential romantic targets who are desirable.
Strategic Self-Presentation
When individuals are not fully confident that the other person will be receptive to their
advances or is attracted to them, Bredow et al. (2008) propose that individuals engage in
strategic self-presentation in an effort to make the self more attractive to others. According to
Bredow et al. (2008), individuals can use three strategies to pique the interest of the other person.
These include appearing likable, competent/capable, or morally virtuous. In order to appear
likeable, people may ingratiate the other person and/or draw attention to their own positive
qualities. To appear competent/capable, people may emphasize those personal qualities that

14

convey talent, resources, and accomplishment. Last, to appear morally virtuous, people may
draw the other person’s attention to those personal qualities that convey honesty and integrity.
Previous research examining the impressions that narcissists make upon unacquainted
individuals has revealed some interesting findings. Oltmanns, Friedman, Fiedler, and Turkheimer
(2004), for instance, examined what traits can be gleaned from 30-second thin slices of behavior
based on videos of military recruits. The military recruits that were nominated by their peers to
be higher in narcissism were rated by unacquainted individuals to be higher in extraversion, to be
more physically attractive, and to be worthy of getting to know better (see also Friedman,
Oltmanns, Gleason, & Turkheimer, 2006). Friedman, Oltmanns, and Turkheimer (2007) further
found that individuals nominated by their peers to be narcissistic (as well as those with higher
scores on the Schedule for Nonadaptive and Adaptive Personality narcissistic subscale) were
perceived by unacquainted coders to take advantage of others, believe they are superior, and
prefer being the center of attention. For the most part, the narcissistic traits connected with
grandiosity therefore appear to be most salient upon initial acquaintance.
Other research has more directly examined how different facets of narcissism are related
to initial personality impressions (e.g., Back, Schmukle, & Egloff, 2010). Back et al. (2010)
showed that individuals’ NPI total scores were positively linked to peers’ ratings of their
popularity in a group setting. Moreover, this connection was primarily driven by the
Exploitativeness/Entitlement dimension of the NPI identified by Emmons (1984, 1987). Using a
lens model approach (see Brunswick, 1947), Back et al. (2010) demonstrated that individuals
with higher levels of Exploitativeness/Entitlement exhibited more humorous verbal expressions,
were flashier and neater in their style of dress, conveyed nonverbal behavior that was more self-

15

assured, and gave more charming facial expressions. These researchers were able to further show
that these behavioral cues were what led to higher popularity ratings.
However, Back et al. (2010) used the Emmons (1984, 1987) dimensional solution for the
NPI, and so their conclusions may need to be qualified because this specification of the
Exploitative/Entitlement dimension actually contains some items representing grandiosity and
exhibitionism (e.g., “I am more capable than other people” and “I get upset when people don't
notice how I look when I go out in public”). Therefore, what may actually be leading to stylish
dress and likeable expressions may be the attention-seeking tendencies associated with
exhibitionism rather than entitled beliefs and exploitative tendencies per se. This would be
consistent with the correlations found between the NPI facet of Self-Admiration/Self-Absorption
and those behavioral cues involving flashy and neat dress, charming facial expression, and selfassured body movements (Back et al., 2010).
In light of the above findings, I hypothesize that the expression of narcissistic grandiosity
(i.e., Grandiose Exhibitionism and Narcissistic Grandiosity) will be associated with selfpresentational strategies that involve competence/capability. Given their need for admiration,
people with greater levels of grandiosity will be likely to boast about their accomplishments and
draw attention to their highly desirable attributes. As such, they will likely take great pains to
appear physically attractive and to convey a sense of self-assurance throughout the interaction.
Moreover, given their exhibitionistic nature, such people will likely come across as gregarious.
Vulnerable individuals, on the other hand, are likely to use a different strategy. As
discussed previously, vulnerable individuals are expected to be motivated to pursue romantic
relationships to cope with negative emotions and to bolster a relatively inadequate self-image. As
such, they are expected to be particularly attracted to those individuals who possess desirable

16

qualities. Rather than utilizing a strategy that involves presenting the self as competent and/or
capable, it is more likely that these individuals will use ingratiation as a strategy to win over the
“perfect” potential partner. Combined with their relatively introverted nature, I therefore predict
that individuals with higher levels of narcissistic vulnerability will engage in warm, admiring,
and submissive behaviors towards those whom they perceive to be desirable.
In summary, I first predict that individuals with higher levels of grandiose traits (i.e.,
Grandiose Exhibitionism and Narcissistic Grandiosity) will use self-presentational strategies that
involve the demonstration of competence and/or capability. As a result of this, I expect such
individuals to be perceived as self-assured, extraverted, charming, and good looking during an
initial encounter. Moreover, given the nature of the strategy being used, I also expect individuals
with greater levels of grandiose traits to be perceived as arrogant. Second, I predict that
individuals with higher levels of vulnerable traits (i.e., Entitlement/Exploitativeness and
Narcissistic Vulnerability) will use self-presentational strategies that involve ingratiation when
they are interacting with someone they perceive to be desirable. Consequently, I expect
individuals with higher levels of these traits to be perceived as friendly, submissive, and
admiring by those individuals they consider to be desirable.
Building Rapport and Setting the Stage for a Second Encounter
The establishment of rapport is generally considered important for the initial
development of a romantic relationship (Bredow et al., 2008). Therefore, it is advantageous for
individuals to be able to initiate a conversation with the other person and work to foster mutual
feelings of affinity. As was discussed previously, strategic self-presentation is one method that
can be used to generate feelings of attraction. Bredow et al. (2008) also assert that affinity can be

17

effectively engendered within an interaction by showing interest, being attentive, finding
common areas of interest, and engaging in self-disclosure.
Self-presentational strategies that involve charm, confidence, and expressiveness have
been found to result in higher ratings of likeability (Back et al., 2010). Likewise, being wellgroomed and physically attractive also can lead to higher ratings of interpersonal attractiveness
(Back et al., 2010; Vazire, Naumann, Rentfrow, & Gosling, 2008). Based on these findings, it is
likely that individuals with higher levels of grandiose traits (i.e., Grandiose Exhibitionism and
Narcissistic Grandiosity) will elicit higher ratings of romantic attraction from their interaction
partners and that this relation will be mediated by their interaction partners’ perceptions of them
as being more gregarious and desirable (i.e., charming, popular, and good looking).
In the previous section, I predicted that individuals with higher levels of narcissistic
vulnerability will engage in warm, admiring, and submissive behaviors towards those whom they
perceive to be desirable. Previous research has found ingratiation to be a rather effective strategy
to foster attraction so long as it is not too extreme (Vonk, 2002). That being said, I expect that
individuals’ levels of narcissistic vulnerability will be related to their interaction partners’ ratings
of romantic attraction. Importantly, however, this will only be the case when vulnerable
individuals perceive their partners to be desirable (i.e., it is not expected that they will ingratiate
other interaction partners that they do not perceive to be desirable). Therefore, I predict that the
relation between individuals’ levels of narcissistic vulnerability and their interaction partners’
ratings of romantic attraction will be moderated by the individuals’ perceptions of their
interaction partners being desirable. Moreover, in the cases where the individual does perceive
the partner to be desirable, the relation between the individuals’ degree of vulnerable narcissism

18

and their partners’ ratings of romantic attraction will be mediated by the interaction partners’
perceptions of the individual as being submissive, warm, and admiring.
As mentioned previously, expressing interest in the other person and engaging in selfdisclosure also work to foster attraction. Indeed, self-disclosure is connected to enhanced
perceptions of intimacy (Laurenceau, Barrett, & Pietromonaco, 1998) and the establishment of
such a connection is likely to lead to increased attraction. However, past research suggests that
individuals with narcissistic traits are unlikely to desire intimacy (Emmons, 1989). Indeed,
Campbell (1999) argued that intimacy represents a detriment to self-enhancement as any flaws
associated with the narcissist will be exposed. Thus, I predict that the grandiose and vulnerable
narcissistic traits will be linked to lower rates of self-disclosure. Moreover, given their excessive
self-preoccupation, individuals with higher levels of the grandiose and vulnerable narcissistic
traits will also be perceived by their partners to be less responsive. Consequently, although the
self-presentational strategies associated with the narcissistic traits may help to foster initial
attraction, it is unlikely that such encounters will be characterized by feelings of closeness.
In summary, several predictions can be made concerning the role of narcissistic traits in
the establishment of rapport during initial encounters. First, individuals with higher levels of
grandiose traits (i.e., Grandiose Exhibitionism and Narcissistic Grandiosity) will be rated as
more attractive because they are perceived to be more gregarious and desirable. Second, when
individuals with higher levels of vulnerable traits (i.e., Entitlement/Exploitativeness and
Narcissistic Vulnerability) perceive potential partners to be desirable, these partners will rate
them as more attractive because they perceive the individuals with vulnerable traits to be
submissive, warm, and admiring. Third, the measures of grandiose and vulnerable narcissism
will be negatively related to self-reported rates of disclosure during initial encounters. Fourth,

19

individuals with greater levels of grandiose and vulnerable narcissism will be perceived by their
partners to be less responsive. Fifth, the measures of grandiose and vulnerable narcissism will be
negatively related to feelings of closeness.
Present Research
The present research was designed to clarify the roles played by normal narcissism,
narcissistic grandiosity, and narcissistic vulnerability in the romantic relationship initiation
process. Study 1 examines whether these traits are differentially associated with separate motives
for entering relationships. It also evaluates whether individuals’ attraction towards potential
romantic partners with particular characteristics represents an actual means for grandiose and
vulnerable individuals to regulate self-esteem. Study 2 uses a paradigm similar to speed-dating to
further examine the role of these narcissistic traits in the attraction process in real-life social
settings, as well as to investigate whether the grandiose and vulnerable expressions of narcissism
are differentially related to strategies involved in self-presentation and the building of rapport.
Table 1 presents a summary of the hypotheses for each of the narcissistic attributes at the
different stages of romantic relationship initiation.

20

CHAPTER 2: INTRODUCTION TO STUDY 1
Study 1 aimed to illuminate the roles of normal, grandiose, and vulnerable narcissistic
traits in the appraisal of initial attraction. As was discussed previously, it is plausible that each of
the narcissistic traits may be associated with different motives for relationship initiation. Those
narcissistic traits indicative of inflated self-views and approach-related behavioral strategies (i.e.,
Grandiose Exhibitionism and Narcissistic Grandiosity), for instance, will likely be associated
with relationship motives that reflect self-enhancement. In contrast, those narcissistic traits
reflective of depleted self-views and self-protective behavioral strategies (i.e.,
Entitlement/Exploitativeness and Narcissistic Vulnerability) will likely be connected to
relationship motives that represent an escape from or amelioration of negative self-states.
In addition to demonstrating that each of these narcissistic traits is associated with
somewhat different motives for initiating a romantic relationship, Study 1 investigates whether
each of these traits is also associated with attraction towards different types of romantic targets. I
hypothesize that those traits indicative of grandiosity will be positively associated with attraction
towards romantic targets who provide admiration. In contrast, those traits indicative of
vulnerability should be positively associated with attraction towards romantic targets who are
considered desirable. Last, I predict that each of the measured narcissist traits will be negatively
associated with attraction towards romantic targets who are considered caring.
Although previous research has assumed that narcissists use their romantic relationships
as a means to regulate self-esteem (see Campbell, 1999), this assumption has never been
empirically tested. The present research therefore experimentally manipulates participants’ selfesteem levels to evaluate whether the relations between narcissistic traits and attraction towards
particular romantic targets become stronger. The presence of an interaction between the self-

21

esteem manipulation and the relations of these narcissistic traits to attraction towards romantic
targets would provide good evidence that such partner choices actually represent a self-esteem
regulation strategy.

22

CHAPTER 3: METHOD FOR STUDY 1
Study 1 involved two phases. In the first phase, participants were asked to complete the
Narcissistic Personality Inventory and the Pathological Narcissism Inventory online prior to a
visit to the laboratory. In the second phase, participants were asked to come into the laboratory to
complete a measure of relationship motivations (i.e., the Relationship Initiation Motives
Inventory) and to make ratings of attraction towards three different romantic targets described to
be desirable, caring, or admiring. Importantly, before participants made these ratings of
attraction, their self-esteem levels were manipulated via the provision of false feedback
regarding their performance on a cognitive task.
The primary independent variables in this study included the NPI traits of
Leadership/Authority, Grandiose Exhibitionism, and Entitlement/Exploitativeness, and the PNI
traits of Narcissistic Grandiosity and Narcissistic Vulnerability. The primary dependent variables
included the relationship initiation motivation variables (i.e., Intimacy, Hedonism, Enhancement,
Coping, and Self-Affirmation), attachment-related avoidance, and participants’ ratings of
romantic attraction towards the desirable target, admiring target, and caring target.

2

Participants
All participants were recruited through the Human Participation Research system
maintained by the psychology department at Michigan State University. The study description on
the website requested students who were not currently involved in a romantic relationship and
who were at least 18 years old (see Appendix 1). A total of 504 participants signed up for the

2

All materials used in Study 1 are contained in Appendix 1.
23

3

first phase of the study. Of these participants, 414 of them completed the second phase of the
study. Unfortunately, nine of these participants could not be linked up to the online data because
the emails that they provided at the second phase did not match those provided in the first phase.
Further, six participants were dropped because they did not respond affirmatively to the question,
“Did you answer these questions honestly?” on the online questionnaire. Finally, out of the 394
participants remaining, 48 were dropped due to their relationship status (i.e., 43 were dating
exclusively, four were living with their significant other, and one was divorced).
The sample that I used for the analyses therefore consisted of 346 undergraduate college
students. Most of the participants were women (60.4%; two participants did not report their
gender) and were 18 (28.6%), 19 (23.1%), 20 (21.7%), and 21 (26.3%) years old (one participant
did not report her/his age). Most participants (78%) were White (Not Hispanic/Not Latino) and
reported being either Single and Not Dating (80.3%) or Dating casually (19.7%). Almost all of
the participants self-identified as heterosexual (97.1%).
Materials
Narcissistic Personality Inventory (NPI). The NPI (Raskin & Terry, 1988) consists of a
set of 40 forced-choice items that aim to measure narcissism as a trait in nonclinical populations.
For each item, participants were asked to choose between a narcissistic alternative (e.g., “I like to
be the center of attention”) and a non-narcissistic alternative (e.g., “I prefer to blend in with the
crowd”). Previous work using the NPI has computed either the NPI total score (i.e., the sum or
average of participants’ responses to all of the items) or subscales derived from published
dimensional solutions (e.g., Emmons, 1984, 1987; Raskin & Terry, 1988).

3

This number includes participants that originally signed up for Study 2 but participated in
Study 1.
24

The present research made use of the dimensional solution provided in Ackerman et al.
(2011). The NPI subscales therefore included Leadership/Authority (11 items; α = .77, average
inter-item r = .23; M = 0.52, SD = 0.26), Grandiose Exhibitionism (10 items; α = .71; average
inter-item r = .19; M = 0.37, SD = 0.24), and Entitlement/Exploitativeness (4 items; α = .40;
average inter-item r = .15; M = 0.20, SD = 0.24). Participants’ scores for each of these subscales
were the average number of narcissistic alternatives that they endorsed, with higher scores
reflecting higher levels of the trait. Note that men and women did not significantly differ in their
average levels of Leadership/Authority (men: M = 0.53, SD = 0.27; women: M = 0.52, SD =
0.26; t(342) = 0.38, p = .703), Grandiose Exhibitionism (men: M = 0.38, SD = 0.25; women: M =
0.36, SD = 0.22; t(342) = 0.55, p = .584), and Entitlement/Exploitativeness (men: M = 0.21, SD =
0.23; women: M = 0.20, SD = 0.25; t(342) = 0.44, p = .661).
Pathological Narcissism Inventory (PNI). This measure consists of 52 items intended
to assess the pathological expression of narcissism (Pincus et al., 2009). Participants were asked
to rate how well each statement described them on a 6-point scale that ranged from 0 (Not at all
like me) to 5 (Very much like me). Pincus et al. (2009) originally established the existence of
seven subscales (see Figure 1). Recent work on the PNI, however, has established the existence
of higher order factors of Narcissistic Grandiosity and Narcissistic Vulnerability (Wright et al.,
2010).
Whereas Narcissistic Grandiosity (18 items; α = .81, average inter-item r = .19; M = 2.89,
SD = 0.60) is comprised of subscales that reflect maladaptive self-enhancement strategies (i.e.,
Grandiose Fantasy, Exploitativeness, and Self-Sacrificing Self-Enhancement), Narcissistic
Vulnerability (34 items; α = .93, average inter-item r = .29; M = 2.28, SD = 0.74) is comprised of
subscales that reflect deficient emotional and self-regulation (i.e., Entitlement-Rage, Hiding the

25

Self, Contingent Self-Esteem, and Devaluing). These superordinate scales of Narcissistic
Grandiosity and Narcissistic Vulnerability were used in this research. Note that men and women
did not significantly differ in their average levels of Narcissistic Grandiosity (men: M = 2.92, SD
= 0.60; women: M = 2.87, SD = 0.60; t(342) = 0.66, p = .511) and Narcissistic Vulnerability
(men: M = 2.22, SD = 0.74; women: M = 2.32, SD = 0.74; t(342) = -1.25, p = .212).
Relationship Initiation Motives Inventory. This 25-item measure was adapted from the
Sex Motives Scale developed by Cooper et al. (1998). Given that the original scale by Cooper et
al. (1998) was designed to assess individuals’ reasons for having sex, several modifications were
made so that the scale would reflect reasons for initiating romantic relationships. These
modifications included changing the instructions, changing the scale anchors, dropping
inappropriate subscales, adding a new subscale, and changing the wording for particular items.
Altogether, two scales assessed self-focused, approach-oriented reasons for beginning
relationships (i.e., the Hedonism Motive and Enhancement Motive subscales), two scales
assessed self-focused, avoidance-oriented reasons for beginning relationships (i.e., the SelfAffirmation Motive and Coping Motive subscales), and one scale assessed other-focused,
approach-oriented reasons for beginning relationships (i.e., the Intimacy Motive subscale).
Participants were first presented with the following stem: “I often begin romantic relationships
with new partners …” Participants were then asked to indicate their degree of agreement with a
series of statements on a 5-point Likert-type scale that ranged from 1 (Strongly Disagree) to 5
(Strongly Disagree).
The Intimacy Motive subscale (α = .86, average inter-item r = .56; M = 3.97, SD = 0.71)
included five items that assessed participants’ motive to develop emotional intimacy (e.g., “To
better understand the other person” and “To have another person truly understand me”). The

26

Self-Affirmation Motive subscale (α = .87, average inter-item r = .58; M = 2.57, SD = 0.93)
included five items that assessed participants’ motive to use the relationship to bolster their less
than adequate self-image (e.g., “Because it makes me feel like I’m a more interesting person”
and “To reassure myself that I am desirable”). The Hedonism Motive subscale (α = .78, average
inter-item r = .40; M = 2.93, SD = 0.83) included five items that assessed participants’ motive to
pursue excitement and physical pleasure (e.g., “Just for the excitement of it” and “To satisfy my
sexual needs”). The Enhancement Motive subscale (α = .84, average inter-item r = .52; M = 2.29,
SD = 0.88) included five items that assessed participants’ motive to affirm and enhance their
inflated self-image (e.g., “To impress others” and “Because they remind me I’m a great person”).
The Coping Motive subscale (α = .85, average inter-item r = .54; M = 2.33, SD = 0.88) included
five items that assessed participants’ intentions to use relationships as a means to regulate
negative affect (e.g., “To cope with upset feelings” and “To cheer myself up”). Participants’
scores for each subscale were computed as the mean of their responses, with higher scores
reflecting higher levels of the motive.
Attachment-Related Avoidance. To provide an additional means of assessing
participants’ desire for intimacy, the attachment-related avoidance subscale of the Experiences in
Close Relationships (ECR; Brennan, Clark, & Shaver, 1998) questionnaire was included in the
current study. This subscale assesses participants’ discomfort with closeness. Participants were
asked to indicate how much they agreed with 18 statements on a 7-point scale that ranged from 1
(Disagree Strongly) to 7 (Agree Strongly). Example items included, “I prefer not to show others
how I feel deep down” and “I get uncomfortable when someone wants to be very close to me.”
Each participant’s score was the average of these items (α = .91, average inter-item r = .36; M =
3.40, SD = 1.00), with higher scores indicating higher levels of attachment-related avoidance.

27

Remote Associations Test (RAT). This test, created by Mednick (1962), was used as
part of a procedure to manipulate participants’ state levels of self-esteem. The RAT consists of a
series of three words that are somehow linked together. When presented with the three words
(e.g., Car-Swimming-Cue) on the computer, each participant was asked to come up with the
fourth word that links them together (e.g., “Pool”) and to type it in the blank space on the screen.
The present research used normative data compiled by Bowden and Jung-Beeman (2003) on
RAT item difficulties to construct a cognitive task consisting of 10 easy items, seven difficult
items, and seven moderately difficult items. Between 85% and 100% of the participants in
Bowden and Jung-Beeman (2003) solved the 10 easy items within a 30-second time span.
Similarly, the seven difficult items were correctly solved by 0% to 10% of the participants in the
30-second time span, and the seven moderately difficult items were correctly solved by 49% to
51% of the participants in the 30-second time span. Participants were presented with these 24
items in a mixed order.
Manipulation Check. The manipulation check consisted of the State Self-Esteem Scale
(Heatherton & Polivy, 1991) and two additional questions reported by Jarry and Kossert (2007)
to ensure that participants paid attention to the feedback that was provided by the experimenter.
The State Self-Esteem Scale included 20 items (α = .90, average inter-item r = .33; M = 3.59, SD
= 0.61) that assessed participants’ current levels of self-esteem at the moment (e.g., “I feel good
about myself” and “I feel that others respect and admire me”). Although the State Self-Esteem
Scale technically consists of three separate subscales (i.e., Social Self-Esteem [7 items; α = .84,
average inter-item r = .43; M = 3.73, SD = 0.76], Performance Self-Esteem [7 items; α = .84,
average inter-item r = .44; M = 3.56, SD = 0.73], and Appearance Self-Esteem [6 items; α = .82,
average inter-item r = .43; M = 3.46, SD = 0.69]), some previous research using the scale as a

28

manipulation check has collapsed across the subscales (e.g., Heatherton & Vohs, 2000). In
contrast, other research has examined the subscales separately and has found the Performance
Self-Esteem subscale to be the most impacted by the manipulation (Jarry & Kossert, 2007).
The present research examined each of the three subscales with the expectation that the
Performance Self-Esteem subscale would be the most affected by the self-esteem manipulation.
The last two questions posed to participants in the manipulation check were taken from Jarry and
Kossert (2007) and concern participants’ attention to their performance on the RAT (note that the
RAT is presented to participants as a test of Integrative Orientation or IO for short): “How many
items on the IO test do you think you solved?” and “Please provide your own assessment of your
IO ability using the following 0 to 10 scale, with 0 being the worst and 10 being the best.”
Romantic Targets. Developed by Campbell (1999), this stimulus consisted of a written
scenario and the presentation of three potential romantic targets (the target described as ‘needy’
in Campbell [1999] was not found to be connected to any of the NPI facets and was therefore not
used in the present research). As reported in Campbell (1998, p. 78), the scenario was as follows:
Imagine the following situation: You are at a party with some people you know and some
people you don’t know. A friend of yours introduces you to [three] people of the opposite
sex who are single—we’ll call them by their initials: J.G., C.F., and B.W. Later, you run
into your friend in the kitchen and she tells you about these people in more detail. Your
friend describes the [three] single people you met— J.G., C.F., and B.W.—as follows…
At this point, participants were asked to refer to a handout that contained the three
romantic targets (note that photographs were not attached to the descriptions of the targets). As
was done in Campbell (1999), each target was identified by a pair of initials so that participants
could easily envision the target being a man or a woman. Moreover, the order in which the three

29

romantic targets was presented on the sheet was counterbalanced across participants. The
Admiring target was presented as follows: “J.G. told me that he/she thought you were terrific!
J.G. thought you were the best looking person at the party and that you were very charming. J.G.
also thought you were one of the most intelligent people he/she had ever met.” The Desirable
target was presented to participants as follows: “C.F. is really a great person. C.F. is very popular
and good-looking. He/she is also one of the top students at the school and is rated as one of the
best soccer players in the nation.” The Caring target was presented as follows: “B.W. is a really
caring person. B.W. told me that he/she’s looking for a person to date with whom he/she could
be close, intimate friends. B.W. is one of the most sensitive, sharing, compassionate, and friendly
people I know.”
Attraction Scale. This scale was developed by Campbell (1999) to assess participants’
attraction to the different romantic targets. It consists of five items on 7-point Likert-type scales.
Participants were asked to use a scale with the anchors of 1 (not at all) and 7 (very) to answer
three questions: “How attractive do you find this person?”, “How desirable would you find this
person as a dating partner?”, and “How much would you actually like to date this person?”
Participants were then asked to use the scale anchors of 1 (very bad) and 7 (very good) to
answer, “How would you feel about yourself if you were dating this person?”, and to use the
scale anchors of 1 (disapproving of me) and 7 (approving of me) to answer, “How do you think
your friends would feel about you if you were dating this person?” Participants’ scores for the
attraction scales were the mean of their responses to these five questions, with higher scores
reflecting greater levels of attraction. Attraction scores were computed for the admiring target (α
= .87, average inter-item r = .56; M = 4.73, SD = 1.13), the desirable target (α = .88, average

30

inter-item r = .60; M = 5.12, SD = 1.25), and the caring target (α = .91, average inter-item r =
.68; M = 5.26, SD = 1.20).
Procedure
Before signing up for an in-lab session, participants were asked to complete the NPI and
PNI online. This was done to ease the burden placed on participants’ time in the laboratory when
they came in for the second session. Participants arrived to the laboratory in groups of one to
four. The laboratory was housed in the basement of the psychology building and consisted of a
large room that was connected to a hallway with three smaller rooms. Each of the three smaller
rooms contained a desk with a computer. The larger room contained a long table in the center
and computers along the perimeter of the room. Upon arrival, participants were asked to sit at the
table in the larger part of the laboratory. After providing informed consent, the experimenter told
participants:
We are interested in the relations between different aspects of personality and romantic
attraction. This study will have three parts. In the first part, we will ask you to complete
some measures related to motivation. In the second part, we will ask you to take a test
that assesses an aspect of your cognitive ability. Finally, in the third part of the study, you
will complete some other personality measures and we will ask you to read several
different personality profiles and rate how attractive you find each one.
The experimenter then randomly assigned participants to one of four rooms and asked
them to complete the Relationship Initiation Motives Inventory on the computer. Once
participants completed the Relationship Initiation Motives Inventory, the experimenter
implemented a procedure designed to manipulate participants’ state levels of self-esteem. The
self-esteem manipulation used in this study was very similar to procedures that have been used in

31

previous research (Heatherton & Vohs, 2000; Jarry & Kossert, 2007). Participants completed a
cognitive task that contained a combination of easy, difficult, and moderately difficult questions,
and were then provided with false feedback that ostensibly informed them of their standing on a
cognitive ability. This study used a cover story to ensure that participants believed the cognitive
ability being tested was important and relevant to their lives. The cover story used in this study
drew heavily on the script reported in Jarry and Kossert (2007; see pp. 42-43). Each participant
was told:
As mentioned at the beginning, I will now ask you to take a test that assesses an
important cognitive ability called Integrative Orientation, or IO for short. The IO ability
involves the capacity to discern the relationships among different kinds of information
and to use that information to effectively solve analytical problems. IO has been
extensively researched for the past 10 years and has been found to be an important
predictor of success in college and in future employment. Have you heard of the IO
ability in your psychology courses [asked casually]?
After this speech, the experimenter introduced participants to the test purportedly used to
assess IO (i.e., the RAT). Participants were then presented with two example RAT problems to
ensure that they understood the task. After participants conveyed that they understood, the
experimenter provided them with the Partner Descriptions sheet with the romantic targets that
they would need to refer to later in the study. The experimenter then escorted participants back to
their computers to complete the IO test. Participants were then presented with a set of RAT
problems of varying difficulty (i.e., 10 easy items, 7 difficult items, and 7 moderately difficult
items).

32

Depending upon the room that participants were randomly assigned to at the beginning of
the study, participants experienced one of three experimental conditions: (a) the Success
Condition, (b) the Control Condition, or (c) the Failure Condition. Once participants completed
their set of problems, the computer provided them with immediate feedback on their purported
performance. In the Success Condition, the computer informed participants, “You performed
very well and scored within the top 15% of all students tested at Michigan State University.” In
the Control Condition, the computer did not provide participants with any feedback. In the
Failure Condition, the computer informed participants, “You did not perform very well and
scored within the bottom 30% of all students tested at Michigan State University.”
After providing feedback, the computer directed participants to complete the State Self-Esteem
Scale and then to answer the two questions explicitly connected to the IO task. After this,
participants were asked to refer to the Partner Descriptions sheet that they were given before
with the three descriptions of the romantic targets and they were asked to rate how attractive they
found each one using the Attraction Scale. Once they completed this, participants were thanked
and fully debriefed.

33

CHAPTER 4: RESULTS FOR STUDY 1
Narcissism and Relationship Initiation Motivations
Before proceeding to analyze associations between the narcissistic dimensions and the
Relationship Initiation Motives Inventory subscales, an exploratory factor analysis (EFA) was
conducted on the Pearson product-moment correlation matrix of the 25 items from the
Relationship Initiation Motives Inventory to evaluate the dimensional structure of the instrument.
Path analyses were then used to test the first set of hypotheses concerning the relations of the
narcissistic dimensions to the various relationship initiation motives. To provide an additional
test of the relation between these narcissistic dimensions and the desire for intimacy, multiple
regression analyses were performed with attachment-related avoidance as the criterion. Each set
of analyses was evaluated for gender moderation.
Exploratory factor analysis of relationship initiation motives inventory. SPSS
version 19.0 was used to conduct the EFA. Principal axis factoring was used as the estimation
method and the factors were rotated using the oblique promax procedure. Based on guidelines
proposed by Fabrigar, Wegener, MacCallum, and Strahan (1999), the number of factors to
extract was decided upon through a combination of examination of the scree plot of eigenvalues
and evaluation of the interpretability of the factors in the solution. Results revealed four
eigenvalues above 1.0 (8.835, 3.505, 1.727, and 1.502). Figure 2 displays a scree plot of the
eigenvalues. As can be seen, the last drop occurs between the fifth and fourth eigenvalues,
therefore suggesting a four-factor solution. In addition, the most noticeable bends occurred
between the second and third eigenvalues (difference = 1.778) and between the fourth and fifth
eigenvalues (difference = 0.522). Together with the fact that five dimensions were initially

34

anticipated, two-, four-, and five-factor solutions were therefore extracted and subsequently
evaluated in regards to their interpretability.
The two-factor solution appeared to isolate the intimacy content from the selfaffirmation, coping, enhancement, and hedonism content. Thus, it represented a clean break
between the self-focused motives and the other-focused motive. Nonetheless, the self-focused
factor was somewhat difficult to interpret given that it combined diverse content related to
sensation-seeking, coping, and self-enhancement. The four-factor solution improved upon this
because in addition to isolating the intimacy content, it divided up the self-focused motivation
content into areas reflecting sensation-seeking, coping, and self-enhancement. The five-factor
solution did not appear to improve upon the four-factor solution. That is, it represented a
virtually identical solution with the exception that the item “Because I feel horny” defined an
additional fifth factor. In light of this, a four-factor solution was extracted. Table 2 shows the
pattern loadings for this solution.
As Table 2 shows, there was clear evidence for the existence of the Intimacy,
Enhancement, Coping, and Hedonism motive subscales. However, items from the SelfAffirmation subscale appeared to primarily load on the Coping factor. I therefore decided to
retain the Relationship Initiation Motives Inventory subscales for Intimacy, Enhancement, and
Hedonism for use in the analyses. In addition, I decided to calculate an aggregated Coping/SelfAffirmation subscale that averaged across the Coping and Self-Affirmation subscale items (10
items; M = 2.45, SD = 0.85; α = .91, average inter-item r = .51). This Coping/Self-Affirmation
subscale consequently reflected coping with negative self-images and affect in general.
Zero-order correlations between the narcissism measures, the relationship initiation
motives inventory measures, and the attachment-related avoidance measure. Table 3

35

presents zero-order correlations between the NPI subscales and the PNI subscales. The NPI
subscales were all moderately correlated with one another, with the largest association occurring
between Leadership/Authority and Grandiose Exhibitionism. Moreover, the PNI Vulnerability
scale was strongly correlated with the PNI Grandiosity scale. Further, whereas the NPI subscales
were all moderately correlated with the PNI Grandiosity scale, only the NPI
Entitlement/Exploitativeness scale was associated with the PNI Vulnerability scale.
Table 4 shows the zero-order correlations between the primary dependent variables. As
can be seen, all of the self-focused Relationship Initiation Motives Inventory subscales were
strongly associated with each other. In contrast, the Intimacy subscale was modestly associated
with Coping/Self-Affirmation and moderately correlated with Enhancement. Attachment-related
avoidance demonstrated a modest positive association with Coping/Self-Affirmation.
Table 5 displays the zero-order correlations between the narcissism measures and the
primary dependent variables. NPI Grandiose Exhibitionism was positively connected to
Hedonism and Enhancement. Moreover, NPI Entitlement/Exploitativeness was positively
connected to each of the self-focused motives and negatively connected to the Intimacy motive
subscale. PNI Grandiosity was positively connected to all of the self-focused motives. PNI
Vulnerability demonstrated a similar pattern of relations with the exception that it exhibited a
modest positive association with the Intimacy motive subscale. Table 5 also shows that NPI
Grandiose Exhibitionism was negatively related to attachment-related avoidance whereas PNI
Vulnerability was positively related to avoidance.
Path analysis evaluating relations between NPI traits and relationship initiation
motivations. A path analysis was used to test the hypotheses concerning the relation of the NPI
traits to the various relationship initiation motives. As shown in Figure 3, the NPI subscales of

36

Leadership/Authority, Grandiose Exhibitionism, and Entitlement/Exploitativeness were specified
as exogenous variables in the model. Moreover, the four Relationship Initiation Motives
Inventory subscales of Intimacy, Coping/Self-Affirmation, Hedonism, and Enhancement were
treated as the endogenous variables. Correlations were freely estimated between the NPI
subscales. Likewise, correlations were freely estimated between the residuals for the
Relationship Initiation Motives Inventory subscales. As such, this model was fully saturated.
Table 6 shows the parameter estimates for this model (estimated using AMOS version
19.0). As hypothesized, Leadership/Authority negatively predicted the Coping/Self-Affirmation
motive. Similarly, Grandiose Exhibitionism positively predicted the Hedonism and Enhancement
motives whereas Entitlement/Exploitativeness positively predicted the Coping/Self-Affirmation
motive and negatively predicted the Intimacy motive. In contrast to what was predicted,
however, there was no evidence that the Leadership/Authority or Grandiose Exhibitionism
subscales were connected to lower levels of the Intimacy motive.
Although not hypothesized, Leadership/Authority also negatively predicted the
Enhancement motive. Given the zero-order relation between these variables, this finding
provides evidence of a suppression effect and suggests that controlling for the overlap with
grandiosity and vulnerability makes the connection between normal narcissism and this motive
clearer. In addition, Entitlement/Exploitativeness positively predicted the Hedonism and
Enhancement motives. This suggests that self-enhancement may be another important reason that
entitled individuals initiate their romantic relationships.
Gender moderation. Gender moderation was evaluated for in the previous analysis by
performing a multiple-groups analysis. Table 7 displays the parameter coefficients from the
unconstrained multiple-groups analysis model for men and women. Imposing equality

37

constraints on the structural weights in the model (i.e., the coefficients for the paths between the
NPI subscales and the Relationship Initiation Motives Inventory subscales) for men and women
2

adversely affected model fit, χ (12) = 27.00, p = .008. Examination of the modification indices
suggested removing equality constraints for the following paths: (a) Leadership/Authority to
Enhancement; (b) Entitlement/Exploitativeness to Coping/Self-Affirmation; and (c)
Entitlement/Exploitativeness to Intimacy. Freely estimating these path coefficients for men and
2

women resulted in reasonable model fit, χ (9) = 16.34, p = .06, thereby suggesting that men and
women differed in regards to these particular variable relations.
Examination of the unconstrained parameters from this less restricted model revealed that
Leadership/Authority was not significantly related to the Enhancement motive for men, b = 0.26, SE = .26, β = -.08, p = .301; however, Leadership/Authority was significantly related to the
Enhancement motive for women, b = -0.72, SE = .21, β = -.22, p < .001. This model also showed
that the relation between Entitlement/Exploitativeness and Coping/Self-Affirmation was stronger
for men (b = 1.16, SE = .25, β = .31, p < .001) than for women (b = 0.56, SE = .21, β = .17, p =
.009). Finally, whereas there was no significant relation between Entitlement/Exploitativeness
and the Intimacy motive for women (b = -0.18, SE = .21, β = -.06, p = .39), men with higher
levels of Entitlement/Exploitativeness reported lower levels of Intimacy motivations (b = -0.76,
SE = .24, β = -.26, p = .002).
Path analysis evaluating relations between PNI traits and relationship initiation
motivations. A path analysis was also used to evaluate the hypotheses concerning relations
between the PNI traits and the relationship initiation motives. Figure 4 shows the specified
model. As shown, the PNI scales of Narcissistic Grandiosity and Narcissistic Vulnerability were
specified as exogenous variables. In addition, the four Relationship Initiation Motives Inventory
38

subscales of Intimacy, Coping/Self-Affirmation, Hedonism, and Enhancement were treated as
the endogenous variables. The PNI scales were allowed to correlate. Likewise, correlations were
freely estimated between the residuals for the Relationship Initiation Motives Inventory
subscales. As before, this path analysis was a fully saturated model.
Table 8 presents the parameter estimates. There was no support for the hypothesis that
higher levels of Narcissistic Grandiosity are connected to greater levels of self-enhancing
relationship initiation motivations. Likewise, there was no support for the hypothesis that
Narcissistic Grandiosity and Narcissistic Vulnerability would be negatively related to the
Intimacy motive. In fact, there was even a weak positive relation between Narcissistic
Vulnerability and the Intimacy motive. Nevertheless, the hypothesis that Narcissistic
Vulnerability would positively predict the Coping/Self-Affirmation motive was supported in
these data. Although not originally hypothesized, Narcissistic Vulnerability also positively
predicted the Hedonism and Enhancement motives. These findings suggest that the relationship
initiation motivations endorsed by individuals with higher levels of Narcissistic Vulnerability
serve both self-enhancing and self-protective functions.
Gender moderation. Gender moderation was evaluated for in the previous analysis by
conducting a multiple-groups analysis. Table 9 presents the parameter coefficients for men and
women from the unconstrained multiple-groups analysis model. Imposing equality constraints on
2

the structural weights in the model for men and women did not adversely affect model fit, χ (8)
= 6.39, p = .603. Thus, there was no evidence that gender moderated these effects.
Summary for analyses involving narcissism and the relationship initiation motives
inventory subscales. The measures of normal narcissism, grandiose narcissism, and vulnerable
narcissism were each hypothesized to demonstrate a unique pattern of relations with the
39

relationship initiation motive measures. As hypothesized, Leadership/Authority negatively
predicted the coping relationship initiation motive. Leadership/Authority also negatively
predicted the enhancement relationship initiation motivation. Although results showed that NPI
Grandiose Exhibitionism positively predicted the Hedonism and Enhancement relationship
initiation motives, PNI Narcissistic Grandiosity was unrelated to the self-focused relationship
initiation motives. Consistent with predictions, the vulnerable narcissistic traits positively
predicted the coping relationship initiation motivation. However, whereas NPI
Entitlement/Exploitativeness negatively predicted the intimacy motive, PNI Narcissistic
Vulnerability positively predicted it. Finally, although not originally predicted, the vulnerable
narcissistic traits both positively predicted the self-enhancing relationship motives.
Narcissism and attachment-related avoidance. As an additional test of the association
between narcissism and the desire for intimacy, a series of multiple regression analyses were
performed to evaluate the independent associations of the NPI and PNI traits with the ECR
subscale of attachment-related avoidance. A multiple regression analysis using the NPI traits to
predict attachment-related avoidance found that Leadership/Authority was unrelated to
attachment-related avoidance, b = 0.20, SE = .22, β = .05, p = .367. In contrast, Grandiose
Exhibitionism negatively predicted attachment-related avoidance, b = -1.14, SE = .25, β = -.27, p
< .001; further, Entitlement/Exploitativeness positively predicted avoidance, b = 0.56, SE = .23,
β = .13, p = .014. None of these effects were significantly moderated by gender. Thus, across
both men and women, individuals with higher levels of Entitlement/Exploitativeness reported
more discomfort with intimacy whereas individuals with higher levels of Grandiose
Exhibitionism reported less discomfort with intimacy.

40

A multiple regression analysis using the PNI traits to predict attachment-related
avoidance found that Narcissistic Grandiosity negatively predicted attachment-related avoidance,
b = -0.44, SE = .10, β = -.26, p < .001. Conversely, Narcissistic Vulnerability positively predicted
scores on attachment-related avoidance, b = 0.63, SE = .08, β = .47, p < .001. Additional
analyses including participant gender revealed significant interactions between Narcissistic
Grandiosity and participant gender, b = -0.38, SE = .10, p < .001, and between Narcissistic
Vulnerability and participant gender, b = 0.18, SE = .08, p = .027. To break down these
interactions, simple slopes for men and women were estimated for these relations between the
PNI traits and attachment-related avoidance.
The simple slope for women for the relation between PNI Narcissistic Grandiosity and
avoidance was, b = -0.12, SE = .13, β = -.07, p = .367. Moreover, the simple slope for men for
the relation between PNI Narcissistic Grandiosity and avoidance was, b = -0.88, SE = .15, β = .53, p < .001. Thus, the relation between PNI Narcissistic Grandiosity and attachment-related
avoidance appeared to only hold true for men. The simple slope for women for the relation
between PNI Narcissistic Vulnerability and avoidance was, b = 0.48, SE = .10, β = .36, p < .001.
Further, the simple slope for men for the relation between PNI Narcissistic Vulnerability and
avoidance was, b = 0.83, SE = .12, β =.62, p < .001. This suggests that although the relation
between Narcissistic Vulnerability and attachment-related avoidance was positive for men and
women, the magnitude of the association was stronger for men.
In sum, each of the narcissistic dimensions exhibited somewhat divergent relations with
attachment-related avoidance. Leadership/Authority was unrelated to attachment-related
avoidance. Moreover, individuals who possessed greater levels of grandiose narcissism reported
lower levels of attachment-related avoidance, therefore suggesting that they are more

41

comfortable with intimacy. However, evidence for the relation between PNI Grandiosity and
attachment-related avoidance only held for men. In contrast, individuals who possessed greater
levels of vulnerable narcissism reported higher levels of attachment-related avoidance; this
relation was particularly pronounced for men.
Narcissism and Attraction towards the Romantic Targets
The second set of hypotheses in Study 1 concerned whether the different narcissistic
dimensions are associated with differential attraction towards particular romantic targets. Before
testing these hypotheses, the effect of target presentation order on participants’ attraction towards
the romantic targets was evaluated. Multiple regression analyses were then used to test the
primary hypotheses regarding the relations between the narcissistic dimensions and attraction
towards the admiring, desirable, and caring romantic targets.
Main effect of target presentation order on attraction towards the romantic targets.
As discussed earlier, participants received presentations of the admiring, desirable, and caring
romantic targets that differed in their ordering. Altogether, six different presentation orders were
possible (1 = Admiring, desirable, and caring; 2 = Admiring, caring, and desirable; 3 =
Desirable, admiring, and caring; 4 = Desirable, caring, and admiring; 5 = Caring, desirable, and
admiring; 6 = Caring, admiring, and desirable). A series of one-way between-subjects ANOVAs
were used to assess the impact of ordering presentation on participants’ ratings of attraction
towards the different romantic targets. Table 10 shows the means and standard deviations for
participants’ ratings of attraction towards the different targets as a function of the order in which
they were presented.
Ordering was not observed to impact participants’ ratings of attraction towards the
admiring target, F (5, 336) = 0.11, MSE = 1.26, p = .99. Similarly, ordering did not significantly

42

affect participants’ ratings of attraction towards the caring target, F (5, 336) = 1.48, MSE = 1.40,
4

p = .195. However, ordering did impact participants’ ratings of attraction towards the desirable
target, F (5, 336) = 3.30, MSE = 1.49, p = .006. A post-hoc Tukey analysis revealed that
presentation order 2 (where the admiring target was presented first, the caring target was
presented second, and the desirable target was presented last) significantly differed from
presentation order 4 (where the desirable target was presented first, the caring target was
presented second, and the admiring target was presented last). There was no evidence of
differences for the other combinations. This suggests that the position of the description for the
desirable target influenced participants’ subsequent attraction ratings towards the desirable
target. Target presentation order did not moderate any of the effects in the following analyses
examining participants’ attraction towards the desirable target. Unless otherwise noted,
controlling for target presentation order also did not substantially alter results for analyses
involving attraction towards the desirable target.
Zero-order correlations between the narcissism measures and the attraction
measures. Table 11 shows the zero-order correlations between the attraction-towards-target
measures. As can be seen, participants’ scores for the attraction measures concerning the three
romantic targets were modestly related. Moreover, Table 12 shows that NPI
Leadership/Authority was positively correlated with attraction towards a desirable partner. NPI
Grandiose Exhibitionism was also positively correlated with attraction towards a desirable
partner, and to a weaker extent, attraction towards an admiring target. Last, PNI Grandiosity was
positively correlated with attraction towards a desirable partner.
4

A Brown-Forsyth (1974a) test indicated heterogeneity of variance, F (5, 336) = 3.46, MSE =
0.50, p = .005. Nevertheless, the Brown and Forsyth (1974b) robust test of equality of means
revealed a similar pattern of results, FBrown-Forsyth (5, 282.04) = 1.48, p = .197.
43

Multiple regression analyses using NPI subscales to predict attraction towards the
romantic targets. Table 13 shows the results from three multiple regression analyses in which
the three NPI subscales were used to predict participants’ attraction towards each of the romantic
targets. Note that these analyses were examining the independent effects of each of these NPI
subscales collapsing across all three RAT conditions. As can be seen, none of my hypotheses
were supported.
Instead, Table 13 shows that Leadership/Authority significantly predicted participants’
attraction towards the desirable target, such that participants with higher levels of
Leadership/Authority reported being more attracted to the desirable target. Although not
statistically significant, the coefficient for Grandiose Exhibitionism was virtually identical in
magnitude to that of Leadership/Authority. When the effect of target presentation order was
controlled for in these analyses, the relation between Grandiose Exhibitionism and attraction
towards the desirable target became significant.
Gender moderation. Each of the above analyses was rerun with gender as a moderator of
each of the effects. The multiple regression analysis with attraction towards the admiring target
as an outcome revealed a significant interaction between gender and Leadership/Authority, b =
0.61, SE = .27, p = .025, and a significant interaction between gender and Grandiose
Exhibitionism, b = -0.63, SE = .30, p = .035. To follow up these interactions, simple slopes for
men and women were estimated for the relations between the NPI subscales and attraction
towards the admiring target.
The simple slope for women for the effect of Leadership/Authority on attraction towards
the admiring target was, b = -0.34, SE = .32, β = -.08, p = .296. In contrast, the simple slope for
men for this relation was, b = 0.89, SE = .44, β = .21, p = .045. Thus, it is only for men that

44

higher levels of Leadership/Authority predict greater attraction towards the admiring target. The
simple slope for women for the effect of Grandiose Exhibitionism on attraction was, b = 0.91, SE
= .38, β = .19, p = .017. The simple slope for men for this relation was, b = -0.35, SE = .46, β = .07, p = .45. These results indicate that only women with higher levels of Grandiose
Exhibitionism report greater attraction towards the admiring target.
A second multiple regression analysis with attraction towards the desirable target as the
outcome revealed a significant interaction between gender and Leadership/Authority, b = 0.85,
SE = .30, p = .005. To break this down, simple slopes for men and women were again estimated.
The simple slope for women was, b = -0.05, SE = .35, β = -.01, p = .893. The simple slope for
men was, b = 1.65, SE = .48, β =.34, p = .001. Consequently, only men with higher levels of
Leadership/Authority reported greater attraction towards the desirable target.
A third multiple regression analysis with attraction towards the caring target as the
outcome did not reveal any significant interactions with gender. Thus, there was no evidence of
gender moderation for these effects.
Multiple regression analyses using PNI scales to predict attraction towards the
romantic targets. Table 14 shows results from three multiple regression analyses in which the
two PNI scales were used to predict participants’ attraction towards each of the romantic targets.
Again, these analyses examined the independent effects of the PNI traits collapsing across the
three RAT conditions. As can be seen in Table 14, none of my hypotheses were supported.
Instead, the PNI Grandiosity scale significantly predicted attraction towards the desirable target,
such that individuals with higher levels of PNI Grandiosity reported being more attracted to the
desirable target.

45

Gender moderation. The above analyses were rerun with gender as a moderator of the
effects. None of the analyses revealed any significant interactions with gender.
Summary for analyses using the narcissism variables to predict attraction towards
the romantic targets. Each of the narcissistic dimensions was hypothesized to be associated
with attraction towards a particular romantic target. In partial support of my hypotheses, I found
that women with higher levels of Grandiose Exhibitionism reported greater levels of attraction
towards the admiring target. However, no other hypotheses were supported. Instead, participants
with higher levels of normal and grandiose narcissistic traits reported greater attraction towards
the desirable target.
Regulation of Self-Esteem through Romantic Attraction
The third set of hypotheses in Study 1 dealt with whether attraction towards particular
romantic targets can serve as a means for narcissists to regulate self-esteem. To examine this,
participants were randomly assigned to one of three RAT conditions before providing their
ratings of attraction towards the three romantic targets. Before performing the main analyses, the
manipulation checks for the RAT condition were assessed. After determining the efficacy of the
self-esteem manipulation, multiple regression analyses with the RAT condition as a moderator
were used to evaluate the main hypotheses. Gender moderation was evaluated for in each of the
primary analyses.
Manipulation checks. Two questions were posed to participants after the IO task to
assess whether they paid attention to the feedback that they had received. The first question
asked participants to report how many problems they believed they had solved successfully.
Because some participants wrote words instead of typing in numbers in response to this question,
some of the data had to be modified. Numbers written in words were transformed into numbers

46

(n = 10). Moreover, expressions such as, “about half” and “75% of them” (n = 11) were treated
as their corresponding numerical values (i.e., 12 and 18, respectively). Other expressions that
conveyed ambiguity (e.g., “most” and “not many”) or were nonsensical (e.g., “40545”) were
treated as missing data (n = 19).
A one-way between-subjects ANOVA was used to determine the effect of the RAT
condition (i.e., success vs. failure vs. control) on participants’ responses to this first question.
Results revealed that the RAT condition had a significant effect on their responses, F (2, 324) =
5.78, MSE = 13.08, p = .003. A post-hoc Tukey analysis showed that participants in the success
condition reported solving significantly more problems (n = 144, M = 7.23, SD = 4.10) than
participants in the failure condition (n = 96, M = 5.67, SD = 3.05), Cohen’s d = 0.42. However,
the success condition did not differ significantly from the control condition (n = 87, M = 6.20,
SD = 3.34), Cohen’s d = 0.27, and the control condition did not differ significantly from the
failure condition, Cohen’s d = 0.17.
The second question after the IO task asked participants to provide their own assessment
of their IO ability using a 10-point scale that ranged from 0 (the worst) to 10 (the best). A oneway between-subjects ANOVA was used to evaluate the effect of the RAT condition on
participants’ responses to this second question. Results revealed that the RAT condition had a
significant effect on participants’ responses, F (2, 343) = 21.05, MSE = 2.69, p < .001. A posthoc Tukey analysis showed that participants in the success condition reported having a
significantly higher IO ability (n = 152, M = 5.28, SD = 1.73) than participants in the control
condition (n = 91, M = 4.62, SD = 1.60), Cohen’s d = 0.39, and the failure condition (n = 103, M
= 3.93, SD = 1.54), Cohen’s d = 0.82, conditions. Likewise, participants in the control condition
reported having a significantly higher IO ability than participants in the failure condition,

47

Cohen’s d = 0.44. In sum, participants’ responses to these first two questions suggest that
participants’ opinions of their performance and ability were indeed impacted by the RAT
condition that they were placed in during the study.
The next set of manipulation checks determined whether the RAT condition impacted
participants’ state levels of self-esteem. A one-way between-subjects ANOVA evaluating the
effect of the RAT condition on participants’ total state self-esteem levels revealed that the RAT
condition did not have an overall impact on the omnibus measure of state self-esteem, F (2, 343)
= 2.82, MSE = 0.36, p = .061. Additional one-way between-subjects ANOVAs showed that the
RAT condition also did not significantly affect participants’ social state self-esteem levels, F (2,
343) = 1.46, MSE = 0.57, p = .233, or their appearance state self-esteem levels, F (2, 343) = 0.06,
MSE = 0.48, p = .938. However, the RAT condition did significantly affect participants’
performance state self-esteem levels, F (2, 343) = 7.26, MSE = 0.52, p = .001.

5

A post-hoc Tukey analysis revealed that participants in the success condition (n = 152, M
= 3.72, SD = 0.63) reported significantly higher performance state self-esteem than participants
in the failure condition (n = 103, M = 3.40, SD = 0.81), Cohen’s d = 0.45, or the control
condition (n = 91, M = 3.47, SD = 0.75), Cohen’s d = 0.37. However, there was no evidence that
participants’ performance state self-esteem levels were different between the failure and control
RAT conditions, Cohen’s d = -0.09. Taken together, the results provide evidence for the efficacy
of the self-esteem manipulation. Although it did not impact the omnibus measure of state selfesteem, the RAT condition significantly affected participants’ performance state self-esteem
levels as expected.
5

A Brown-Forsyth (1974a) test indicated heterogeneity of variance, F (2, 343) = 3.49, MSE =
0.22, p = .032. Nevertheless, the Brown and Forsyth (1974b) robust test of equality of means
revealed a similar pattern of results, FBrown-Forsyth (2, 280.17) = 6.86, p = .001.
48

Main effect of RAT condition on attraction towards the romantic targets. Table 15
shows the means and standard deviations for attraction towards each romantic target as a
function of the RAT condition. A one-way between-subjects ANOVA evaluating the effect of
the RAT condition on attraction towards the admiring target did not reveal a significant effect of
condition, F (2, 343) = 0.21, MSE = 1.27, p = .81. Likewise, a one-way between-subjects
ANOVA evaluating the effect of the RAT condition on attraction towards the desirable target did
not reveal a significant effect of condition, F (2, 343) = 1.68, MSE = 1.56, p = 19. However, a
one-way between-subjects ANOVA testing the effect of RAT condition on attraction towards the
caring target revealed a significant effect for condition, F (2, 343) = 3.05, MSE = 1.43, p = .049.
A post-hoc Tukey analysis revealed a marginally significant difference between the
success and failure conditions (p = .05), such that participants receiving negative feedback
regarding their performance on the IO task reported greater levels of attraction towards the
caring target. In contrast, no significant differences were found between the success and control
conditions or between the failure and control conditions.
Does the self-esteem manipulation moderate the relations between the narcissism
variables and romantic attraction towards the admiring, desirable, and caring targets? A
series of multiple regression analyses were used to evaluate whether the narcissism-attraction
associations were moderated by the self-esteem manipulation. Two dummy codes were created
to represent the RAT condition: Success (i.e., 1 for the success condition and 0 otherwise) and
Failure (i.e., 1 for the failure condition and 0 otherwise). The reference group when including
these dummy codes in the analysis was therefore the control condition. The NPI and PNI
subscales were each grand-mean centered on their respective means. Interaction terms were

49

formed between each of the dummy coded RAT condition variables and the centered NPI and
PNI scales.
Using the NPI to predict attraction towards the romantic targets. Each multiple
regression analysis used the dummy codes for the RAT condition, the centered NPI subscales,
and the interaction terms between the centered NPI subscales and the dummy codes for the RAT
condition to predict participants’ attraction towards the romantic target. Hierarchical regression
analyses were used to evaluate the incremental change in R-squared that resulted from including
the interaction terms in the model. Similar to the previous multiple regression analyses
examining the independent effects of the NPI variables, I was interested in the extent to which a
block of interaction terms between a particular NPI variable (e.g., Grandiose Exhibitionism) and
the RAT condition dummy codes explained additional variance in the outcome over and above
the inclusion of the other NPI variables and their interaction terms with the RAT condition
dummy codes. Thus, each test of R-squared change for a block of interaction terms corresponds
to the case when it is the final block in the model.
Table 16 presents results from a multiple regression analysis in which participants’
attraction towards the admiring target was the outcome. As can be seen, each of the blocks
containing interactions between the RAT condition dummy codes and the NPI variables did not
account for significant additional variance in participants’ attraction towards the admiring target.
Thus, there was no evidence that the self-esteem manipulation moderated the narcissismattraction relations.
Table 17 presents results from a multiple regression analysis in which participants’
attraction towards the desirable target was the outcome. Analyses showed that the blocks
containing the interactions between the RAT condition dummy codes and the NPI variables did

50

not contribute significant additional variance to the outcome. However, when target presentation
order was controlled for in the analysis, the block containing the interactions between the RAT
condition dummy codes and Entitlement/Exploitativeness explained significant additional
2

variance in participants’ attraction towards the desirable target, ΔR = .018, ΔF (2, 329) = 3.18, p
= .043. This finding was followed up by computing simple slopes for the
Entitlement/Exploitativeness-Attraction relation in the success, failure, and control RAT
conditions. As Table 18 shows, Entitlement/Exploitativeness negatively predicted attraction
towards the desirable target in the failure condition. Thus, individuals with higher levels of
Entitlement/Exploitativeness reported being less attracted to the desirable target when they
received negative feedback regarding their performance on the IO task. In contrast,
Entitlement/Exploitativeness was unrelated to attraction towards the desirable target in the
success and control conditions.
Table 19 presents results from a multiple regression analysis in which participants’
attraction towards the caring target was the outcome. As seen in Table 19, none of the blocks
containing interactions between the RAT condition dummy codes and the NPI variables
explained significant additional variance in the outcome. Thus, there was little evidence to
suggest that the self-esteem manipulation moderated the relations between the NPI variables and
participants’ attraction towards the caring target.
Gender moderation. Additional hierarchical regression analyses were used to evaluate
whether gender moderated any of the two-way interactions between the NPI subscales and RAT
condition. Out of nine tests, only one suggested moderation. The block containing the effects of
the two-way interactions between Leadership/Authority and the RAT condition dummy codes on
2

attraction towards the caring target was significantly moderated by gender, ΔR = .026, ΔF (2,
51

326) = 4.49, p = .012. To follow this up, simple slopes for the two-way interaction between
gender and Leadership/Authority were estimated for each RAT condition. The simple slope for
the control condition was, b = -0.11, SE = .55, p = .847. The simple slope for the success
condition was, b = -0.25, SE = .42, p = .56. The simple slope for the failure condition was, b =
1.67, SE = .54, p = .002. These results indicate that the relation between Leadership/Authority
and attraction towards the caring target only differed for men and women when they were given
negative feedback on the IO task.
This significant two-way interaction was further broken down by computing simple
slopes for men and women. The simple slope for women in the failure RAT condition was, b = 0.27, SE = .34, β = -.06, p = .439. Moreover, the simple slope for men was, b = 0.31, SE = .47, β
=.07, p = .509. Thus, whereas women tended to report being less attracted to the caring target in
the failure condition, men tended to report being more attracted to the caring target. Because
both of these simple slopes were not significantly different from zero, however, the finding that
these relations differed for men and women should be treated with caution.
Using the PNI to predict attraction towards the romantic targets. I used the same
strategy outlined before for the PNI scales. Table 20 presents results from a multiple regression
analysis in which attraction towards the admiring target was the outcome. Moreover, Table 21
presents results from a multiple regression analysis in which participants’ attraction towards the
desirable target was the outcome. As can be seen, none of the blocks containing interactions
between the PNI variables and the RAT condition dummy codes accounted for significant
additional variance in the outcomes.
Table 22 presents results from a multiple regression analysis in which participants’
attraction towards the caring target was the outcome. The block containing the two-way

52

interactions between Narcissistic Vulnerability and the RAT condition dummy codes explained
significant additional variance in participants’ attraction towards the caring target. Simple slopes
analysis was used to follow up these significant interactions. Results are presented in Table 23
and show that Narcissistic Vulnerability positively predicted attraction in the failure RAT
condition. Thus, individuals with higher levels of Narcissistic Vulnerability reported being more
attracted to the caring target when they received negative feedback regarding their performance
on the IO task. The other relations were negative but non-significant.
Gender moderation. Hierarchical regression analyses were again used to assess whether
gender moderated any of the two-way interactions between the PNI scales and the RAT
condition dummy codes. None of the tests revealed evidence of gender moderation.
Summary of analyses evaluating moderation of the narcissism-attraction relations
by the self-esteem manipulation. It was hypothesized that lowering participants’ state selfesteem levels would strengthen the magnitudes of the previously proposed associations between
the grandiose and vulnerable narcissistic dimensions and attraction towards particular romantic
targets. Overall, there was little support for this hypothesis. The self-esteem manipulation did not
moderate any of the associations between the grandiose narcissism measures and attraction
towards the romantic targets. Moreover, in contrast to what was hypothesized, individuals with
higher levels of Entitlement/Exploitativeness reported being less attracted to the desirable target
when their state self-esteem was lowered. In a similar manner, individuals with higher levels of
Narcissistic Vulnerability reported being more attracted to the caring target when their state selfesteem was lowered. Such findings run counter to what was originally predicted.

53

CHAPTER 5: DISCUSSION FOR STUDY 1
Study 1 aimed to better understand the roles of normal, grandiose, and vulnerable
narcissistic traits in the appraisal of initial attraction. Towards this end, Study 1 examined
whether these traits were predictive of several relationship initiation motivations. It also
investigated what types of partners are considered attractive by people with differing levels of
these narcissistic dimensions and whether attraction towards these different romantic partners
can serve as a means to regulate self-esteem. Below I summarize the main results from Study 1
for each of the narcissistic dimensions and then briefly discuss conclusions.
Normal Narcissism and the Appraisal of Initial Attraction
A robust finding in the literature is that people with higher levels of Leadership/Authority
possess greater levels of self-esteem (Ackerman et al., 2011; Brown, Budzek, & Tamborski,
2009; Rhodewalt & Morf, 1995; Watson & Biderman, 1993) and experience lower levels of
psychological distress (Emmons, 1984; Watson & Biderman, 1993; Watson & Morris, 1991). It
therefore follows that people with higher levels of this trait are unlikely to pursue their romantic
relationships to improve their self-image or to ameliorate negative emotions. Consistent with
this, participants with higher levels of Leadership/Authority reported less frequently pursuing
their romantic relationships to cope with negative emotions, to bolster an inadequate sense of
self, and to enhance a potentially inflated self-image.
This research also found that men with higher levels of Leadership/Authority were more
attracted to potential romantic partners that were described to be admiring or desirable. Indeed,
romantic attraction towards such partners may represent adaptive self-esteem regulation
strategies, such as asserting dominance (for attraction towards admiring partners) and
achievement striving (for attraction towards perfect partners; Pincus & Lukowitsky, 2010). The

54

finding that more dominant men are attracted to desirable partners is also consistent with
theorizing by evolutionary psychologists that dominant and powerful men pursue especially
attractive female mates (see Kenrick, Trost, & Sheets, 1996).
Grandiose Narcissistic Traits and the Appraisal of Initial Attraction
Participants with higher levels of Grandiose Exhibitionism reported being more likely to
pursue their romantic relationships for hedonistic reasons and to enhance their self-image. This
suggests that grandiose individuals, at least to some extent, construe romantic relationships as an
avenue to increase positive feelings and sentiments towards the self. This is consistent with other
work that has found vanity and exhibitionism to be moderately positively associated with
strivings for self-presentation (Emmons, 1989). This conclusion needs to be tempered by the
fact, however, that the same results were not obtained for Narcissistic Grandiosity as assessed by
the PNI. Although both measures assess maladaptive self-enhancement strategies, the NPI
Grandiose Exhibitionism subscale appears to capture features that are more explicitly concerned
with the admiration of others—namely, exhibitionism and vanity. As such, these features are
likely to be more closely connected to the adoption of motivations that emphasize the need for
self-enhancement or excitement.
Interestingly, the grandiose narcissistic traits were both negatively related to attachmentrelated avoidance. Therefore, people who possessed higher levels of grandiose traits tended to be
more comfortable with intimacy and closeness. This was true across both sexes for Grandiose
Exhibitionism and it was true in men for Narcissistic Grandiosity. Although I had originally
expected that grandiose narcissism would be positively related to attachment-related avoidance,
the current findings are actually consistent with previous research. Indeed, Dickinson and Pincus

55

(2003) found that the majority of the participants that they designated as grandiose narcissistic
subtypes reported possessing secure adult attachment styles.
The present study further hypothesized that grandiose narcissistic traits would be
positively associated with attraction towards admiring targets. Results from Study 1 provided
some support for this hypothesis: Women with higher levels of Grandiose Exhibitionism
reported being more attracted to the targets that were described as admiring. Thus, admiration in
a potential romantic partner may represent an important resource for women who possess greater
levels of vanity and exhibitionism. Unlike Grandiose Exhibitionism, Narcissistic Grandiosity
was unrelated to attraction towards the admiring target.
Although not hypothesized, one of the consistent findings to emerge from Study 1 is that
participants with higher levels of grandiose narcissistic traits are more attracted to targets that are
described to be desirable. Indeed, both NPI Grandiose Exhibitionism and PNI Narcissistic
Grandiosity positively predicted attraction towards the desirable target. This finding is consistent
with research by Campbell (1999) that found that narcissists use identification as one means to
self-enhance in romantic relationships. The primary caveat to this conclusion is that this means
of self-enhancement appears to be more appropriately reserved for grandiose narcissistic traits as
opposed to vulnerable narcissistic traits.
Vulnerable Narcissistic Traits and the Appraisal of Initial Attraction
The hypothesis that vulnerable narcissistic traits would be positively related to the
endorsement of self-protective relationship initiation motivations was supported in the present
research. Participants with higher levels of NPI Entitlement/Exploitativeness and PNI
Narcissistic Vulnerability reported more frequently beginning their romantic relationships to
cope with negative emotions and to bolster an inadequate sense of self. Much previous research

56

has documented associations between narcissistic traits related to entitlement and
exploitativeness and increased negative affect (Emmons, 1987; Watson, Little, Sawrie, &
Biderman, 1992; Watson & Morris, 1991) and unstable self-views (Ackerman et al., 2011;
Brown et al., 2009; Watson et al., 1992). Likewise, research by Pincus and his colleagues (2009)
has shown that facets belonging to the PNI Narcissistic Vulnerability factor are associated with
lower levels of self-esteem and increased shame. Based on these findings, it seems reasonable
that an important criterion for vulnerable individuals in initiating a romantic relationship is the
possibility of ameliorating negative emotions and coping with negative self-views.
Although not hypothesized, there was evidence that participants with higher levels of
vulnerable narcissistic traits also appear to adopt self-enhancing relationship motivations. Indeed,
both NPI Entitlement/Exploitativeness and PNI Narcissistic Vulnerability positively predicted
the Hedonism and Enhancement relationship initiation motives. These findings suggest that
people with higher levels of vulnerable narcissistic traits are also concerned with using their
relationships to enhance their self-image and pursue excitement.
This research further predicted that participants with higher levels of vulnerable
narcissistic traits would possess diminished intimacy motivations. The results showed that the
vulnerable narcissistic traits did indeed exhibit significant relations with the intimacy
relationship initiation motive variable; however, the two operationalizations of vulnerable
narcissism exhibited divergent relations. Men with higher levels of NPI
Entitlement/Exploitativeness reported being less likely to pursue their relationships to develop
emotional intimacy with their partners. Thus, at least for men with higher levels of this
normative expression of vulnerable narcissism, emotional intimacy represents an atypical motive
for initiating relationships. This would be consistent with arguments made by Campbell (1999)

57

that intimacy represents a detriment to self-enhancement for narcissists. Indeed, increased
intimacy may exacerbate vulnerability and risk further injury to an already brittle sense of self.
On the other hand, participants with higher levels of PNI Narcissistic Vulnerability
reported pursuing their romantic relationships more often to develop emotional intimacy. To
explore why this might be, I examined partial correlations between the facets of Narcissistic
Vulnerability and the intimacy relationship initiation motive variable after controlling for the
effects of Narcissistic Grandiosity and gender. Whereas the intimacy motive was unrelated to
Entitlement-Rage (r = .02, p = .700), Devaluing (r = .05, p = .323), and Hiding the Self (r = .07,
p = .174), it was significantly related to Contingent Self-Esteem (r = .14, p = .011). This suggests
that intimacy represents a more important motivation for beginning romantic relationships for
individuals whose self-esteem is labile and contingent upon external sources.
As an additional test of whether the vulnerable narcissistic dimensions would be
associated with diminished intimacy motivations, I examined connections between the
vulnerable narcissistic traits and attachment-related avoidance. The vulnerable narcissistic traits
were positively related to attachment-related avoidance, indicating that people who possess
greater levels of Entitlement/Exploitativeness and Narcissistic Vulnerability experience greater
discomfort with intimacy in general. This finding is consistent with research by Dickinson and
Pincus (2003) that has found that vulnerable narcissistic subtypes more frequently report
possessing a fearful avoidant adult attachment style.
When investigating the effects of decreasing participants’ levels of self-esteem on
romantic attraction, I found that the relation between NPI Entitlement/Exploitativeness and
attraction towards the desirable target was moderated by the self-esteem manipulation.
Specifically, people with higher levels of NPI Entitlement/Exploitativeness reported being less

58

attracted to desirable partners when their self-esteem levels were lowered. Given their higher
self-reported levels of vulnerability and depression (Ackerman & Donnellan, 2011), participants
with higher levels of Entitlement/Exploitativeness may have viewed the desirable targets as a
potential threat to their sense of self-worth. This therefore hints at the participants’ use of
devaluation as a strategy to undermine this threat. Another potential explanation is that, in
general, threatened people will denigrate people of higher status (see Crocker, Thompson,
McGraw, & Ingerman, 1987).
Related to this, participants with higher levels of Narcissistic Vulnerability reported being
more attracted to the caring target when they experienced a temporary decrease in self-esteem.
Although inconsistent with my predictions, this result is congruent with the previous finding that
Narcissistic Vulnerability is positively related to the intimacy relationship initiation motive.
Taken together, the results suggest that people with higher levels of Narcissistic Vulnerability are
more likely to seek out partners that offer the potential for intimacy when they feel wounded.
Conclusions
Broadly speaking, many of the outcomes investigated in this research can be considered
either self-focused (e.g., pursuing a new romantic relationship to reduce one’s own negative
affect) or other-focused (e.g., pursuing a new romantic relationship to become close to the other
person). Furthermore, the self-focused outcomes can be thought to reflect either self-enhancing
or self-protecting opportunities. Whereas self-enhancing opportunities represent the potential to
increase positive sentiments towards the self, self-protective opportunities represent the potential
to decrease negative sentiments towards the self.
The present research found that the pattern of correlates for the normal, grandiose, and
vulnerable narcissistic traits were all fairly distinct. Normal narcissism was generally associated

59

with outcomes reflecting a diminished self-focus. The grandiose narcissistic traits, in contrast,
were more consistently associated with outcomes that reflected self-enhancement. Further, the
vulnerable narcissistic traits were characterized by outcomes that reflected a mixture of selfenhancement and self-protection.
Several significant gender differences emerged across the analyses. It is important to
note, however, that none of these gender differences were predicted a priori on the basis of
strong theory. Moreover, the pattern of gender differences was not consistent enough to lend
itself to interpretation. Indeed, given the large number of tests for gender moderation and the few
significant differences that emerged, their occurrence is most likely due to chance. Future
research will be needed to replicate these gender differences and therefore determine whether
they are robust.
An important concern with the Relationship Initiation Motivations measure is that I did
not ask participants how many times they had actually attempted to begin a romantic
relationship. Participants were primarily in their late teens and early twenties, and it is possible
that some of them may not have had much experience in initiating romantic relationships. If this
is true, then the measure may say less about their recollected experiences and more about their
general intentions for beginning romantic relationships in the future. It will be important to
replicate these results in future research with participants that have verified having multiple
experiences with initiating romantic relationships.
Another potential limitation is that the romantic targets in the vignettes were artificial and
unrealistic. Indeed, it is unlikely that many people will encounter replicas of these particular
romantic targets that were described. It will therefore be necessary to make sure that these
findings apply to naturally occurring situations so as to ensure the generalizability of the results.

60

It will also be important use a more natural paradigm that permits the investigation of strategic
self-presentation and the development of rapport in initial interactions.

61

CHAPTER 6: INTRODUCTION TO STUDY 2
The goal of Study 2 was to further illuminate the roles of normal, grandiose, and
vulnerable narcissistic traits in romantic relationship initiation. Unlike Study 1, Study 2
examined the roles of these traits in actual dyadic interactions using a paradigm similar to speeddating. Participants were asked to engage in one-on-one interactions with several members of the
opposite sex and subsequently complete questionnaires (i.e., dating records) regarding the
interaction and their date (i.e., interaction partner). Such a paradigm allows the possibility of
addressing additional questions regarding self-presentation and the generation of rapport in
relationship initiation. To be able to address these questions effectively, however, requires a
good understanding of the data that are generated by participants in these sessions. Indeed, the
perceptions that participants have of their interactions and their dates may reflect several
different sources of influence. Although unique features of the interaction between the two
participants may contribute to the perception, it is also possible that the rating is influenced by
general characteristics of the person making the rating or general characteristics of the date that
is being rated.
The Social Relations Model
The Social Relations Model (SRM; Kenny, Kashy, & Cook, 2006) is a theoretical model
of dyadic behavior and a methodological tool that can illuminate the extent to which
interpersonal perceptions are driven by features of the perceiver, the target, or something unique
about the particular relationship between the perceiver and the target. As an example, let us
imagine that Alfredo has a date with Cecilia and that he afterwards rates how attracted he is to
Cecilia. One potential source contributing to this rating is Alfredo’s general tendency to be
attracted to all women. In SRM terminology, this is called Alfredo’s perceiver effect. It

62

represents the consistency of a person’s rating across all partners. Another potential source
contributing to this rating is Cecilia’s general tendency to be rated as attractive by all men. This
is termed the target effect in the SRM. It reflects the degree to which a person is rated similarly
by her or his partners. The final potential source contributing to Alfredo’s rating of attraction to
Cecilia is Alfredo’s unique level of attraction to Cecilia, over and above Alfredo’s general
tendency to be attracted to all women and over and above Cecilia’s general tendency to be seen
as attractive by all men. In the SRM, this is labeled the relationship effect. It reflects the unique
adjustment that one person makes to another after accounting for the perceiver and target effects.
The SRM estimates the variance in these effects for each group (i.e., dating session) and
then pools these estimates across groups. Perceiver variance assesses the extent to which people
differ in their general tendencies to provide high or low ratings across dates. Significant
perceiver variance in attraction for men would indicate that some men report being more
attracted to all of their dates whereas other men report being less attracted to all of their dates.
Target variance captures consensus among participants in their ratings of a particular date.
Significant target variance in attraction for women would indicate that some women are seen as
more attractive by all of their male partners whereas other women are seen as less attractive by
all of their male partners. Finally, substantial variance in the relationship effects would indicate
that attraction has a fundamentally dyadic component and is to some extent specific to the two
particular individuals interacting. Note that unless there are multiple indicators of the dating
record variable, the relationship variance includes error variance.
In addition to deriving estimates of variance in the perceiver, target, and relationship
(plus error) effects, the SRM also routinely generates estimates of reciprocity. Generalized
reciprocity refers to the association between individuals’ perceiver effects and their target

63

effects. In the case of attraction, a significant and negative generalized reciprocity coefficient for
men would indicate that men who report being more attracted to all of their dates tend to be rated
by those female dates as being less attractive. Dyadic reciprocity refers to the association
between a man’s relationship effect with a particular woman and that woman’s relationship
effect with the man. Thus, a positive and significant dyadic reciprocity coefficient for attraction
would indicate that a man who reports being especially attracted to a particular female date tends
to be seen as especially attractive by that date.
Narcissism and the SRM Effects
An SRM analysis of the dating record variables in Study 2 can provide important
information for the subsequent analyses. From a descriptive standpoint, the relative percentage of
variance in the dating record variables accounted for by the perceiver, target, and relationship
effects can shed insight on how much the variables reflect individual differences versus
relationship-specific phenomenon. In other words, it is possible to examine whether variability in
attraction ratings is more so a function of general characteristics of the two individuals involved
or something specific to their particular relationship. In addition, examination of the SRM
variances has practical implications for the analyses. If enough variance exists in the SRM
effects, it is possible to correlate individual-difference variables (e.g., narcissism) with them.
Correlations between the narcissism variables and the SRM perceiver effects for the
dating record variables would provide information on whether participants’ levels of narcissism
are associated with their general tendencies to perceive their dates in a particular way. For
instance, a negative correlation between Entitlement/Exploitativeness and the SRM perceiver
effect for attraction would indicate that participants with higher levels of
Entitlement/Exploitativeness consistently report being less attracted to their dates.

64

Correlations between the narcissism variables and the SRM target effects for the dating
record variables, on the other hand, would provide information on whether people who are
higher in narcissism are seen by members of the opposite-sex in a certain way. A positive
correlation between Grandiose Exhibitionism and the SRM target effect for attraction, for
example, would indicate that participants with higher levels of Grandiose Exhibitionism are
generally seen as more attractive.
Hypotheses Regarding the Role of Narcissism in Strategic Self-Presentation
As noted previously, because Study 2 involves actual interactions, it allows me to
evaluate the types of self-presentation that individuals with the different narcissistic traits use in
their initial encounters. In addition, I can assess whether certain narcissistic traits are perceived
to be more desirable than others. Last, I can investigate whether the type of self-presentation that
individuals engage in depends upon the impression that they have of their date. Below I present
my hypotheses and my plans for the analyses.
How do narcissists present themselves during the interactions? Given that past
research has revealed a notable connection between certain narcissistic traits related to
grandiosity and flashy styles of dress, charming facial expressions, and self-assured body
postures (Back et al., 2010), I hypothesize that individuals with higher levels of grandiose
narcissistic traits will elicit impressions of greater dominance and gregariousness from their
dates. I also predict that grandiose narcissistic traits will be associated with impressions of
increased arrogance. On the other hand, I expect that more vulnerable narcissistic traits will
generally elicit impressions of decreased warmth and dominance.
As a preliminary step in evaluating these hypotheses, I will examine the zero-order
correlations between the narcissism variables and the SRM target effects for the interpersonal

65

circumplex variables. Because I am primarily interested in the independent effects of grandiosity
and vulnerability, the primary analyses will use multilevel regression with these variables as
predictors. Note that separate analyses will be performed with the NPI traits as predictors and the
PNI traits as predictors.
Are narcissists considered more desirable by their dates? Because past research has
found links between grandiose narcissistic traits and popularity (Back et al., 2010), I hypothesize
that individuals with greater levels of grandiose narcissistic traits will be perceived to be more
desirable (i.e., good looking, charming, and popular) and attractive by their dates. In addition, I
predict that individuals with higher levels of vulnerable narcissistic traits will be perceived by
their dates as less admiring towards them and less attractive.
To evaluate these hypotheses, I will first examine the zero-order correlations between the
narcissism variables and the SRM target effects for these “appeal” variables (i.e., the dates’
perceptions of the participants’ admiration towards the dates, the dates’ perceptions of the
participants’ desirability, and the dates’ attraction towards the participants). The primary
analyses will then use multilevel regression to obtain estimates of the partial relations between
the participants’ levels of grandiose and vulnerable narcissism and their dates’ perceptions of
their appeal. As before, separate analyses will be performed with the NPI traits as predictors and
the PNI traits as predictors.
Are narcissists’ self-presentational strategies dependent upon their impressions of
their dates? It is possible that there are situations in which the self-presentational strategies used
by vulnerable individuals actually work to garner more positive impressions from their dates. For
example, if such individuals perceive a date to be desirable, they may behave in ways that are
more ingratiating and reflect a positive form of submission. In this case, individuals with higher

66

levels of vulnerable traits may be seen as less dominant but also more warm and admiring. I
therefore hypothesize that the relation between participants’ vulnerable traits and their dates’
perceptions of their warmth-agreeableness and admiration will be moderated by the participants’
perceptions of the date being desirable. Note that I do not anticipate such effects for the normal
and grandiose narcissistic traits.
To evaluate these hypotheses, I will perform a series of moderated multilevel regression
analyses in which I assess whether the relations between the participants’ scores on the
vulnerable narcissism variables and their dates’ perceptions of their appeal are moderated by the
participants’ ratings of their dates’ levels of desirability. Two outcomes will be evaluated: the
dates’ perceptions of the participants’ warmth-agreeableness and the dates’ perceptions of the
participants’ admiration towards the date. Again, separate analyses will be performed with the
NPI traits as predictors and the PNI traits as predictors.
Hypotheses Regarding the Processes Linking Narcissism to Attractiveness
One of the other purposes of Study 2 is to examine how each of the narcissistic traits is
related to ratings of attraction in a speed-dating context. Back et al. (2010) were able to show
how particular behavioral cues (e.g., self-assured body postures, charming facial expressions)
emitted by people with higher levels of certain narcissistic traits drive the relations between their
traits and their partners’ ratings of their popularity. In the present work, I am interested in
whether particular interpersonal perceptions mediate the relations between participants’
narcissistic traits and their dates’ attraction towards the participants.
There are three mediational hypotheses altogether. Based on the findings of Back et al.
(2010), I hypothesize that the positive relation between participants’ grandiose narcissism and
their dates’ attraction towards them will be mediated by the dates’ perceptions of their increased

67

levels of gregariousness, dominance, arrogance, and desirability. In addition, I hypothesize that
the negative relation between participants’ vulnerable narcissism and their dates’ attraction
towards them will be mediated by the dates’ perceptions of their decreased levels of dominance,
warmth, and admiration towards the date. Finally, in those cases where participants with higher
levels of vulnerable narcissistic traits perceive their dates to be more desirable, I hypothesize that
the positive relation between participants’ vulnerable narcissism and their dates’ attraction
towards them will be mediated by the dates’ perceptions of their increased levels of warmth and
admiration towards the date.
I will be using the causal steps approach to evaluating mediation proposed by Baron and
Kenny (1986). As will be described later in more detail, two of the requirements that must be
satisfied to be able to conclude that mediation has taken place are that: (a) the independent
variable(s) is significantly related to the outcome; and that (b) the independent variable(s) is
significantly related to the mediator. If, in the previous analyses, I find that participants’ levels of
grandiose and vulnerable narcissistic traits are not significantly related to their dates’ attraction
towards them, or that these traits are not significantly related to their dates’ perceptions of their
interpersonal behaviors or appeal-related behaviors (i.e., dominance, gregariousness, arrogance,
warmth, admiration, and/or desirability, depending upon the mediating variables hypothesized),
then mediational analyses will not be pursued for the first and/or second hypotheses. Likewise, if
I find that the relation between participants’ levels of vulnerable narcissistic traits and their
dates’ impressions of their admiration and warmth are not significantly moderated by the
participants’ perceptions of their dates’ desirability, then mediational analyses for the third
hypothesis will not be pursued.
Hypotheses Regarding the Role of Narcissism in the Building of Rapport

68

Another set of questions concerns whether certain narcissistic traits serve to shape the
rapport-relevant qualities of the interaction. For instance, do particular narcissistic traits lead
people to disclose less in their interactions, trust their dates less, and/or feel less connected to
their dates? Moreover, do dates tend to trust people with certain narcissistic traits less and/or
perceive them to be less responsive during the interactions? Below I detail my hypotheses
regarding these “rapport” variables and discuss the analytic strategies that I plan on pursuing.
Do narcissists disclose less to their dates, trust their dates less, and/or feel less
connected to their dates? Based on previous research (Emmons, 1989), I predict that the
grandiose and vulnerable narcissistic traits will be associated with lower self-reported rates of
disclosure. Similarly, I predict that the grandiose and vulnerable narcissistic traits will be
associated with lower self-reported levels of interpersonal connection. Though admittedly
speculative, it is possible that participants with higher levels of vulnerable traits will be less
likely to trust their interaction partners.
To evaluate connections between the grandiose and vulnerable narcissistic traits and selfreported rates of disclosure, connection, and trust, I will first examine the zero-order correlations
between the narcissism variables and the SRM perceiver effects for the disclosure, connection,
and trust variables. The primary analyses investigating these hypotheses will then use multilevel
regression to predict participants’ self-reported levels of disclosure, connection, and trust from
their own narcissism variables. Separate analyses will be performed using the NPI traits as
predictors and the PNI traits as predictors.
Are narcissists considered less responsive and trustworthy by their dates? Because
of their apparent insensitivity to people’s needs and their increased levels of disagreeableness
(Ackerman et al., 2011), I predict that participants with higher levels of grandiose and vulnerable

69

narcissistic traits will be perceived as less responsiveness by their dates. Moreover, albeit
speculative, I expect that participants with higher levels of grandiose and vulnerable narcissistic
traits will also elicit greater levels of skepticism from their partners.
To evaluate connections between the grandiose and vulnerable narcissistic traits and
informant-reported rates of responsiveness and trust, I will first examine the zero-order
correlations between the narcissism variables and the SRM target effects for the responsiveness
and trust variables. I will then again use multilevel regression analyses to predict the
participants’ dates’ reports of how much they trust the participants and their perceptions of the
participants’ responsiveness from the participants’ narcissism variables. Separate analyses will
be performed with the NPI traits as predictors and the PNI traits as predictors.
Hypotheses Regarding the Role of Narcissism in the Appraisal of Initial Attraction
Finally, Study 2 aimed to further evaluate whether the narcissistic traits are connected
with differential attraction towards particular romantic partners. As in Study 1, I hypothesize that
individuals with higher levels of grandiose narcissistic traits will be more attracted to dates
whom they perceive to be admiring during the interaction. Moreover, I hypothesize that
individuals with higher levels of vulnerable narcissistic traits will be more attracted to dates
whom they perceive to be desirable during the interaction. Last, I hypothesize that individuals
with higher levels of each of the narcissistic traits will be less attracted to dates whom they
perceive to be caring during the interaction.
To evaluate these hypotheses, I will perform a series of moderated multilevel regression
analyses in which the relation between participants’ levels of narcissism and their attraction
towards their dates is moderated by participants’ perceptions of their dates’ levels of admiration
towards the participants, their dates’ levels of desirability, or their dates’ levels of kindness.

70

Separate analyses will be performed with the NPI traits as predictors and the PNI traits as
predictors.

71

CHAPTER 7: METHOD FOR STUDY 2
Study 2 involved two phases. In the first phase that took place online, participants
completed the NPI and PNI. In the second phase, participants were asked to engage in fiveminute dates with several opposite-sex participants and subsequently complete dating records.

6

In the dating record, participants rated how attracted they were to their date (i.e., the Romantic
Attraction to Target Scale); how admiring, desirable, caring, dominant, submissive, cold-hearted,
arrogant, warm, and gregarious they perceived their date to be (i.e., the Partner Description
Inventory); and how much they disclosed to their dates, how responsive their dates were to them,
how close they felt to their dates, and how much they trusted their dates (i.e., the Encounter
Description Inventory).
The primary independent variables in Study 2 included the NPI traits of
Leadership/Authority, Grandiose Exhibitionism, and Entitlement/Exploitativeness, and the PNI
traits of Narcissistic Grandiosity and Narcissistic Vulnerability. The primary dependent variables
in Study 2 included participants’ levels of romantic attraction towards their dates; participants’
perceptions of their dates’ levels of admiration, desirability, dominance, submissiveness,
arrogance, cold-heartedness, warmth, and gregariousness; participants’ self-reported rates of
disclosure during the interaction, participants’ perceptions of their dates’ responsiveness during
the interaction, participants’ perceptions of connection to their dates, and participants’ degree of
trust in their dates.
Participants
All participants were recruited through the Human Participation Research system
maintained by the psychology department at Michigan State University and received course
6

All materials used in Study 2 are contained in Appendix 2.
72

credit or extra credit for their participation. The study description on the website requested
students who were not currently involved in a romantic relationship, who had lived in the
continental United States since they were five years old (so that language barriers would not
interfere with communication on the dates), and who were between 18 and 20 years old. A total
7

of 497 participants signed up for the first phase of the study. Of these participants, 238 of them
8

completed the second phase of the study. However, the online responses for six participants had
to be dropped because they did not respond affirmatively to the question, “Did you answer these
questions honestly?” on the online questionnaire. Thus, complete data from both phases of the
study were available from 232 participants overall.
Data from the second phase of the study came from 33 groups of 6 to 10 undergraduate
college students (i.e., 16 groups of 6, 13 groups of 8, and 4 groups of 10). Each group contained
equal numbers of men and women. Based on the data set of 232 participants, participants were
18 (41.3%) years old, 19 (34.3%) years old, 20 (23.5%) years old, and 21 (0.9%) years or older
(two participants did not report their age). Participants were predominantly White (Not
Hispanic/Not Latino) (82.3%) or Black or African American (11.7%); one participant did not
provide data on her/his ethnicity. Participants reported being Single or Not Dating (78.4%),
Dating casually (13.9%), Dating exclusively (7.4%), and Engaged (0.4%); one participant did
not provide data on her/his relationship status. All participants self-identified as heterosexual.
Materials

7

This number includes participants that signed up for Study 2 but ended up participating in
Study 1.
8
Two additional participants completed the in-lab portion of the study before they completed the
online portion. Because they were privy to the study hypotheses, their online data were not used.
73

Narcissistic Personality Inventory. As in Study 1, subscales for Leadership/Authority
(11 items; α = .78, average inter-item r = .24; M = 0.56, SD = 0.27), Grandiose Exhibitionism
(10 items; α = .69, average inter-item r = .18; M = 0.41, SD = 0.24), and
Entitlement/Exploitativeness (4 items; α = .28, average inter-item r = .09; M = 0.22, SD = 0.23)
were created according to the Ackerman et al. (2011) solution. Participants’ scores for each of
these subscales were the average number of narcissistic alternatives that they endorsed, with
higher scores reflecting higher levels of the trait. Note that men and women significantly differed
in their average levels of Leadership/Authority (men: M = 0.61, SD = 0.26; women: M = 0.50,
SD = 0.26; t(238) = 3.06, p = .002), Grandiose Exhibitionism (men: M = 0.45, SD = 0.23;
women: M = 0.36, SD = 0.23; t(238) = 2.81, p = .005), and Entitlement/Exploitativeness (men: M
= 0.26, SD = 0.24; women: M = 0.18, SD = 0.20; t(238) = 3.05, p = .003).
Pathological Narcissism Inventory. As in Study 1, scales were created for the
superordinate factors of Narcissistic Grandiosity (18 items; α = .84, average inter-item r = .22; M
= 2.93, SD = 0.67) and Narcissistic Vulnerability (34 items; α = .94, average inter-item r = .30;
M = 2.28, SD = 0.77). Participants’ scores were the mean of their responses for each scale, with
higher scores indicating higher levels of the trait. Note that although men and women
significantly differed in their average levels of Narcissistic Grandiosity (men: M = 3.10, SD =
0.62; women: M = 2.76, SD = 0.66; t(238) = 4.10, p < .001), they did not significantly differ in
their average levels of Narcissistic Vulnerability (men: M = 2.21, SD = 0.68; women: M = 2.35,
SD = 0.83; t(238) = -1.40, p = .164).
Romantic Attraction to Target. Participants’ degree of romantic attraction towards each
date with whom they interacted during the lab session was assessed with four questions: (a) “Do
you like this person?”; (b) “Are you attracted to this person?”; (c) “Would you like to date this

74

person?”; and (d) “Do you think this person is desirable?” Participants responded to each
question on a 7-point scale that ranged from 1 (Definitely no) to 7 (Definitely yes). Each
participant’s score for Romantic Attraction was the mean of her or his responses to these four
questions, with higher scores indicating greater levels of attraction (M = 3.90, SD = 1.34; α = .89,
average inter-item r = .66).
Encounter Description Inventory. This measure used and adapted items from Eastwick
and Finkel (2008), Finkel, Eastwick, and Matthews (2007), and Laurenceau et al. (1998) to
describe several aspects of the dyadic interactions during the lab session. These aspects included
Interpersonal Connection, Trust, Disclosure, and Perceived Responsiveness. Participants were
asked to provide their responses to the Interpersonal Connection and Trust items on 7-point
scales that ranged from 1 (Definitely no) to 7 (Definitely yes). Interpersonal Connection assesses
the degree to which participants felt an initial bond with their date (M = 3.49, SD = 1.28; α = .84,
average inter-item r = .64). It was measured with three items (e.g., “Did you feel a real
connection?” and “Were there sparks between you and your partner?”). Trust measured how
much participants were unsuspicious of their dates and was assessed with a single item (i.e.,
“Were you skeptical of your interaction partner?”; reverse scored; M = 5.45, SD = 1.51).
Participants provided responses to the Disclosure and Perceived Responsiveness items on
7-point scales that ranged from 1 (Very Little) to 7 (A Great Deal). Disclosure assessed
participants’ perceptions of their own rate of disclosure during the interaction and was measured
with a single item (i.e., “I disclosed…”; M = 4.10, SD = 1.40). Perceived Responsiveness
measured the degree to which participants perceived their dates to be accepting and responsive to
their disclosure. It was assessed with two items (i.e., “How much did your partner understand
you?” and “Did the other person see you as acceptable?”; M = 5.05, SD = 1.11; α = .77, inter-

75

item r = .64). Participants’ scores for the Interpersonal Connection and Perceived
Responsiveness scales were computed as the mean of their responses.
Partner Description Inventory. This 21-item measure consisted of a list of personality
characteristics and was used to assess how dates were perceived during the lab interactions.
Participants were asked to rate how much each characteristic was descriptive of their date on a 7point scale that ranged from 1 (Extremely Inaccurate) to 5 (Extremely Accurate). One part of this
inventory used some of the language reported in Campbell (1999) to assess how well
characteristics associated with being an admiring, desirable, and caring person described
participants’ dates. Thus, participants were asked to rate how much they believed their date was
Admiring (i.e., “Admires me,” “Interested in me,” and “Loves my company”; 3 items, M = 3.69,
SD = 1.13; α = .87, average inter-item r = .69), Desirable (i.e., “Good looking,” “Charming,” and
“Popular”; 3 items, M = 4.23, SD = 1.17; α = .80, average inter-item r = .57), and Caring (i.e.,
“Friendly,” “Caring,” and “Compassionate”; 3 items, M = 5.10, SD = 0.97; α = .82, average
inter-item r = .61).
A second part of this inventory included items from the Interpersonal Adjective Scale,
Revised (IAS-R; Wiggins, Trapnell, & Phillips, 1988) to assess particular octants of the
interpersonal circumplex (see Figure 5). To save time and lessen the participants’ burden, only
adjectives from those octants relevant for the hypotheses and/or that are necessary to create the
fundamental circumplex dimensions of agency and communion were used. Therefore, six octants
were assessed: Assured-Dominant (PA), Cold-Hearted (DE), Unassured-Submissive (HI),
Arrogant-Calculating (BC), Warm-Agreeable (LM), and Gregarious-Extraverted (NO). To
further save time, the two adjectives from each of these octants that were reported in Wiggins et
al. (1988) to exhibit the highest loadings on their corresponding octant were selected for use in

76

this study. Six scales were computed: Assured-Dominant (i.e., “Assertive” and “Dominant”; M =
3.65, SD = 1.25; α = .71, inter-item r = .55), Cold-Hearted (i.e., “Unsympathetic” and
“Warmthless”; M = 2.24, SD = 0.99; α = .54, inter-item r = .37), Unassured-Submissive (i.e.,
“Unaggressive” and “Shy”; M = 3.85, SD = 1.25; α = .40, inter-item r = .25), ArrogantCalculating (i.e., “Cunning” and “Sly”; M = 2.82, SD = 1.09; α = .57, inter-item r = .40), WarmAgreeable (i.e., “Gentlehearted” and “Tenderhearted”; M = 4.91, SD = 1.09; α = .83, inter-item r
= .71), and Gregarious-Extraverted (i.e., “Friendly” and “Outgoing”; M = 5.21, SD = 1.10; α =
.70, inter-item r = .56). Participants’ scores for each scale were the mean of their responses.
Procedure
Participants completed the NPI and the PNI online and then signed up for the in-lab
session. The in-lab sessions involved the participation of six to ten participants and took place in
a medium-sized room in the Psychology Building. To run a session, it was required that there
were an equal number of men and women and that there was at least three of each gender
present. In the event that not enough participants were available to conduct the session (or
alternatively, that there were extra participants), participants were offered the opportunity to
participate in Study 1 for credit.
The room was set up such that five small tables were evenly spread out along the
perimeter of the room and one longer table was situated in the center of the room. Each of the
tables was adorned with a red table cloth. As participants arrived to the study, they were each
given a bottle of water and were asked to remain quiet until the study began. Once all of the
participants arrived (or it was five minutes past when the study was scheduled to begin), the
experimenter asked participants if any of them knew each other. In the event that any men knew

77

any of the women or any women knew any of the men, one of them was randomly chosen to
participate in Study 1.
Once these preliminary tasks were completed, the experimenter provided participants
with informed consent. After this, the experimenter provided each participant with an
identification tag and a clipboard of dating records. The identification-tags for each participant
had different colors and unique numbers. The clipboard contained several dating records in
which the Romantic Attraction to Target Scale, the Encounter Description Inventory, and the
Partner Description Inventory were embedded.
At this point, the experimenter read the instructions to the participants and explained the
general procedure for the study (see Appendix 3 for the entire script). After explaining the study,
the experimenter asked all of the women (or men) to take a seat at one of the tables along the
perimeter of the room. The experimenter then randomly assigned each of the men (or women) to
take a seat with one of the women (or men). Note that the status of men or women sitting down
or rotating during the study was randomly determined prior to the actual session (see Finkel &
Eastwick, 2009). Once all of the participants had taken a seat, the experimenter told participants:
You will start your first date momentarily. I will warn you when you have one minute
left. I will then ring a bell to let you know to stop your date and to fill out the
questionnaire. I do have one request: Please stop your date within a second or two of
hearing the bell; this is the only way to keep the event on track. While you are actually
interacting with other people, we suggest that you put the clipboard on the floor so it does
not become a distraction. OK, please go ahead and start your first date.
After these five minutes transpired, the experimenter told the participants to stop the
interaction and asked the rotating gender to take a seat at the long table in the center. At this

78

time, participants were told to spend the next five minutes completing a dating record based on
the date with whom they just interacted. After all the participants had completed their dating
records, the rotating gender was asked to rotate clockwise to the next table along the room
perimeter. This cycle of rotating, having a date, and completing a dating record continued until
all the participants had interacted with all other participants of the opposite-sex. Once the session
was over, the experimenter collected all participants’ dating records. Participants were then
thanked and fully debriefed. Note that there was no implemented follow-up as would normally
occur in a real speed-dating event.

79

CHAPTER 8: RESULTS FOR STUDY 2
Preliminary Analyses of the Social Interaction Data
The hypotheses in Study 2 were evaluated using the dating record data obtained from the
speed-dating sessions. As noted in my discussion of the SRM, these ratings reflect multiple
sources of influence, including perceiver effects, target effects, and relationship effects. The
speed-dating design in which each individual interacts with and then rates multiple dates allows
me to tease apart these multiple sources of influence and thereby gain a better understanding of
the dating record variables used in the present research.
The SRM (Kenny et al., 2006) for an asymmetric block design was used to partition
variance in individuals’ social interaction ratings. A block design occurs when one subgroup of
participants (e.g., men) interacts one-on-one with another subgroup of participants (e.g.,
women); it is termed “asymmetric” when the two subgroups can be distinguished on some
characteristic (e.g., gender). Table 24 provides an illustration of the asymmetric block design
used in the present study.
Estimates of the SRM parameters (e.g., perceiver, target, and relationship variance;
reciprocity correlations) were obtained from the program, BLOCKO (Kenny, 1998). Because
BLOCKO does not permit missing data, missing data were mean-imputed (n = 8 for each of the
NPI subscales; n = 10 for PNI Narcissistic Grandiosity; n = 9 for PNI Narcissistic Vulnerability;
and n = 1 for disclosure). In those cases where significant gender differences were observed in
the variable means (i.e., each of the NPI subscales, PNI Narcissistic Grandiosity, and disclosure),
the missing value was imputed with the mean of the variable for the appropriate gender.
Application of the SRM to the dating record variables. Table 25 presents the SRM
variance partitioning for the dating record variables. This table includes the total amount of

80

variance for each variable as well as the relative percentage of variance in the dating record
variable accounted for by the perceiver, target, and relationship (plus error) effects. Note that
estimates of these values are provided for men and women separately. To facilitate comparison,
values for men’s ratings of women and women’s ratings of men are presented one right after the
other. Moreover, only the perceiver and target variances were tested for statistical significance
given that the relationship variances include error.
All of the perceiver variances for the dating record variables were statistically significant.
Participants in the speed-dating sessions were therefore at least somewhat consistent in their
responses across the different dates. For example, in Table 25, close to half of the variance in
participants’ self-reported degree of disclosure was due to their perceiver effects. Thus, some
participants reported disclosing a great deal to all of their dates, but others reported little such
disclosure. For the most part, the relative percentage of variance accounted for by the perceiver
effects for the dating record variables was similar for men and women.
Table 25 also shows that there was significant and substantial target variance for
participants’ ratings of attraction. This indicates that participants within a session tended to agree
with one other on which dates they found attractive. There was also substantial consensus among
participants on who they found to be desirable, dominant, submissive, and gregarious. To a lesser
extent, participants also agreed on who they found to be easy to connect with, who was
responsive to their needs, who was kind, and who was admiring. As Table 25 shows, not all of
the target variances for the dating record variables were statistically significant, thus indicating a
lack of consistency in participants’ ratings of a target for some variables. Indeed, there was little
consensus for cold-heartedness. Again, generally speaking, these relative percentages were
similar for men and women.

81

Table 26 displays the reciprocity correlations for the dating record variables. As Table 26
shows, there was no evidence of generalized reciprocity. Thus, participants’ general perceptions
of their dates were not significantly connected to their dates’ general perceptions of them. In
contrast, there was evidence of significant dyadic reciprocity for romantic attraction,
interpersonal connection, perceived responsiveness, perceived admiration, perceived desirability,
perceived kindness, and perceived warmth-agreeableness. Thus, if the man was especially
attracted to a particular woman, that woman also reported being especially attracted to the man.
This tendency for participants to reciprocate levels of attraction and connection shows that they
were actively engaged during the speed-dating sessions; thus, these findings illustrate the
experimental realism of the paradigm.
Conclusions from SRM analysis. Taken together, these results provide important
baseline information for the study. The findings regarding significant dyadic reciprocity for
variables such as attraction and connection, for instance, show that participants were taking the
task seriously. Moreover, because all of the dating record variables possess a substantial degree
of perceiver variance, it will be possible to examine associations between the narcissism
variables and the perceiver effects to determine whether narcissists see potential dating partners
in a similar way. Further, many of the dating record variables have significant target variance.
Thus, it will be possible to examine whether narcissists are seen in similar ways by potential
dating partners.
Unfortunately, it will not be possible to evaluate whether participants’ levels of the
narcissistic traits are associated with their dates’ perceptions of their cold-heartedness. Related to
this, there are some variables for which the relative percentage of variance accounted for by the
target effects is small for men, or women, or both (i.e., trust, arrogant-calculating, and warm-

82

agreeable). Thus, depending on the dating record variable, there may be some correlations that
cannot be computed due to insufficient variance in the target effects.
Plan of Analysis for Testing Hypotheses
The hypotheses for Study 2 concerned the role of narcissism in self-presentation, the
development of rapport, and the appraisal of initial attraction. As noted in the introduction to
Study 2, the analyses used to test these hypotheses involved some combination of zero-order
correlations and multilevel regression. The zero-order correlations between the narcissism
variables and the SRM perceiver and/or target effects were obtained from the program,
BLOCKO (Kenny, 1998) and are disattenuated for measurement error; moreover, separate
correlations were obtained for men and women and are presented side by side in the
corresponding tables. For each hypothesis that uses both zero-order correlations and multilevel
regression, I present the zero-order correlations first before proceeding to present the results for
the multilevel regression analyses.
Multilevel Modeling (MLM) was used to account for the hierarchical structure of the data
in the primary analyses. The random effects structure was specified to be consistent with the
SRM. Accordingly, separate perceiver and target variances were estimated for men and women.
Moreover, separate covariances between the perceiver and target effects for men and women
were specified to index their corresponding generalized reciprocity coefficients. The relationship
variances for men and women were estimated as residuals in the models and the covariance
estimated between them was used to index dyadic reciprocity. Although a random component for
the group was also initially included in the models, its inclusion was found to often produce nonpositive definite matrices; this was due to the minimal variance across the groups on the outcome
variables (this is sensible given that the groups were randomly formed and the levels of the

83

variables were more or less randomly distributed across members). As a result, the random
component for the group was removed from all the models. Unless otherwise noted, this was the
random structure that was specified across each of the subsequent analyses using MLM. Tests of
the hypotheses concerned the fixed effects in the models.
What Self-Presentational Strategies do Narcissists Generally Use on their Dates?
The first set of hypotheses concerned the self-presentational strategies that narcissists use
when initially interacting with a member of the opposite-sex in a romantic context. Due to the
circumplex structure of the variables examined for this particular set of hypotheses, these
analyses focused on what octant(s) of the interpersonal circle appeared to best describe the selfpresentational interpersonal style of narcissists. Indeed, if a narcissistic dimension is associated
with one octant of the circle, it will likely also be associated with other octants in close
proximity. Consequently, what is important here is the area of the interpersonal circle that is best
predicted by the narcissistic dimension.
Table 27 presents the zero-order correlations between the narcissism variables. As can be
seen, all of the NPI subscales were moderately correlated with each other, with the strongest
association occurring between Leadership/Authority and Grandiose Exhibitionism. Likewise, the
two PNI scales were moderately correlated. Whereas PNI Narcissistic Grandiosity was
moderately correlated with each of the NPI subscales, the PNI Narcissistic Vulnerability scale
was only modestly associated with the NPI Grandiose Exhibitionism and NPI
Entitlement/Exploitativeness subscales.
Do narcissists express the same interpersonal behaviors across their dates?: Zeroorder correlations with SRM target effects. Table 28 presents the zero-order correlations
between the narcissism variables and the SRM target effects for the interpersonal circumplex

84

measures. These correlations assess whether individuals with higher levels of narcissistic traits
9

tended to be seen in similar ways by their dating partners. For instance, are people with higher
levels of grandiose narcissistic traits seen as more arrogant?
As can be seen in Table 28, participants with higher levels of Leadership/Authority were
seen by their dates as more assured-dominant. Moreover, participants with higher levels of
grandiose narcissistic traits tended to be perceived as more assured-dominant. In addition, men
with higher levels of grandiose narcissistic traits were seen by their dates as more gregariousextraverted. With the exception of a negative correlation between NPI
Entitlement/Exploitativeness and unassured-submissiveness for women, the vulnerable
narcissistic traits were unrelated to perceptions of interpersonal behavior.
Do narcissists express the same interpersonal behaviors across their dates?:
Multilevel regression analyses. A series of MLM analyses were conducted to evaluate how
individuals with higher levels of the narcissism dimensions were perceived by their dates during
the speed-dating interactions. Each analysis included the following predictors: (a) the
participants’ gender (effect coded such that men = 1 and women = -1), the participants’ scores on
each of the narcissism scales (either all three NPI subscales or the two PNI scales), and the
interactions between the narcissism scales (either the NPI subscales or the PNI scales) and
participant gender to predict the participants’ dates’ perceptions of the participants’ interpersonal
behaviors. All continuous predictor variables were grand-mean centered prior to forming the
interaction terms. Five analyses were performed for each set of narcissism traits, one for each
9

Note that there was insufficient target variance to compute correlations between the narcissism
variables and the SRM target effects for cold-heartedness. Likewise, there was insufficient target
variance in the arrogant-calculating variable for women and insufficient target variance in the
warm-agreeable variable for men; therefore, correlations were not estimated with these particular
target effects.
85

interpersonal octant as an outcome (i.e., Assured-Dominant, Arrogant-Calculating, UnassuredSubmissive, Warm-Agreeable, and Gregarious-Extraverted).
Table 29 shows results for the MLM analyses using the three NPI traits to predict
participants’ dates’ perceptions of the participants’ interpersonal behaviors, with each column
displaying a separate analysis. Participants with higher levels of Leadership/Authority were seen
by their dates as more assured-dominant and more gregarious-extraverted. Table 29 shows,
however, that there was a significant interaction between Leadership/Authority and participant
gender for perceptions of gregarious-extraversion. To break down this interaction, simple slopes
for the effect of participants’ Leadership/Authority on their dates’ perceptions of gregariousextraversion were computed separately for men and women. The simple slope of
Leadership/Authority on gregarious-extraversion for women was b = 0.03, SE = .26, p = .914. In
contrast, the simple slope for men was b = 0.86, SE = .33, p = .01. Therefore, whereas men with
higher levels of Leadership/Authority were seen as more gregarious-extraverted and assureddominant in these social interactions, women with higher levels of Leadership/Authority were
only seen as more assured-dominant.
Consistent with predictions, participants with higher levels of Grandiose Exhibitionism
were seen by their dates as being more assured-dominant, more arrogant-calculating, more
gregarious-extraverted, and less unassured-submissive. However, in contrast to what was
hypothesized, Table 29 shows that Entitlement/Exploitativeness was not significantly related to
dates’ perceptions of any of the interpersonal traits.
Table 30 shows results for the MLM analyses using the two PNI traits to predict
participants’ dates’ perceptions of the participants’ interpersonal behaviors. As hypothesized,
participants with higher levels of Narcissistic Grandiosity were perceived by their dates to be

86

more gregarious-extraverted during the interactions. However, this effect was qualified by a
significant interaction between Narcissistic Grandiosity and participant gender. To break this
interaction down, simple slopes for the effect of Narcissistic Grandiosity on perceptions of
gregarious-extraversion were computed separately for men and women. The simple slope of
Narcissistic Grandiosity on perceptions of gregarious-extraversion for women was b = -0.02, SE
= .11, p = .859. The simple slope for men was b = 0.58, SE = .13, p < .001. Therefore, it appears
that only men with higher levels of Narcissistic Grandiosity were perceived by their dates to be
more gregarious-extraverted.
Although Narcissistic Grandiosity was not significantly related to dates’ perceptions of
any other interpersonal behaviors, there was a significant interaction between Narcissistic
Grandiosity and participant gender in the prediction of how warm-agreeable participants were
perceived to be during the interactions. Simple slopes of Narcissistic Grandiosity on perceptions
of warmth-agreeableness were again estimated separately for men and women. The simple slope
of Narcissistic Grandiosity for women was b = -0.06, SE = .09, p = .526. The simple slope for
men was b = 0.24, SE = .09, p = .009. Thus, only men with higher levels of Narcissistic
Grandiosity were perceived by their dates to be more warm-agreeable during the interactions.
Unlike Narcissistic Grandiosity, Narcissistic Vulnerability was not significantly related to
dates’ perceptions of any of the interpersonal behaviors. However, there was a marginal
tendency for participants with higher levels of Narcissistic Vulnerability to be perceived by their
dates as less gregarious-extraverted. Thus, at the very most, participants with higher levels of
Narcissistic Vulnerability were perceived to be slightly more aloof and introverted.
Summary of analyses using participants’ narcissism to predict their dates’
perceptions of their interpersonal behavior. The grandiose and vulnerable narcissistic traits

87

were predicted to be associated with different self-presentational strategies. Participants with
higher levels of grandiose narcissistic traits were hypothesized to behave in a more dominant,
gregarious, and arrogant manner. There was reasonable support for this hypothesis. However,
evidence did not suggest that Narcissistic Grandiosity is linked to interpersonal behaviors in the
arrogant-calculating and assured-dominant octants. In addition, participants with higher levels of
vulnerable narcissistic traits were hypothesized to behave in a more submissive and cold manner.
There was no support for this hypothesis.
How Appealing are Narcissists to their Dates?
In addition to predicting general interpersonal styles of presentation, this research
evaluated whether individuals with higher levels of particular narcissistic dimensions are
perceived to be more or less attractive during the interactions. Furthermore, this research
evaluated whether the narcissistic dimensions are differentially associated with perceptions of
admiration and desirability from dates.
Are narcissists considered appealing by their dates?: Zero-order correlations
between the narcissism variables and SRM target effects. Table 31 presents the zero-order
correlations between the narcissism variables and the SRM target effects for the appeal variables.
These correlations assess whether participants’ narcissistic traits are associated with a tendency
for them to be perceived by their dates as more appealing.

10

For instance, are people with higher

levels of grandiose narcissistic traits considered more attractive by all of their dates?

10

Because of insufficient variance in the SRM target effects for the disclosure and trust
variables in women, correlations were not estimated between women’s scores on the narcissism
measures and their SRM target effects for these variables. Likewise, because of insufficient
variance in the SRM target effect for the admiring variable in men, correlations were not
estimated between men’s scores on the narcissism measures and this variable.
88

As seen in Table 31, men with higher levels of Leadership/Authority were perceived by
their dates to be more desirable and attractive. Moreover, participants with higher levels of the
grandiose narcissistic traits tended to be perceived as more desirable and attractive as well.
However, the vulnerable narcissistic traits were unrelated to the SRM target effects for the
appeal variables.
Are narcissists considered appealing by their dates?: Multilevel regression analyses.
Multilevel regression analyses using the same set of predictors from the previous section were
pursued. The outcomes in these analyses included: (a) how much the dates perceived the
participants admired them; (b) how much the dates perceived the participants to be desirable; and
(c) how attracted the dates were to the participants.
Table 32 shows the results for the MLM analyses using the NPI traits as predictors. In
contrast to what was hypothesized, participants with greater levels of
Entitlement/Exploitativeness were not perceived by their dates to be less admiring during the
interactions. Instead, as Table 32 shows, there was a significant interaction between
Leadership/Authority and participant gender on perceived admiration, as well as a marginally
significant interaction between Grandiose Exhibitionism and participant gender on perceived
admiration. To break these interactions down, simple slopes for the effects of
Leadership/Authority and Grandiose Exhibitionism on perceived admiration were estimated
separately for men and women.
The simple slope of Leadership/Authority on perceived admiration for women was b = 0.24, SE = .19, p = .211; the simple slope of Leadership/Authority on perceived admiration for
men was b = 0.39, SE = .22, p = .084. Moreover, the simple slope of Grandiose Exhibitionism on
perceived admiration for women was b = -0.25, SE = .22, p = .256, and the simple slope of

89

Grandiose Exhibitionism on perceived admiration for men was b = 0.38, SE = .23, p = .11. The
simple slopes suggest that whereas women with higher levels of Leadership/Authority and
Grandiose Exhibitionism tended to be perceived by their dates as less admiring, men with higher
levels of Leadership/Authority and Grandiose Exhibitionism were perceived by their dates as
more admiring. Because none of the slopes were significantly different from zero, however,
these interpretations should be treated with caution.
Consistent with predictions, Table 32 shows that participants with higher levels of
Grandiose Exhibitionism were perceived by their dates to be more desirable. Moreover, there
was a marginally significant interaction between Leadership/Authority and participant gender in
the prediction of participants’ dates’ perceptions of the participants’ desirability. To break this
interaction down, simple slopes for the effect of Leadership/Authority on perceived desirability
were computed separately for men and women. The simple slope of Leadership/Authority for
women was b = -0.07, SE = .30, p = .815. The simple slope for men was b = 0.85, SE = .37, p =
.024. Thus, it appears that only men with higher levels of Leadership/Authority were perceived
by their dates to be more desirable.
Also in line with expectations, Table 32 shows that participants with higher levels of
Grandiose Exhibitionism were considered to be more attractive by their dates. However,
Entitlement/Exploitativeness was unrelated to ratings of attractiveness.
Table 33 shows results for the MLM analyses with the PNI traits as predictors. As shown,
neither Narcissistic Grandiosity nor Narcissistic Vulnerability significantly predicted dates’
perceptions of participants’ admiration towards the dates. However, there was a significant
interaction between Narcissistic Grandiosity and gender in the prediction of dates’ perceptions of
participants’ admiration. The simple slope of Narcissistic Grandiosity for women was b = -0.14,

90

SE = .08, p = .083. The simple slope for men was b = 0.11, SE = .08, p = .204. These slope
coefficients suggest that whereas men with higher levels of Narcissistic Grandiosity tended to be
perceived as more admiring by their dates, women with higher levels of Narcissistic Grandiosity
were perceived to be less admiring. Note that although these effects trended in opposite
directions, neither slope was significantly different from zero.
As hypothesized, participants with higher levels of Narcissistic Grandiosity were
perceived by their dates to be more desirable. However, this effect was qualified by a significant
interaction between Narcissistic Grandiosity and participant gender. The simple slope for women
was b = 0.06, SE = .13, p = .656. In contrast, the simple slope for men was b = 0.51, SE = .14, p
< .001. Thus, only men with higher levels of Narcissistic Grandiosity were perceived by their
dates to be more desirable.
Also consistent with hypotheses, Table 33 shows that participants with higher levels of
Narcissistic Grandiosity were considered to be more attractive by their dates. In contrast,
Narcissistic Vulnerability was unrelated to ratings of attractiveness.
Summary of analyses using narcissism to predict appeal. Participants with higher
levels of grandiose narcissistic traits were predicted to be perceived as more desirable and
attractive by their dates. These hypotheses were largely supported in the current study. In
addition, participants with higher levels of vulnerable narcissistic traits were predicted to be
perceived as less admiring by their dates and to elicit lower ratings of attraction. There was no
support for these hypotheses.
Are Narcissists’ Self-Presentational Strategies Conditional?
Another set of hypotheses in Study 2 focused on whether vulnerable narcissists’ use of
particular self-presentational strategies is dependent upon their impressions of their date. I

91

hypothesized that vulnerable individuals would be perceived by their dates to be more admiring
and more warm when the vulnerable individuals perceived their dates to be desirable. These
increased levels of admiration and warmth were then predicted to elicit higher ratings of
attraction from their dates.
A series of MLM analyses were conducted to evaluate whether the relations between
participants’ levels of vulnerable narcissistic traits and their dates’ perceptions of their
admiration and warmth-agreeableness are moderated by the degree to which participants
perceived their dates to be desirable. Note that the parameters of primary interest are the
interactions between the narcissism scales and the variable used to index participants’
perceptions of their dates’ levels of desirability (i.e., perceived desirability).
Table 34 shows the results of an analysis using the NPI traits to predict the participants’
dates’ perceptions of the participants’ warmth-agreeableness. As can be seen, the relations
between participants’ NPI traits and their dates’ perceptions of their warmth-agreeableness did
not depend upon how desirable participants perceived their dates to be.
Table 35 shows the results of an analysis using the NPI traits to predict the participants’
dates’ perceptions of the participants’ admiration towards the date. In contrast to what was
hypothesized, there was no significant interaction between Entitlement/Exploitativeness and
perceived desirability. Nevertheless, there was a significant two-way interaction between
Grandiose Exhibitionism and perceived desirability. To break this interaction down, simple
slopes for the effect of Grandiose Exhibitionism on dates’ perceptions of the participants’
admiration were computed at one standard deviation above and below the mean for perceived
desirability. The simple slope at low levels of perceived desirability was b = -0.12, SE = .23, p =
.608. The simple slope at high levels of perceived desirability was b = 0.61, SE = .22, p = .006.

92

This finding indicates that whereas participants with higher levels of Grandiose Exhibitionism
were perceived as more admiring by their dates when they perceived their dates to be more
desirable, participants’ levels of Grandiose Exhibitionism were unrelated to their dates’
perceptions of their admiration when they perceived their dates to be less desirable.
Table 36 shows the results of an analysis using the PNI traits to predict the participants’
dates’ perceptions of the participants’ warmth-agreeableness. There were no significant
interactions between the PNI dimensions and perceived desirability. However, there was a
marginally significant three-way interaction between Narcissistic Grandiosity, participant
gender, and perceived desirability. To break down this three-way interaction, simple slopes for
the two-way interactions between Narcissistic Grandiosity and perceived desirability were
computed separately for men and women. The simple slope for women was b = 0.06, SE = .05, p
= .27. The simple slope for men was b = -0.13, SE = .08, p = .108. Although the parameter
estimates for the two-way interaction terms for men and women marginally differed, neither was
significantly different from zero; thus, these interactions were not broken down further.
Table 37 shows the results of an analysis using the PNI traits to predict the participants’
dates’ perceptions of the participants’ admiration towards the date. As Table 37 shows, there
were no significant interactions between the PNI traits and perceived desirability.
Summary of analyses examining moderation of the relations between narcissism and
self-presentation. Broadly speaking, the hypothesis that more vulnerable individuals would be
perceived as more warm and admiring by their dates when they perceived their dates to be more
desirable was not supported. Because there was no evidence for this hypothesis, the proposed
additional analyses involving mediation were not pursued. Instead, the results revealed that

93

individuals with higher levels of Grandiose Exhibitionism were perceived by their dates to be
more admiring when they perceived their dates to be more desirable.
Why is Grandiose Narcissism Related to Enhanced Attraction?
This next set of analyses evaluated the mediational process hypothesized to account for
the relations between the grandiose narcissism dimensions and ratings of attraction. I expected
that participants’ levels of grandiose narcissism would lead their dates to perceive higher levels
of assured-dominance, gregarious-extraversion, arrogance-calculation, and desirability, and
thereby lead their dates to perceive them as more attractive. Before performing the mediational
analyses, zero-order correlations between the proposed mediators and outcome were first
inspected to evaluate whether the mediators were sufficiently distinct from the outcome and
whether the mediators were sufficiently distinct from one another.
Zero-order correlations between the proposed mediators and outcome. Table 38
presents the zero-order correlations between the variables. As would be expected given the
circumplex structure of the IAS-R, ratings of gregarious-extraversion, assured-dominance, and
arrogance-calculation were all correlated with one another. To prevent redundancy in the
analyses, I chose to focus on perceptions of gregarious-extraversion as a mediator given that
previous analyses revealed relations with this variable were common to both measures of
grandiose narcissism (see Tables 29 and 30).
Table 38 shows that dates’ ratings of participants’ levels of gregarious-extraversion and
desirability were both strongly associated with dates’ ratings of attraction to the participants.
Nevertheless, dates’ ratings of participants’ desirability were so strongly related to dates’ ratings
of attraction to the participants that a mediational analysis with these two variables would
provide little meaningful information. Accordingly, mediational analyses were only pursued with

94

gregarious-extraversion as a mediator of the relation between participants’ levels of grandiose
narcissism and their dates’ ratings of attraction.
Baron and Kenny (1986) steps for mediation. Baron and Kenny (1986) outlined a
causal steps approach to evaluating mediation that has been widely adopted in the social
sciences. According to these authors, a series of conditions must be met to claim that mediation
has taken place. First, the independent variable must significantly predict the outcome variable
(i.e., coefficient c in Panel A in Figure 6; the “total effect”). The logic behind this is that it does
not make sense to speak of mediation if there is not first a path to be mediated (MacKinnon,
2008). Second, the independent variable must significantly predict the mediating variable (i.e.,
coefficient a in Panel B in Figure 6). Third, the mediating variable must significantly predict the
outcome variable while controlling for the independent variable (i.e., coefficient b in Panel B in
Figure 6). Finally, for there to be complete mediation, the regression coefficient for the
independent variable obtained from regressing the outcome variable on the independent variable
and the mediating variable must be near zero (i.e., coefficient c′ in Panel B in Figure 6; the
“direct effect”).
In many cases, researchers are interested in the size of the indirect effect and whether it is
statistically different from zero. This is computed as either the difference in the total and direct
effects (i.e., c - c′) or the product of the a and b paths (i.e., ab). In the multiple regression
context, these two methods of computing the indirect effect are algebraically equivalent.
However, in the context of multilevel modeling, they are not, and methodologists report using
the ab product as the estimate of the indirect effect (MacKinnon, 2008).

95

The Sobel (1982) test is used to test the statistical significance of the indirect effect. The
standard error for the product of paths a and b is given by the following equation (MacKinnon,
2008; p. 52),
2 2

2 2

sab = square root of a sb + b sa ,
where sa is the standard error of the a path and sb is the standard error of the b path. The quotient
of these values (i.e., ab/sab) is approximately distributed as a standard normal score.
It should be noted that the topic of mediation is considerably more complex when
considered within the context of a MLM. This is because mediation can take place across
different levels (e.g., an upper-level independent variable predicting a lower-level mediator that
predicts a lower-level outcome) and the paths between lower-level variables (e.g., the path
connecting a lower-level mediator to a lower-level outcome) can have random effects.
Methodologists are just beginning to grapple with this complexity and offer solutions (see Bauer,
Preacher, & Gil, 2006; Kenny, Korchmaros, & Bolger, 2003; Krull & MacKinnon, 2001).
One solution proposed by Krull and MacKinnon (1999) involves the estimation of three
MLM equations that are analogous to those used in single-level mediation for non-hierarchical
data (see the equations presented in Figure 6). By estimating these three equations with a
properly specified random effects structure, researchers are able to derive appropriate standard
errors of the a and b parameters. These standard errors can then be used to derive an appropriate
standard error for the indirect effect using the Sobel (1982) test. Note that a restriction of this
method for multilevel mediation is that it assumes that the b parameter is fixed across the upperlevel units (see Bauer et al., 2006). Figure 7 shows the general mediational model that was
evaluated in the subsequent analyses.

96

Mediational hypothesis with NPI Grandiose Exhibitionism. To estimate path c, a MLM
was specified in which participants’ dates’ attraction to the participants was predicted to be a
function of the participants’ levels of Leadership/Authority, Grandiose Exhibitionism, and
Entitlement/Exploitativeness. Participants’ levels of Grandiose Exhibitionism were found to
significantly predict their dates’ levels of attraction towards them (b = 0.67, SE = .30, p = .026).
The first step in the Baron and Kenny (1986) causal steps approach was therefore satisfied.
To estimate path a, a MLM was specified in which participants’ dates’ perception of the
participants’ gregarious-extraversion was predicted to be a function of the participants’ levels of
Leadership/Authority, Grandiose Exhibitionism, and Entitlement/Exploitativeness. Participants’
levels of Grandiose Exhibitionism significantly predicted their dates’ perceptions of their
gregarious-extraversion (b = 0.46, SE = .23, p = .048). Consequently, the second step in the
Baron and Kenny (1986) causal steps approach was also satisfied.
To estimate paths b and c′, a MLM was specified in which participants’ dates’ attraction
to the participants was predicted to be a function of the participants’ levels of
Leadership/Authority, Grandiose Exhibitionism, and Entitlement/Exploitativeness, as well as
participants’ dates’ perceptions of the participants’ gregarious-extraversion. Path b in this model
was significant (b = 0.56, SE = .03, p < .001). This finding therefore satisfied the third step in
Baron and Kenny’s (1986) causal steps approach. In addition, path c′ was non-significant (b =
0.39, SE = .23, p = .094). According to Baron and Kenny (1986), there is consequently evidence
for mediation.
To test the significance of the indirect effect, the ab product term was first formed by
multiplying the values obtained from paths a and b (i.e., ab = 0.26). The standard error of the
mediated effect was derived using the formula for the Sobel test presented earlier (sab = 0.13).
97

The quotient of these values resulted in a z-obtained of 1.98. There was thus evidence that
participants’ dates’ perceptions of the participants’ gregarious-extraversion mediated the relation
between participants’ levels of Grandiose Exhibitionism and their dates’ attraction towards them.
Mediational hypothesis with PNI Narcissistic Grandiosity. To estimate path c, a MLM
was specified in which participants’ dates’ attraction to the participants was predicted to be a
function of the participants’ levels of Narcissistic Grandiosity and Narcissistic Vulnerability.
Participants’ levels of Narcissistic Grandiosity were found to significantly predict their dates’
levels of attraction towards them (b = 0.25, SE = .11, p = .018). The first step in the Baron and
Kenny (1986) causal steps approach was therefore satisfied.
To estimate path a, a MLM was specified in which participants’ dates’ perception of the
participants’ gregarious-extraversion was predicted to be a function of the participants’ levels of
Narcissistic Grandiosity and Narcissistic Vulnerability. Participants’ levels of Narcissistic
Grandiosity significantly predicted their dates’ perceptions of their gregarious-extraversion (b =
0.19, SE = .08, p = .022). Consequently, the second step in the Baron and Kenny (1986) causal
steps approach was also satisfied.
To estimate paths b and c′, a MLM was specified in which participants’ dates’ attraction
to the participants was predicted to be a function of the participants’ levels of Narcissistic
Grandiosity and Narcissistic Vulnerability, as well as the participants’ dates’ perceptions of the
participants’ gregarious-extraversion. Path b in this model was significant (b = 0.56, SE = .03, p
< .001). This therefore satisfied the third step in Baron and Kenny’s (1986) causal steps
approach. In addition, path c′ was non-significant (b = 0.15, SE = .08, p = .078). According to
Baron and Kenny (1986), there is consequently evidence for mediation.

98

To test the significance of the indirect effect, the ab product term was first formed by
multiplying the values obtained from paths a and b (i.e., ab = 0.11). The standard error of the
mediated effect was derived using the Sobel test (sab = 0.05). The quotient of these values
resulted in a z-obtained of 2.29. There was therefore evidence that participants’ dates’
perceptions of the participants’ gregarious-extraversion mediated the relation between
participants’ levels of Narcissistic Grandiosity and their dates’ attraction towards them.
Summary of mediational analyses. The hypothesis that the expression of interpersonal
behaviors related to gregarious-extraversion would mediate the connection between grandiose
narcissistic traits and ratings of attraction was supported in this research. Indeed, both measures
of grandiose narcissism provided converging evidence for this hypothesis.
How Does Narcissism Contribute to the Building Up of Rapport in the Social Interaction?
Data from the speed-dating interactions were used to further evaluate how the different
narcissistic dimensions contribute to rapport-relevant features of an initial romantic encounter.
The outcomes examined in this section include self-reported rates of disclosure, trust, and
interpersonal connection, and informant-reported levels of trust and responsiveness. Before
presenting the primary MLM analyses, zero-order correlations between the narcissism variables
and the SRM perceiver effects (for the disclosure, trust, and connection variables) and the SRM
target effects (for the trust and responsiveness variables) are examined.
Do narcissists disclose less to their dates, trust their dates less, and feel less
connected to their dates?: Zero-order correlations between narcissism variables and SRM
perceiver effects. Table 39 presents the zero-order correlations between the narcissism
variables and the SRM perceiver effects for the rapport variables. These correlations capture
whether the narcissism dimensions are associated with a generalized tendency to see interactions
99

as close and characterized by rapport. For example, do people with higher levels of vulnerable
narcissistic traits report feeling less connected to all of their dates?
As seen in Table 39, men with higher levels of Leadership/Authority were more skeptical
of their dates. Moreover, participants with higher levels of grandiose narcissistic traits tended to
also be more skeptical of their dates. Interestingly, whereas women with higher levels of
Entitlement/Exploitativeness trusted their dates more, women with higher levels of Narcissistic
Vulnerability were more skeptical of their dates. All of the narcissism variables were unrelated to
the SRM perceiver effects for disclosure and connection.
Are narcissists trusted less by their dates and considered less responsive?: Zeroorder correlations between narcissism variables and SRM target effects. Table 40 presents
the zero-order correlations between the narcissism variables and the SRM target effects for the
rapport variables. These correlations assess whether participants’ narcissistic traits are associated
with a tendency for their dates to report having interactions with them that are characterized by
more or less rapport.

11

For instance, are people with higher levels of grandiose narcissistic traits

considered less responsive by all of their dates?
As Table 40 shows, whereas men with higher levels of the grandiose narcissistic traits
were perceived by their dates to be more responsive, women with higher levels of the grandiose
narcissistic traits tended to be perceived as less responsive (though the associations for women
were non-significant). However, the normal and vulnerable narcissistic traits were unrelated to
the SRM target effects for the trust and responsiveness variables.

11

Because of insufficient variance in the SRM target effects for the trust variable in women,
correlations were not estimated between women’s scores on the narcissism measures and their
SRM target effect for this variable.
100

Does narcissism contribute to the development of rapport?: Multilevel regression
analyses. Table 41 displays the results for three MLM analyses in which the NPI traits were
used to predict self-reported levels of disclosure, trust, and connection within the social
interactions.

12

As can be seen, none of the NPI traits significantly predicted rates of disclosure

or connection. On the other hand, participants with higher levels of Leadership/Authority
reported being less trusting of their dates. In addition, there was a significant interaction between
Entitlement/Exploitativeness and participant gender in the prediction of trust.
To break down this significant interaction, simple slopes for the effect of
Entitlement/Exploitativeness on trust were computed separately for men and women. The simple
slope for women was b = 1.30, SE = .57, p = .025. The simple slope for men was b = -0.14, SE =
.42, p = .738. Thus, only women with higher levels of Entitlement/Exploitativeness reported
being more trusting of their dates.
Table 42 displays the results for two MLM analyses in which the NPI traits were used to
predict the dates’ trust in the participant and the dates’ perceptions of the participants’
responsiveness.

13

None of the participants’ NPI traits were significantly related to their dates’

levels of trust. However, there was a significant interaction between Grandiose Exhibitionism
and participant gender. The simple slope for women was b = -0.44, SE = .32, p = .171. The
simple slope for men was b = 0.96, SE = .36, p = .009. These results indicate that women
reported being more trusting of men with higher levels of Grandiose Exhibitionism.

12

The women’s target variance for trust was fixed to zero in the random effects statement due to
minimal variance in this parameter.
13
The women’s perceiver variance for dates’ trust in the participant was fixed to zero in the
random effects statement due to minimal variance in this parameter.
101

As Table 42 shows, none of participants’ NPI traits were significantly related to their
dates’ perceptions of their responsiveness. However, there was a significant interaction between
Grandiose Exhibitionism and participant gender. The simple slope for women was b = -0.50, SE
= .24, p = .037. The simple slope for men was b = 0.58, SE = .27, p = .035. These findings
suggest that whereas men perceived women with higher levels of Grandiose Exhibitionism to be
less responsive, women perceived men with higher levels of Grandiose Exhibitionism to be more
responsive.
Table 43 displays the results for three MLM analyses in which the PNI traits were used to
predict participants’ self-reported levels of disclosure, trust, and connection within the social
interactions.

14

Neither of the PNI traits predicted self-reported rates of disclosure or trust in the

interactions. Interestingly, however, participants with higher levels of Narcissistic Grandiosity
reported feeling more connected to their dates.
Table 44 presents the results for two MLM analyses in which the PNI traits were used to
predict the dates’ trust in the participant and the dates’ perception of the participants’
responsiveness.

15

As can be seen, neither PNI dimension was significantly related to

participants’ dates’ levels of trust in the participant or their dates’ perceptions of the participants’
responsiveness during the interaction.
Summary of analyses using narcissism to predict rapport. I hypothesized that the
grandiose and vulnerable narcissistic traits would be related to lower levels of self-reported
disclosure during the interactions. I also hypothesized that the grandiose and vulnerable

14

The women’s target variance for trust was fixed to zero in the random effects statement due to
minimal variance in this parameter.
15
The women’s perceiver variance for dates’ trust in the participant was fixed to zero in the
random effects statement due to minimal variance in this parameter.
102

narcissistic traits would be related to lower self-reported levels of interpersonal connection.
There was no support for these hypotheses. In fact, participants with higher levels of Narcissistic
Grandiosity reported feeling more connected to their dates.
In addition, I predicted that participants with higher levels of vulnerable narcissistic traits
would be less trusting of their dates. To the contrary, women with higher levels of
Entitlement/Exploitativeness reported having more trust in their dates. Related to this, I expected
that participants would be less trusting of dates who possessed greater levels of grandiose and
vulnerable narcissistic traits. In contrast, men with higher levels of Grandiose Exhibitionism
were trusted more by their dates. Last, I hypothesized that the grandiose and vulnerable
narcissistic traits would be associated with lower levels of informant-reported responsiveness.
On the one hand, women with higher levels of Grandiose Exhibitionism were perceived by their
dates to be less responsive. Men with higher levels of Grandiose Exhibitionism, however, were
perceived by their dates to be more responsive.
Who do Narcissists Find Attractive in the Social Interactions?
The final set of hypotheses evaluated whether the narcissism dimensions are associated
with differential attraction towards certain romantic partners. For this set of analyses, the
relations between participants’ scores on the narcissism variables and their ratings of attraction
towards their dates were evaluated for moderation by the participants’ ratings of their dates’
levels of admiration, desirability, and kindness. Separate analyses were performed for perceived
admiration as a moderator, perceived desirability as a moderator, and perceived kindness as a
moderator. Moreover, separate analyses were performed with the NPI subscales as predictors and
the PNI scales as predictors, and gender was included as a moderator of each of the effects. In

103

each case, the parameters of primary interest were the two-way interactions between the
narcissism variables and the date perception variable.
Are grandiose narcissists more attracted to admiring dates?: Multilevel regression
analyses. Table 45 shows the results of an MLM analysis using the NPI traits to predict
participants’ attraction towards their dates. There was no significant interaction between
Grandiose Exhibitionism and perceived admiration. Similarly, no significant two-way
interactions were observed between the other NPI traits and perceived admiration.
Table 46 shows the results from an MLM analysis using the PNI traits to predict
participants’ attraction towards their dates. In contrast to what was hypothesized, there was no
significant two-way interaction between Narcissistic Grandiosity and perceived admiration.
However, there was a significant two-way interaction between Narcissistic Vulnerability and
perceived admiration. To break down this interaction, simple slopes were computed for the effect
of Narcissistic Vulnerability on attraction at one standard deviation above and below the mean of
perceived admiration. The simple slope at low levels of perceived admiration was b = 0.15, SE =
.11, p = .169. The simple slope at high levels of perceived admiration was b = -0.17, SE = .10, p
= .093. Although there was a trend for participants with higher levels of Narcissistic
Vulnerability to be less attracted to dates that they perceived to be more admiring during the
interaction, this effect was not significant.
Are vulnerable narcissists more attracted to desirable dates?: Multilevel regression
analyses. Table 47 presents results from an MLM analysis using the NPI traits to predict
participants’ attraction towards their dates.

16

There were no significant two-way interactions

between the NPI traits and perceived desirability. However, there was a significant three-way
16

In this analysis, the men’s target variance for attraction was fixed to zero in the random effects
statement due to minimal variance in this parameter.
104

interaction between Leadership/Authority, perceived desirability, and gender. To break this
interaction down, the simple slope of the two-way interaction between Leadership/Authority and
perceived desirability was computed separately for men and women. The simple slope for men
was b = 0.19, SE = .17, p = .27. The simple slope for women was b = -0.24, SE = .12, p = .045.
This finding indicates that the relation between participants’ levels of Leadership/Authority and
their attraction towards their dates is only dependent upon how desirable they perceive their
dates to be for women.
To further break down the two-way interaction between Leadership/Authority and
perceived desirability for women, simple slopes for the effect of Leadership/Authority on
attraction towards the date were computed at one standard deviation above and below the mean
of perceived desirability. The simple slope at high levels of perceived desirability was b = -0.71,
SE = .26, p = .008. The simple slope at low levels of perceived desirability was b = -0.14, SE =
.25, p = .591. These results indicate that whereas women with higher levels of
Leadership/Authority report being less attracted to their dates when they perceive them to be
more desirable, women’s levels of Leadership/Authority are unrelated to their attraction towards
their dates when they perceive them to be less desirable.
Table 48 presents results from an MLM analysis using the PNI traits to predict
participants’ attraction towards their dates.

17

As can be seen, there were no significant two-way

interactions between the PNI traits and perceived desirability.
Are narcissists less attracted to caring dates?: Multilevel regression analyses. Table
49 shows the results of an MLM analysis using the NPI traits as predictors. As can be seen, there
were no significant two-way interactions between the NPI traits and perceived kindness.
17

Note that the men’s target variance for attraction was fixed to zero in the random effects
statement due to minimal variance in this parameter.
105

Nevertheless, there was a significant three-way interaction between Grandiose Exhibitionism,
gender, and perceived kindness. To break down this interaction, simple slopes for the two-way
interaction between Grandiose Exhibitionism and perceived kindness were computed separately
for men and women. The simple slope for women was b = -0.26, SE = .22, p = .24. The simple
slope for men was b = 0.86, SE = .34, p = .011. Thus, it is only for men that the relation between
Grandiose Exhibitionism and attraction towards their date is dependent upon how caring they
perceive their date to be during the interaction.
To further break down the significant two-way interaction between Grandiose
Exhibitionism and perceived kindness for men, simple slopes for the effect of Grandiose
Exhibitionism on attraction for men were computed at one standard deviation above and below
the mean for perceived kindness. The simple slope at low levels of perceived kindness was b = 1.24, SE = .48, p = .01. The simple slope at high levels of perceived kindness was b = 0.42, SE =
.44, p = .343. These findings indicate that whereas men with higher levels of Grandiose
Exhibitionism are less attracted to their dates when they perceive them to be less caring during
the interaction, men’s levels of Grandiose Exhibitionism are unrelated to their attraction towards
their dates when they perceive them to be more caring.
Table 50 shows the results of an MLM analysis using the PNI traits to predict
participants’ attraction towards their dates. As Table 50 shows, there were no significant twoway interactions between the PNI traits and perceived kindness. However, there was a significant
three-way interaction between Narcissistic Vulnerability, gender, and perceived kindness. To
break this interaction down, simple slopes for the two-way interaction between Narcissistic
Vulnerability and perceived kindness were computed separately for men and women. The simple
slope for women was b = 0.09, SE = .08, p = .231. The simple slope for men was b = -0.16, SE =

106

.10, p = .098. Therefore, although the interaction terms trended in opposite directions for men
and women, neither was significantly different from zero for either gender. These interactions
were consequently not broken down further.
Summary for analyses involving the narcissism dimensions and attraction towards
dates in social interactions. Individuals with higher levels of grandiose narcissistic traits were
predicted to be more attracted to dates who they perceived to be more admiring during the
interactions. The results did not provide any support for this hypothesis. In addition, participants
with higher levels of vulnerable narcissistic traits were hypothesized to be more attracted to dates
who they perceived to be more desirable during the interactions. There was no support for this
hypothesis. Instead, women with higher levels of Leadership/Authority were found to be less
attracted to their dates when they perceived their dates to be more desirable during the
interactions. Finally, participants with higher levels of each of the narcissistic traits were
expected to be less attracted to dates who they perceived to be more caring. Results did not
provide any support for this hypothesis. To the contrary, findings indicated that men who
perceived their dates to be less caring during the interactions reported even lower levels of
attraction towards these dates.

107

CHAPTER 9: DISCUSSION FOR STUDY 2
Study 2 sought to better understand the roles of the normal, grandiose, and vulnerable
narcissistic traits in strategic self-presentation (i.e., how narcissists are viewed by others) and the
development of rapport (i.e., how narcissists perceive their interactions and how narcissists’
dates perceive their interactions with the narcissists). It also aimed to further illuminate the roles
of these traits in the narcissists’ appraisals of initial attraction towards their dates. To accomplish
these aims, Study 2 used a speed-dating paradigm to investigate the manifestations of these traits
in a real-world context related to romantic relationship initiation. Below I present the main
findings for each of the narcissism dimensions and then offer some basic conclusions.
Normal Narcissism in the Speed-Dating Sessions
Participants with higher levels of Leadership/Authority were perceived by their dates to
express interpersonal behaviors characterized by increased levels of assured-dominance when
viewed through the lens of the interpersonal circumplex (Wiggins et al., 1988). These findings
are consistent with self-report data showing that people with higher levels of
Leadership/Authority report higher levels of surgency and dominance (Emmons, 1984). They are
also consistent with findings by Back et al. (2010) that people with higher levels of
Leadership/Authority exhibit more self-assured body movements during group interactions. In
general, these findings suggest that an important goal for people with greater levels of this trait—
even within initial encounters with a potential romantic partner—is to establish dominance.
Men with higher levels of Leadership/Authority were also perceived by their dates to be
more gregarious-extraverted. Research has shown that people with higher levels of
Leadership/Authority report being more extraverted (Ackerman et al., 2011; Corry et al., 2008;
Emmons, 1984) and also have informants that report them being more extraverted (Ackerman et

108

al., 2011; Hill & Roberts, in press). However, previous work has not generally uncovered gender
differences in the relation between Leadership/Authority and Extraversion. It is possible that the
present finding can be explained in terms of gender role violations. Indeed, the expression of
dominant traits in men within an explicitly romantic context (as opposed to the expression of
such traits in women within the same context) was probably viewed upon more favorably given
that such behaviors are consistent with the traditional masculine gender role.
Along with this, men with higher levels of Leadership/Authority were also perceived by
their dates to be more desirable. This finding contradicts work by Back et al. (2010), which
failed to find links between Leadership/Authority and ratings of likeability. Perhaps the more
explicit nature of the romantic context was responsible for the finding. Indeed, evolutionary
psychologists have long argued that women are more attracted to men with greater levels of
dominance (e.g., Havlicek, Roberts, & Flegr, 2005).
Interestingly, participants with higher levels of Leadership/Authority were more skeptical
of all of their dates. Past research has shown, however, that people with higher levels of
Leadership/Authority report possessing lower levels of Machiavellianism (Ackerman et al.,
2011). This finding was therefore unexpected and seems to run counter to the notion that people
with higher levels of this trait should be less cynical of their dates’ intentions. Nevertheless, this
finding may have emerged due to the particular item that was used to assess trust (i.e., were you
skeptical of your interaction partner?). Instead of assessing trust per se, this item may tap into a
person’s ability to detect the degree of self-presentation that is occurring in these brief social
encounters. In this case, it may be that participants with higher levels of Leadership/Authority
are better able to detect that their partners are trying to make a good impression. An important

109

next step in this research is to determine whether such participants are actually accurate in their
inferences about the underlying motivations of their dates.
Finally, women with higher levels of Leadership/Authority reported being less attracted
to dates that they perceived to be more desirable. Thus, women who report being more dominant
and having greater leadership potential find men who are more charming, popular, and good
looking to be less attractive. It is possible that such dates represent a potential threat to the
participants’ assertion of dominance. Even if this is true, however, it remains unclear why the
same finding did not emerge for men.
Grandiose Narcissism in the Speed-Dating Sessions
Participants with higher levels of Grandiose Exhibitionism were perceived by their dates
to display more agentic interpersonal behaviors. Specifically, they were seen as more arrogant,
dominant, and gregarious during the interactions. Such results are consistent with findings that
individuals with higher levels of Grandiose Exhibitionism tend to be more extraverted
(Ackerman et al., 2011; Hill & Roberts, in press).
Interestingly, men with higher levels of Narcissistic Grandiosity were perceived by their
dates to display relatively more communal interpersonal behaviors. They were seen as more
gregarious and warm during the interactions. These findings conflict with other research using
the HEXACO model of personality which has shown that Narcissistic Grandiosity is largely
unrelated to self-reports of Extraversion and Agreeableness (instead, Narcissistic Grandiosity is
negatively related to Honesty-Humility and positively related to Emotionality; Bresin & Gordon,
2011). The present findings suggest that, at least within the context of an initial interaction with a
potential romantic partner, men with higher levels of Narcissistic Grandiosity behave in a more
agreeable fashion.

110

Related to this, women tended to agree with each other that men with higher levels of
Narcissistic Grandiosity were more desirable. Such results are consistent with findings by Back
et al. (2010) that participants who are more self-absorbed and self-admiring exhibit more
charming facial expressions during group settings. However, Back et al. (2010) found that they
were not considered more popular, which is somewhat at odds with what was found here. In
addition, Back et al. (2010) did not report gender differences in this result.
Although not hypothesized, an interesting finding to emerge from Study 2 was that
participants with higher levels of Grandiose Exhibitionism were perceived by their dates to be
more admiring when the participants perceived that their dates were more desirable. Thus,
participants with higher levels of this trait more freely expressed their admiration for their dates
when they perceived them to be more charming, good looking, and popular. It therefore appears
that they were attempting to win over this “perfect” partner via admiration.
Another group of hypotheses in Study 2 concerned whether dates would be more
attracted to participants with higher levels of grandiose narcissistic traits due to the participants’
enhanced levels of self-assuredness, gregariousness, arrogance, and desirability. To prevent
redundancy and ensure that the mediational analyses were meaningful, I chose to focus on
gregarious-extraversion as a mediator. Results showed that the relation between participants’
grandiose narcissism measures and their dates’ attraction to them was indeed mediated by their
dates’ perceptions of their gregarious-extraversion.
The other hypotheses in Study 2 focused on the development of rapport within the speeddating interactions. Consistent with hypotheses, women with higher levels of Grandiose
Exhibitionism were perceived to be less responsive by their dates. This is congruent with work
that suggests that people with higher levels of Grandiose Exhibitionism tend to be oblivious to

111

the needs of others (Ackerman et al., 2011). However, men with higher levels of Grandiose
Exhibitionism were perceived to be more responsive by their dates. Similarly, women reported
being less skeptical of men with higher levels of Grandiose Exhibitionism. These findings are
difficult to explain in the context of past research, but appear to fit a general trend in the present
research where men with higher levels of grandiose and dominant traits are perceived more
positively than women with these traits.
It was originally expected that participants with higher levels of grandiose narcissistic
traits would report feeling lower levels of connection and attraction towards caring dates due to
their diminished intimacy motivations. In contrast to what was hypothesized, participants with
higher levels of Narcissistic Grandiosity actually reported feeling more connected to their caring
dates. Likewise, men with higher levels of Grandiose Exhibitionism were less attracted to their
dates when they perceived them to be less caring.
Vulnerable Narcissism in the Speed-Dating Sessions
None of the hypotheses regarding the role of vulnerable narcissistic traits in relationship
initiation were supported in the present study. This suggests the need for seriously reconsidering
how participants with greater levels of vulnerable narcissistic traits initiate their relationships
with romantic partners. It could be that the particular strategies adopted by such participants
were not well-captured by the instruments used in this research. Alternatively, such traits may
matter more for the later stages of relationship development. Apart from these possibilities,
however, there are a few findings worthy of mention.
When the other NPI traits were not controlled, participants with higher levels of
Entitlement/Exploitativeness were perceived by their dates to be less submissive. Thus, dates
tended to agree with one another that more entitled participants were not shy and unaggressive in

112

their interactions. Research using informants that have known the participants for longer periods
of time (e.g., roommates, mothers) has generally found that participants with higher levels of
Entitlement/Exploitativeness are seen by informants as less agreeable (Ackerman et al., 2011;
Hill & Roberts, in press).
Participants with higher levels of Narcissistic Vulnerability, on the other hand, were
perceived by their dates to be less gregarious. Interestingly, Pincus et al. (2009) found that
participants with higher levels of traits related to Narcissistic Vulnerability generally reported
experiencing more interpersonal problems characterized by distrust, difficulty with being
affectionate, difficulty with initiating interactions, and gullibility. The present research
corroborates the notion that more vulnerable individuals would experience difficulty with
initiating interactions as they were found to be more withdrawn during their dates.
I had originally hypothesized that the vulnerable narcissistic traits would be associated
with lower self-reported levels of trust in dates. In contrast, women with higher levels of
Entitlement/Exploitativeness reported being less skeptical of their dates. Again, this finding may
be due to how trust was measured in the present study. Indeed, it may simply mean that women
with higher levels of entitled beliefs were better detectors of their dates’ self-presentational
tendencies. Future research will be needed to determine whether a similar finding is replicated
with a more face-valid measure of trust.
Conclusions
Broadly speaking, each of the narcissistic traits demonstrated reasonably unique
associations with outcomes in the speed-dating interactions. Normal narcissism was associated
with the expression of dominant behaviors and greater levels of skepticism in dates. The
grandiose narcissistic traits were generally linked to the expression of dominance and

113

gregariousness; moreover, participants with greater levels of the grandiose traits were considered
more attractive and desirable by their dates. Unfortunately, the findings for the vulnerable
narcissistic traits were not as straightforward as there were few connections between these traits
and outcomes in the interpersonal interactions.
Interestingly, there was a general trend for men with greater levels of normal and
grandiose traits to be construed more positively by their female dates. Women with greater levels
of these traits did not accrue such important benefits. As mentioned previously, a potential
reason for this discrepancy resides in the norms that are expected for each gender. Simply put,
men are expected to be agentic whereas women are not (Eagly & Steffen, 1984).
All in all, Study 2 provided an important extension to Study 1 by examining the
manifestations of these narcissistic traits within a speed-dating paradigm. Future research would
benefit from recording participants’ actual behaviors during the speed-dating interactions. This
information would provide added insight into what behavioral cues are being used by
participants to inform their impressions of the participants (see Back et al., 2010).

114

CHAPTER 10: GENERAL DISCUSSION
The purpose of this dissertation was to illuminate the role of narcissism within romantic
relationship initiation. Despite having clear conceptual links to interpersonal behaviors, relatively
little research has investigated the connections between narcissism and close relationships. There
is even less research that has examined the role of narcissism in the beginning stages of a close
relationship. Unlike much of the current work in social and personality psychology which
typically relies solely on one instrument (Cain et al., 2008), I used multiple measures of
narcissism and attempted to delineate the unique roles of normal narcissism (characterized by the
use of adaptive self-esteem regulation strategies and the presence of a healthy self-image),
grandiose narcissism (characterized by the use of maladaptive self-enhancement strategies and
the presence of an inflated self-image), and vulnerable narcissism (characterized by the use of
maladaptive self-protection strategies and the presence of a depleted self-image) within the
initiation of relationships. Using the conceptual stages of romantic relationship initiation
suggested by Bredow et al. (2008), I investigated the roles of these narcissism dimensions within
attraction, self-presentation, and the development of rapport.
Below I summarize my findings for the roles of normal, grandiose, and vulnerable
narcissism in these three stages of relationship initiation. I then discuss the implications of these
results for better understanding the role of narcissism within close relationships. Finally, I cite
chief strengths and limitations of the present research and recommend areas for future study.
The Role of Narcissism in Romantic Relationship Initiation
Bredow et al. (2008) developed a conceptual model that outlines general stages involved
in romantic relationship initiation. According to these authors, romantic first encounters begin
with attraction, potentially proceed to strategic self-presentation, and possibly end with a

115

successful encounter that holds the promise for future relationship development. Although it is
possible that individuals can transition to earlier stages in the course of a romantic first
encounter, it is useful to review these stages in order for the sake of expositional clarity.
The beginning stage in Bredow et al.’s (2008) conceptual model of relationship
initiation—the Appraisal of Initial Attraction—involves processes related to romantic attraction.
This stage is concerned with the motives that individuals have for beginning their relationships
and whether they are congruent with the attributes that a potential romantic partner possesses. In
this dissertation, I predicted that each of the narcissism dimensions would be differentially
related to broader self-enhancement and self-protection motives that would in turn impact the
motives that were adopted for beginning a romantic relationship. The present research found
relatively unique connections between each of the three dimensions of narcissism and motives
for relationship initiation. It also found a few connections between some of the narcissism
dimensions and the partner attributes that were considered most attractive.
Generally speaking, normal narcissists did not pursue their romantic relationships for
self-focused reasons. There was some evidence, however, that grandiose narcissists were
concerned about self-enhancement when they pursued their romantic relationships and that they
were attracted to partners that would fulfill their needs for self-enhancement. In a similar
manner, the motives endorsed by vulnerable narcissists suggest that they pursue their romantic
relationships to both bolster and enhance their self-image.
A subsequent stage in Bredow et al.’s (2008) conceptual model of romantic relationship
initiation—Strategic Self-Presentation—centers on the presentation of the self. Bredow et al.
(2008) proposed that individuals will attempt to make themselves more attractive to others when
they are not fully confident that the other person is attracted to them. Bredow et al. (2008) further

116

suggested that this “strategic self-presentation” can take three forms: (a) appearing likeable, (b)
appearing competent/capable, and (c) appearing morally virtuous. I expected that the relations
between the narcissistic traits and the form of self-presentation adopted would hinge upon
whether the trait was more closely aligned with interpersonal motives involving selfenhancement or self-protection. If the trait was more closely aligned with self-enhancement, I
predicted that the self-presentational style would entail competence; if the narcissistic trait was
more closely aligned with self-protection (and the narcissistic individual perceived the partner to
be very desirable), I predicted that the self-presentational style would entail likeability.
The current results suggest that normal narcissists generally portray themselves as
competent and capable. This particular self-presentational style appeared to be effective for
normal narcissists as their dates considered them to be rather desirable. Similarly, the gregarious
self-presentation and ingratiation tactics used by grandiose individuals were effective in
generating positive impressions. Broadly speaking, however, there was not good evidence that
the vulnerable narcissistic traits are connected to a particular form of strategic selfpresentation—at least not one that was readily detected by the particular measures used. The
failure to find consistent relations between these traits and interpersonal perceptions may signal
something about the traits themselves. Indeed, unlike some of the traits associated with grandiose
narcissism, the vulnerable narcissistic traits do not have considerable associations with
dispositions like Extraversion. Thus, the behaviors associated with them may be less likely to be
observed in these brief five-minute interactions.
The final stage in Bredow et al.’s (2008) conceptual model of romantic relationship
initiation is the Build-Up of Rapport. To increase the likelihood of a second encounter with
another person, it is generally considered important to develop rapport during the initial

117

encounter. Besides strategic self-presentation, other ways to develop rapport within an
interaction include engaging in self-disclosure and being responsive to the other person’s
disclosure. Further, interactions with increased levels of rapport are likely to be characterized by
higher levels of trust and greater feelings of connection between the two partners involved.
Given the connections between the grandiose and vulnerable narcissistic traits and excessive
self-focus, I predicted that these traits would engender outcomes that negatively impacted
rapport during these first encounters.
Normal narcissism and vulnerable narcissism did not exhibit many connections with the
rapport-relevant variables in the present research. In contrast, the grandiose traits were associated
with interactions that could be characterized by higher levels of rapport (at least for men).
Although not originally predicted, these findings are consistent with recent work which suggests
that the expression of confidence in an initial encounter is construed positively unless there is
other information to suggest that the confidence may be unfounded (Tenney & Spellman, 2011).
Indeed, within the speed-dating interactions, it is plausible that dates generally presumed that the
grandiose individual was a good judge of his own abilities. It may not be until the date obtains
more accurate information over time that impressions of the grandiose date may change.
In sum, the narcissism dimensions were observed to possess rather unique relations with
several outcomes involved in romantic relationship initiation. Normal narcissism was generally
associated with favorable outcomes. Although the grandiose narcissistic traits demonstrated
many similar outcomes as normal narcissism, these behaviors were likely motivated by selfenhancement concerns. Finally, the vulnerable narcissistic traits were characterized by motives
for beginning relationships that reflected tendencies towards self-protection and self-

118

enhancement; however, these traits had few connections with interpersonal outcomes related to
strategic self-presentation and rapport.
Rethinking the Role of Narcissism in Romantic Relationships
Close relationships constitute an important arena for self-regulation and clarifying the
role of narcissism within this arena therefore provides important insight into how narcissists
regulate their self-image. This general idea prompted Campbell et al. (2006) to develop the
Agency Model of Narcissism, which attempts to explain the interpersonal self-regulatory
strategies used by narcissists in their romantic relationships.
According to Campbell et al. (2006), narcissists possess certain fundamental qualities
that make them successful in the beginning stages of a relationship but which eventually
deteriorate the quality of the relationship as time progresses. These fundamental qualities
include: a preoccupation with agentic issues over communal issues; an inflated view of the self
across myriad domains; a sense of entitlement; the possession of approach-oriented motivations;
and the exhibition of self-regulatory processes that emphasize the acquisition of self-esteem.
Taken together, these qualities are proposed to be associated with interpersonal skills
such as confidence, charisma, attractiveness, and charm (Campbell et al., 2006), as well as
interpersonal strategies such as game playing, self-promotion, and the attainment of trophy
partners (Campbell et al., 2006). Within the model, Campbell et al. (2006) argue that the
possession of these interpersonal skills give rise to the interpersonal strategies that narcissists
engage in within their relationships. Likewise, engagement in these interpersonal strategies and
the successful attainment of narcissistic esteem reinforce the original interpersonal skills.
The findings from the present research suggest that the normal, grandiose, and vulnerable
narcissistic traits do not reflect the same fundamental qualities and are implicated differently

119

with separate elements of relationships. Whereas normal narcissism is characterized by outcomes
that reflect diminished tendencies towards self-enhancement and self-protection, grandiose
narcissism is more closely connected to self-regulatory processes emphasizing the acquisition of
self-esteem using self-enhancement. Further, vulnerable narcissism is characterized by the
acquisition of self-esteem using both self-enhancement and self-protection strategies. In addition,
the normal, grandiose, and vulnerable narcissistic traits appear to be connected to different sets
of interpersonal skills and interpersonal strategies. Whereas normal narcissism is primarily
linked to the expression of dominance in romantic first encounters, grandiose narcissism is
related to the expression of charm, gregariousness, and dominance. However, vulnerable
narcissism is not clearly related to any particular form of strategic self-presentation. All in all,
these findings suggest that the Agency Model of Narcissism would benefit from making a
distinction between the grandiose and vulnerable expressions of narcissism (as well as between
normal and pathological narcissism more broadly) more explicit.
As it stands, the Agency Model of Narcissism also does not offer much insight into how
the vulnerable expressions of narcissism play out in romantic relationships. One of the insights
gained from the present work is that it is primarily the normal and grandiose narcissistic traits
that are important for the beginning stages of romantic relationships. It may be that the
vulnerable narcissistic traits play a more important role in the later stages of relationship
development. Indeed, research that has examined the role of Entitlement/Exploitativeness in
roommate relationships (where roommates had lived with each other for at least a few weeks)
has shown that this variable has negative implications for relationship satisfaction (Ackerman et
al., 2011). This insight may have important implications for understanding the developmental
course of narcissism within romantic relationships.

120

That being said, it is important to note that the Agency Model of Narcissism was
informed by the paradoxical finding by Paulhus (1998) regarding the impressions that narcissists
make upon their peers. Paulhus (1998) found that although narcissists (operationalized using the
NPI total score) are perceived more favorably upon first acquaintance, they are later perceived
quite negatively. The findings from this dissertation suggest that this conclusion regarding the
longitudinal impact of narcissism on relationships may be a bit misleading. Indeed, whereas
normal and grandiose narcissism may facilitate the initial development of relationships, it may be
vulnerable narcissism (and not normal or grandiose narcissism) which deteriorates or impedes
the long-term stability and quality of relationships.
At a broader level, the findings regarding normal narcissism beg the question of whether
this dimension should truly be considered narcissism. It could be argued that
Leadership/Authority is a poor operationalization of normal narcissism. Indeed, it is often
difficult to distinguish this variable from constructs such as dominance and self-esteem.
Nevertheless, even if new measures were developed to capture the broader features of normal
narcissism, it is debatable whether retaining the term “normal narcissism” would still be useful.
To be fair, many clinical theorists have suggested the existence of “normal narcissism” (e.g.,
Kernberg, 1975; Pincus & Lukowitsky, 2010; Stone, 1998); however, in these contexts, normal
narcissism is generally synonymous with healthy self-esteem regulation. Perhaps the term
narcissism is best reserved for describing those compensatory self-regulatory efforts that stem
from an underlying sense of vulnerability. Rather than labeling adaptive self-esteem regulation
strategies as narcissistic, it may make more sense to study the purported features of normal
narcissism on their own terms. That is, instead of studying normal narcissism, researchers could
study achievement striving, asserting dominance, etc. To continue labeling these behaviors as

121

“narcissistic,” even if preceded by the label “normal,” only serves to obfuscate the
conceptualization of narcissism.
Strengths, Weaknesses, and Future Directions
A major strength of this research was that it investigated the role of narcissism in several
different stages of romantic relationship initiation. In addition to investigating the role of this
construct in romantic attraction, it also examined its manifestations in self-presentation and the
development of rapport. As such, this research provided enhanced insight into the relationship
initiation process. Another important strength was the inclusion of multiple measures of
narcissism. A frequent criticism of research within social and personality psychology is that there
is an overreliance on the NPI to assess narcissism. By including both the NPI and the PNI, I was
able to obtain a broader understanding of the role of narcissism in relationships. Moreover, by
dividing the NPI into subscales, I was able to investigate the differential associations of the
personality constructs embedded within this instrument with the relationship outcomes. Use of
the NPI total score would have obscured these differences.
One limitation across both of the studies was the low alpha coefficient associated with the
NPI Entitlement/Exploitativeness subscale. The low reliabilities of the NPI subscales are a
ubiquitous criticism of the instrument (del Rosario & White, 2005). It is important to point out,
however, that the chief concern with having a low reliability is that it sets an upper limit on the
possible magnitude of the correlations. Nevertheless, the Entitlement/Exploitativeness subscale
still predicted several of the hypothesized criteria. I therefore do not view this limitation as a
serious concern.
Another criticism that could be raised is that the Relationship Initiation Motivations
measure in Study 1 has not been independently validated in separate research. Indeed, it is

122

possible that the motives I used are not representative. However, previous research has generally
uncovered similar motives (see Clark, Shaver, & Abrahams, 1999; Mongeau, Serewicz, &
Therrien, 2004). For instance, Clark et al. (1999) found that some common goals reported by
undergraduate students when they initiated romantic relationships included having sex (similar to
Hedonism), experiencing love (similar to Intimacy), and impressing others outside of
relationship (similar to Enhancement).
The use of the self-esteem manipulation in Study 1 was a notable methodological
strength. An untested assumption in previous research is that narcissists use their romantic
relationships to regulate their self-esteem (e.g., Campbell, 1999; Campbell et al., 2006). By
manipulating participants’ levels of state self-esteem, I was able to empirically evaluate whether
romantic attraction towards particular romantic partners (e.g., grandiose narcissists’ attraction
towards admiring partners) becomes stronger for narcissists whose self-esteem levels are
lowered. Unfortunately, none of the proposed hypotheses were supported.
Examination of the manipulation checks suggested that the self-esteem manipulation was
reasonably effective. Moreover, there were other relations between the narcissism dimensions
and attraction that were strengthened as a result of the self-esteem manipulation. Specifically,
when participants’ self-esteem levels were lowered, those with higher levels of
Entitlement/Exploitativeness became less attracted to the desirable partners and those with
higher levels of Narcissistic Vulnerability became more attracted to caring partners. Taken
together, these facts suggest that my hypotheses may simply have been wrong. It will be
important for future research to examine the possibility that self-esteem regulation is manifested
in areas of close relationships other than romantic attraction.

123

The use of the speed-dating paradigm in Study 2 was an important strength. With this
design, participants were able to interact with several real-life potential romantic partners rather
than viewing and rating artificial person profiles. Nevertheless, one potential concern connected
with the design was the amount of time allotted for participants to interact with their dates. As
the results showed, the vulnerable narcissistic traits demonstrated a remarkable dearth of
associations with the interpersonal perception variables. The five-minute interactions were
perhaps too brief to uncover the strategies used by vulnerable narcissists. Although it is probably
ill-advised to extend the length of time for the speed-dating interactions, use of another paradigm
(e.g., the experimental generation of closeness; see Aron, Melinat, Aron, Vallone, & Bator,
1997) may help to uncover the strategies used by vulnerable narcissists in relationship initiation.
Another important limitation of Study 2 was that self-presentation was inferred through
participants’ perceptions of their dates’ behaviors. That is, I did not record the specific behaviors
that participants engaged in during the interactions. Therefore, I do not technically have evidence
of the particular ways in which participants objectively presented themselves to their dates. It
will be useful for future research to videotape or audio record the sessions so that more objective
qualities of the interaction can be assessed. Indeed, behaviors such as gaze aversion, animation,
and physical proximity can be measured and used to index characteristics of self-presentation
and/or rapport.
Future research should aim to study the role of narcissism at other stages of romantic
relationship initiation. For instance, a neglected stage in the current research was the decision to
make an overture. There are several different ways in which the different narcissism dimensions
can impact this decision. For instance, the increased confidence associated with grandiose
narcissistic traits may make it more likely that a decision will be made to try to initiate a

124

romantic relationship. It will also be important for future research to gain more insight into the
period immediately following romantic relationship initiation. One simple way to do this is to
follow-up the participants from the speed dates after the sessions. By following up participants
after the speed date, there is an enhanced opportunity to better understand the role of narcissism
in relationship development.
General Conclusions
This dissertation demonstrated that the different narcissistic dimensions play unique and
important roles in the romantic relationship initiation process. People with higher levels of each
of these traits were found to have differing motives for beginning relationships. They were also
found to engage in different self-presentational strategies to garner positive impressions.
This dissertation also provided insights into the potential mechanisms responsible for the
deterioration of narcissists’ relationships over time. Along with the differential sets of findings
for the normal, grandiose, and vulnerable narcissistic traits, such insights lend further support to
the argument that researchers who study narcissism should be investigating these dimensions
separately. In the end, further research examining the impact of the grandiose and vulnerable
narcissistic traits on the relationship initiation process will facilitate a better understanding of
narcissism as well as relationship initiation in general.

125

Table 1
Summary of Hypotheses for the Different Narcissism Dimensions at Each Stage of Romantic Relationship Initiation
Narcissism Dimension
Normal Narcissism

Relationship
Initiation

Vulnerable Narcissism

Measure
Stage of

Grandiose Narcissism
Measure

Measure

NPI Leadership/

NPI Grandiose

PNI Narcissistic

NPI Entitlement/

PNI Narcissistic

Authority

Exhibitionism

Grandiosity

Exploitativeness

Vulnerability

Appraisal of

1. Decreased adoption

1. Decreased adoption of Intimacy

Initial Attraction

of Intimacy and Coping

Motive

1. Decreased adoption of Intimacy Motive

Motives
2. Decreased attraction

Enhancement Motives

Affirmation Motives

3. Increased attraction towards admiring

3. Increased attraction towards desirable

targets

targets

4. Decreased attraction towards caring

4. Decreased attraction towards caring

targets

Presentation

2. Increased adoption of Coping and Self-

towards caring targets

Strategic Self-

2. Increased adoption of Hedonism and

targets

1. Increased expressions of dominance,

1. Generally speaking, decreased

gregariousness, arrogance, and charm

expressions of warmth and dominance

No hypotheses

126

Table 1 (cont’d)
Narcissism Dimension
Normal Narcissism

Relationship
Initiation

Vulnerable Narcissism

Measure
Stage of

Grandiose Narcissism
Measure

Measure

NPI Leadership/

NPI Grandiose

PNI Narcissistic

NPI Entitlement/

PNI Narcissistic

Authority

Exhibitionism

Grandiosity

Exploitativeness

Vulnerability

2. In cases where vulnerable participants
Strategic Self-

perceive their date to be desirable,

Presentation

increased expressions of warmth and
admiration

Building Rapport

No hypotheses

1. Will be considered more attractive by their

1. Generally speaking, will be considered

dates

less attractive by their dates

2. Relation between participants’ levels of
grandiose narcissism and their dates’
attraction towards them will be mediated by
their dates’ perceptions of the participants’
levels of gregariousness and desirability
3. Will disclose less to their dates
127

2. In cases where vulnerable participants
perceive their date to be desirable, will be
considered more attractive by their dates,
and this relation will in turn be mediated
by their dates’ perceptions of their warmth
and admiration
3. Will disclose less to their dates

Table 1 (cont’d)
Narcissism Dimension
Normal Narcissism

Relationship
Initiation

Vulnerable Narcissism

Measure
Stage of

Grandiose Narcissism
Measure

Measure

NPI Leadership/

NPI Grandiose

PNI Narcissistic

NPI Entitlement/

PNI Narcissistic

Authority

Exhibitionism

Grandiosity

Exploitativeness

Vulnerability

4. Will be seen as less responsive

4. Will be seen as less responsive

5. Will feel less connected to their dates

Building Rapport

5. Will feel less connected to their dates

Note. NPI = Narcissistic Personality Inventory. PNI = Pathological Narcissism Inventory.

128

Table 2
Pattern Coefficients from an Exploratory Factor Analysis of the Relationship Initiation Motives Inventory (RIM)
RIM Item

RIM Subscale

I

II

III

IV

To receive compliments

Enhancement

1.00

-.21

-.05

-.01

To impress others

Enhancement

.80

-.11

-.08

.06

Because they remind me I’m a great person

Enhancement

.75

-.00

.09

-.05

To be admired

Enhancement

.64

.10

.13

.03

Enhancement

.57

-.03

.20

-.07

Self-Affirmation

.47

.16

-.20

.16

Self-Affirmation

.39

.37

-.08

.17

Coping

-.23

.96

-.12

-.13

Coping

-.10

.83

-.14

-.09

Coping

.08

.75

.07

-.03

Coping

-.03

.64

.20

.06

To have other person treat me like the
center of his/her world
To prove to myself that the other person
thinks I’m attractive
To reassure myself that I am desirable
To help me deal with disappointment in my
life
To cope with upset feelings
Because it helps me feel better when I’m
feeling low
Because it helps me feel better when I’m
lonely

129

Table 2 (cont’d)
RIM Item

RIM Subscale

I

II

III

IV

To help me feel better about myself

Self-Affirmation

.38

.57

-.02

-.04

To cheer myself up

Coping

.42

.53

-.01

-.09

Self-Affirmation

.05

.43

.09

.34

Self-Affirmation

.17

.42

.10

.20

Intimacy

.02

-.10

.85

-.02

Intimacy

-.11

.11

.79

.01

Intimacy

.17

-.14

.77

-.17

Intimacy

.07

-.12

.67

.08

Intimacy

-.04

.10

.65

-.01

Hedonism

-.07

-.07

.08

.83

Because it makes me feel more selfconfident
Because it makes me feel like I’m a more
interesting person
To feel close with the other person
Because I want to feel connected to the
other person
Because I want to open up to the other
person
To better understand the other person
To have another person truly understand
me
Just for the excitement of it

130

Table 2 (cont’d)
RIM Item

RIM Subscale

I

II

III

IV

Just for the thrill of it

Hedonism

.05

-.11

-.06

.80

To satisfy my sexual needs

Hedonism

.05

-.05

.00

.72

Because I feel “horny”

Hedonism

.07

-.04

-.23

.57

Because it feels good

Hedonism

-.16

.09

.36

.43

Note. All pattern loadings ≥ .50 are bolded. r between I and II = .69. r between I and III = .18. r between I and IV = .63. r between II
and III = .19. r between II and IV = .60. r between III and IV = .09.

131

Table 3
Zero-order correlations between primary independent variables in Study 1
1

2

3

4

5

NPI Scales
1. Leadership/Authority
2. Grandiose Exhibitionism

.43**

3. Entitlement/Exploitativeness

.22**

.24**

4. Grandiosity

.36**

.28**

.22**

5. Vulnerability

.02

.07

.29**

PNI Scales
.54**

Note. NPI = Narcissistic Personality Inventory. PNI = Pathological Narcissism Inventory. **p <
.01.

132

Table 4
Zero-order correlations between RIM variables and attachment-related avoidance in Study 1
Variable

1

2

3

4

1. RIM Intimacy
2. RIM Coping/Self-Affirmation

.13*

3. RIM Hedonism

.07

.56**

4. RIM Enhancement

.21**

.69**

.49**

5. Attachment Avoidance

-.06

.11*

.09

.09

Note. RIM = Relationship Initiation Motives Inventory. *p < .05. **p < .01.

133

5

Table 5
Zero-order correlations between primary independent variables and dependent variables in
Study 1
NPI Scales
Leadership/
Authority

Grandiose

PNI Scales
Entitlement/

Exhibitionism Exploitativeness

Grandiosity Vulnerability

RIM Variables
Intimacy

-.09

-.09

-.16**

.08

.13*

Coping/Self-

-.09

.06

.21**

.25**

.53**

.06

.21**

.23**

.21**

.29**

-.02

.17**

.25**

.32**

.49**

-.04

-.22**

Affirmation
Hedonism
Enhancement
Avoidance

.08

-.01

.33**

Note. RIM = Relationship Initiation Motives Inventory. NPI = Narcissistic Personality Inventory.

134

Table 6
Parameter Estimates from Path Analysis using the Narcissistic Personality Inventory (NPI) Traits to Predict Relationship Initiation
Motivations
NPI Scales
Leadership/Authority
RIM Scales

Grandiose Exhibitionism

b

SE

β

b

SE

β

-0.12

.16

-.04

-0.12

.18

-.04

-0.57**

.19

-.18

0.31

.21

.09

Hedonism

-0.21

.18

-.07

0.68**

.20

Enhancement

-0.51**

.19

-.15

0.66**

.21

Intimacy
Coping/
Self-Affirmation

Entitlement/Exploitativeness
b

SE

β

-0.41*

.17

-.14

0.80**

.19

.23

.19

0.69**

.19

.20

.18

0.88**

.20

.24

Note. RIM = Relationship Initiation Motives Inventory. SE = Standard error of the unstandardized regression coefficient. *p < .05.
**p < .01.

135

Table 7
Parameter Estimates from an Unconstrained Multiple-Groups Analysis for Men and Women using the Narcissistic Personality
Inventory Subscales to Predict Relationship Initiation Motivations
Gender
Women
b

SE

Men
β

b

SE

β

Leadership/Authority (LA)
LA

Intimacy

-0.33

.21

-.11

0.38

.25

.16

LA

Coping/Self-Affirmation

-0.97**

.23

-.30

0.07

.32

.02

LA

Hedonism

-0.36

.22

-.12

-0.06

.32

-.02

LA

Enhancement

-0.98**

.23

-.29

0.20

.34

.06

-0.05

.25

-.02

-0.37

.26

-.14

-0.22

.33

-.06

Grandiose Exhibitionism (GE)
GE

Intimacy

GE

Coping/Self-Affirmation

0.60*

.27

.16

GE

Hedonism

0.67*

.26

.19

0.70*

.34

.20

GE

Enhancement

0.74**

.27

.19

0.34

.36

.10

-0.20

.22

-.07

-0.82**

.25

-.28

Entitlement/Exploitativeness (EE)
EE

Intimacy

EE

Coping/Self-Affirmation

0.41

.23

.12

1.33**

.32

.35

EE

Hedonism

0.46*

.23

.14

1.07**

.32

.28

EE

Enhancement

0.72**

.24

.21

1.07**

.34

.27

Note. SE = Standard error of the unstandardized regression coefficient. *p < .05. **p < .01.
136

Table 8
Parameter Estimates from Path Analysis using the Pathological Narcissism Inventory (PNI)
Traits to Predict Relationship Initiation Motivations
PNI Scales
Grandiosity
RIM Scales

Vulnerability

SE

β

0.01

.08

.01

-0.08

.08

Hedonism

0.11

Enhancement

0.11

Intimacy
Coping/
Self-Affirmation

b

SE

β

0.13*

.06

.13

-.06

0.64**

.06

.56

.09

.08

0.28**

.07

.25

.08

.07

0.53**

.07

.45

b

Note. RIM = Relationship Initiation Motives Inventory. SE = Standard error of the
unstandardized regression coefficient. *p < .05. **p < .01.

137

Table 9
Parameter Estimates from an Unconstrained Multiple-Groups Analysis for Men and Women using the Pathological Narcissism
Inventory Scales to Predict Relationship Initiation Motivations
Gender
Women
b

SE

Men
β

b

SE

β

Narcissistic Grandiosity (NG)
NG

Intimacy

0.01

.10

.01

0.02

.11

.02

NG

Coping/Self-Affirmation

-0.17

.10

-.12

-0.02

.11

-.01

NG

Hedonism

-0.00

.11

-.00

0.22

.14

.15

NG

Enhancement

0.03

.11

.02

0.17

.13

.11

Narcissistic Vulnerability (NV)
NV

Intimacy

0.17*

.08

.18

0.03

.09

.03

NV

Coping/Self-Affirmation

0.63**

.08

.57

0.71**

.09

.61

NV

Hedonism

0.28**

.09

.26

0.34**

.11

.28

NV

Enhancement

0.55**

.09

.48

0.55**

.11

.44

Note. SE = Standard error of the unstandardized regression coefficient. *p < .05. **p < .01.

138

Table 10
Means and Standard Deviations for Romantic Targets as a Function of Target Presentation Order
Attraction Ratings Towards Target
Admiring

Desirable

Caring

Target Order

n

Mean

SD

Mean

SD

Mean

SD

1. Admiring, Desirable, and Caring

50

4.82

0.97

5.10

1.17

5.27

1.12

2. Admiring, Caring, and Desirable

51

4.78

1.28

4.69

1.19

4.90

1.50

3. Desirable, Admiring, and Caring

50

4.70

1.29

5.01

1.29

5.25

1.06

4. Desirable, Caring, and Admiring

97

4.73

1.04

5.49

1.08

5.39

1.16

5. Caring, Desirable, and Admiring

49

4.76

1.14

5.17

1.34

5.46

1.07

6. Caring, Admiring, and Desirable

45

4.68

1.03

4.97

1.36

5.32

1.14

139

Table 11
Zero-order correlations between primary dependent variables in Study 1
Variable

1

2

1. Admiring Partner
2. Desirable Partner

.16**

3. Caring Partner

.11*

-.05

Note. *p < .05. **p < .01.

140

3

Table 12
Zero-order correlations between primary independent variables and dependent variables in
Study 1
NPI Scales
Leadership/
Authority

Grandiose

PNI Scales
Entitlement/

Exhibitionism Exploitativeness

Admiring Partner

.06

.11*

.04

Desirable Partner

.16**

.15**

Caring Partner

-.05

Grandiosity Vulnerability
.11

-.02

.16**

.05

-.07

-.10

.06
.07

.03

Note. NPI = Narcissistic Personality Inventory. PNI = Pathological Narcissism Inventory. *p <
.05. **p < .01.

141

Table 13
Multiple Regression Analyses using the Narcissistic Personality Inventory Subscales to Predict
Attraction towards the Romantic Targets
Narcissistic Personality Inventory Subscales
Leadership/

Entitlement/

Authority
Romantic Targets

Grandiose
Exhibitionism

Exploitativeness

b

SE

β

b

SE

β

Admiring

0.07

.26

.02

0.47

.29

Desirable

0.62*

.28

.13

0.61

.28

-.01

-0.42

Caring

-0.04

b

SE

β

.10

0.08

.26

.02

.31

.12

-0.39

.29

-.07

.31

-.08

-0.22

.28

-.04

Note. SE = Standard error of the unstandardized regression coefficient. *p < .05.

142

Table 14
Multiple Regression Analyses using the Pathological Narcissism Inventory Scales to Predict
Attraction towards the Romantic Targets
Pathological Narcissism Inventory Scales
Grandiosity
Romantic Targets

b

Vulnerability

SE

β

b

SE

β

Admiring

0.01

.12

.01

0.15

.10

.10

Desirable

0.40**

.13

.19

-0.09

.11

-.05

Caring

0.16

.13

.08

-0.03

.10

-.02

Note. SE = Standard error of the unstandardized regression coefficient. **p < .01.

143

Table 15
Means and Standard Deviations of Attraction towards Romantic Targets
Remote Associations Test Condition
Success

Control

(n = 152)
Romantic Target

Failure
(n = 103)

(n = 91)

Mean

SD

Mean

SD

Mean

Admiring

4.69

1.14

4.75

1.14

4.78

1.10

Desirable

5.13

1.20

5.26

1.22

4.94

1.34

Caring

5.08

1.22

5.44

1.24

5.34

1.09

144

SD

Table 16
Predicting Attraction towards Admiring Target with Narcissistic Personality Inventory Traits
SE

b
Model Intercept

4.80**

2

ΔR

0.12

Remote Associations Test Condition
Success Condition

-0.12

0.16

Failure Condition

-0.08

0.17

Leadership/Authority

0.08

0.52

Grandiose Exhibitionism

0.56

0.58

-0.08

0.50

Leadership/Authority by RAT Condition
Leadership/Authority by Success Condition

0.03

0.65

Leadership/Authority by Failure Condition

-0.00

0.72

Grandiose Exhibitionism by RAT Condition
Grandiose Exhibitionism by Success Condition

-0.36

0.74

Grandiose Exhibitionism by Failure Condition

0.25

0.77

Entitlement/Exploitativeness by RAT Condition
Entitlement/Exploitativeness by Success Condition

0.14

0.65

Entitlement/Exploitativeness by Failure Condition

0.25

0.73

.001

Narcissistic Personality Inventory Traits

Entitlement/Exploitativeness

.013

Interactions between Variables
a

.000

b

.002

c

145

.000

Table 16 (cont’d)
2

Note. SE = Standard error of the unstandardized regression coefficient. **p < .01. a = ΔR in this row represents the change in Rsquared obtained by including the Leadership/Authority by RAT condition dummy codes as the final block in the model (i.e., after the
blocks containing the interactions between Grandiose Exhibitionism and the RAT condition dummy codes and the interactions
2

between Entitlement/Exploitativeness and the RAT condition dummy codes). b = ΔR in this row represents the change in R-squared
obtained by including the Grandiose Exhibitionism by RAT condition dummy codes as the final block in the model (i.e., after the
blocks containing the interactions between Leadership/Authority and the RAT condition dummy codes and the interactions between
2

Entitlement/Exploitativeness and the RAT condition dummy codes). c = ΔR in this row represents the change in R-squared obtained
by including the Entitlement/Exploitativeness by RAT condition dummy codes as the final block in the model (i.e., after the blocks
containing the interactions between Leadership/Authority and the RAT condition dummy codes and the interactions between
Grandiose Exhibitionism and the RAT condition dummy codes).

146

Table 17
Predicting Attraction towards Desirable Target with Narcissistic Personality Inventory Traits
SE

b
Model Intercept

4.92**

0.14

Success Condition

0.21

0.17

Failure Condition

0.28

0.18

Leadership/Authority

0.61

0.57

Grandiose Exhibitionism

0.30

0.63

Entitlement/Exploitativeness

0.64

0.54

Leadership/Authority by RAT Condition
Leadership/Authority by Success Condition

0.18

0.70

Leadership/Authority by Failure Condition

-0.34

0.78

Grandiose Exhibitionism by RAT Condition
Grandiose Exhibitionism by Success Condition

-0.05

0.80

Grandiose Exhibitionism by Failure Condition

0.81

0.84

Entitlement/Exploitativeness by RAT Condition
Entitlement/Exploitativeness by Success Condition

-1.16

0.70

Entitlement/Exploitativeness by Failure Condition

-1.83*

0.79

2

ΔR

Remote Associations Test Condition
.010

Narcissistic Personality Inventory Traits

.036**

Interactions between Variables
a

.002

b

.004

c

147

.016

Table 17 (cont’d)
2

Note. SE = Standard error of the unstandardized regression coefficient. *p < .05. **p < .01. a = ΔR in this row represents the change
in R-squared obtained by including the Leadership/Authority by RAT condition dummy codes as the final block in the model (i.e.,
after the blocks containing the interactions between Grandiose Exhibitionism and the RAT condition dummy codes and the
2

interactions between Entitlement/Exploitativeness and the RAT condition dummy codes). b = ΔR in this row represents the change in
R-squared obtained by including the Grandiose Exhibitionism by RAT condition dummy codes as the final block in the model (i.e.,
after the blocks containing the interactions between Leadership/Authority and the RAT condition dummy codes and the interactions
2

between Entitlement/Exploitativeness and the RAT condition dummy codes). c = ΔR in this row represents the change in R-squared
obtained by including the Entitlement/Exploitativeness by RAT condition dummy codes as the final block in the model (i.e., after the
blocks containing the interactions between Leadership/Authority and the RAT condition dummy codes and the interactions between
Grandiose Exhibitionism and the RAT condition dummy codes).

148

Table 18
Simple Slopes for each Remote Associations Test (RAT) Condition for the Relation between the
Narcissistic Personality Inventory Subscale of Entitlement/Exploitativeness and Attraction
towards the Desirable Target
Entitlement/Exploitativeness
b

SE

β

Success

-0.52

.44

-.10

Failure

-1.19*

.57

-.23

Control

0.64

.54

.12

RAT Condition

Note. SE = Standard error of the unstandardized regression coefficient. *p < .05.

149

Table 19
Predicting Attraction towards Caring Target with Narcissistic Personality Inventory Traits
SE

b
Model Intercept

5.31**

2

ΔR

.13

Remote Associations Test Condition
Success Condition

-0.24

.16

Failure Condition

0.16

.18

0.83

.55

-0.66

.61

0.03

.53

Leadership/Authority by RAT Condition
Leadership/Authority by Success Condition

-1.46*

.68

Leadership/Authority by Failure Condition

-0.80

.76

Grandiose Exhibitionism by RAT Condition
Grandiose Exhibitionism by Success Condition

0.64

.77

Grandiose Exhibitionism by Failure Condition

-0.16

.81

Entitlement/Exploitativeness by RAT Condition
Entitlement/Exploitativeness by Success Condition

-0.88

.68

Entitlement/Exploitativeness by Failure Condition

0.36

.76

.017*

Narcissistic Personality Inventory Traits
Leadership/Authority
Grandiose Exhibitionism
Entitlement/Exploitativeness

.012

Interactions between Variables
a

.013

b

.004

c

150

.010

Table 19 (cont'd)
2

Note. SE = Standard error of the unstandardized regression coefficient. *p < .05. **p < .01. a = ΔR in this row represents the change
in R-squared obtained by including the Leadership/Authority by RAT condition dummy codes as the final block in the model (i.e.,
after the blocks containing the interactions between Grandiose Exhibitionism and the RAT condition dummy codes and the
2

interactions between Entitlement/Exploitativeness and the RAT condition dummy codes). b = ΔR in this row represents the change in
R-squared obtained by including the Grandiose Exhibitionism by RAT condition dummy codes as the final block in the model (i.e.,
after the blocks containing the interactions between Leadership/Authority and the RAT condition dummy codes and the interactions
2

between Entitlement/Exploitativeness and the RAT condition dummy codes). c = ΔR in this row represents the change in R-squared
obtained by including the Entitlement/Exploitativeness by RAT condition dummy codes as the final block in the model (i.e., after the
blocks containing the interactions between Leadership/Authority and the RAT condition dummy codes and the interactions between
Grandiose Exhibitionism and the RAT condition dummy codes).

151

Table 20
Predicting Attraction towards Admiring Target with Pathological Narcissism Inventory Traits
SE

b
Model Intercept

4.78**

2

ΔR

0.12

Remote Associations Test Condition
Success Condition

-0.10

0.15

Failure Condition

-0.05

0.16

-0.02

0.23

0.16

0.20

Narcissistic Grandiosity by RAT Condition
Narcissistic Grandiosity by Success Condition

-0.09

0.31

Narcissistic Grandiosity by Failure Condition

0.25

0.32

Narcissistic Vulnerability by RAT Condition
Narcissistic Vulnerability by Success Condition

-0.02

0.25

Narcissistic Vulnerability by Failure Condition

0.06

0.26

.001

Pathological Narcissism Inventory Traits
Narcissistic Grandiosity
Narcissistic Vulnerability

.011

Interactions between Variables
a

.004

b

152

.000

Table 20 (cont’d)
2

Note. SE = Standard error of the unstandardized regression coefficient. **p < .01. a = ΔR in this row represents the change in Rsquared obtained by including the Narcissistic Grandiosity by RAT condition dummy codes as the final block in the model (i.e., after
2

the blocks containing the interactions between Narcissistic Vulnerability and the RAT condition dummy codes). b = ΔR in this row
represents the change in R-squared obtained by including the Narcissistic Vulnerability by RAT condition dummy codes as the final
block in the model (i.e., after the blocks containing the interactions between Narcissistic Grandiosity and the RAT condition dummy
codes).

153

Table 21
Predicting Attraction towards Desirable Target with Pathological Narcissism Inventory Traits
SE

b
Model Intercept

4.95**

0.13

Success Condition

0.20

0.17

Failure Condition

0.29

0.18

0.56*

2

ΔR

0.26

Remote Associations Test Condition
.010

Pathological Narcissism Inventory Traits
Narcissistic Grandiosity
Narcissistic Vulnerability

-0.29

0.22

Narcissistic Grandiosity by RAT Condition
Narcissistic Grandiosity by Success Condition

-0.23

0.34

Narcissistic Grandiosity by Failure Condition

-0.23

0.35

Narcissistic Vulnerability by RAT Condition
Narcissistic Vulnerability by Success Condition

0.29

0.28

Narcissistic Vulnerability by Failure Condition

0.24

0.29

.027*

Interactions between Variables
a

.002

b

.003
2

Note. SE = Standard error of the unstandardized regression coefficient. *p < .05. **p < .01. a = ΔR in this row represents the change
in R-squared obtained by including the Narcissistic Grandiosity by RAT condition dummy codes as the final block in the model (i.e.,

154

Table 21 (cont’d)
2

after the blocks containing the interactions between Narcissistic Vulnerability and the RAT condition dummy codes). b = ΔR in this
row represents the change in R-squared obtained by including the Narcissistic Vulnerability by RAT condition dummy codes as the
final block in the model (i.e., after the blocks containing the interactions between Narcissistic Grandiosity and the RAT condition
dummy codes).

155

Table 22
Predicting Attraction towards Caring Target with Pathological Narcissism Inventory Traits

Model Intercept

2

SE

b
5.35**

ΔR

0.13

Remote Associations Test Condition
Success Condition

-0.25

0.16

Failure Condition

0.10

0.17

0.30

0.25

-0.32

0.21

Narcissistic Grandiosity by RAT Condition
Narcissistic Grandiosity by Success Condition

-0.07

0.32

Narcissistic Grandiosity by Failure Condition

-0.30

0.33

Narcissistic Vulnerability by RAT Condition
Narcissistic Vulnerability by Success Condition

0.15

0.27

Narcissistic Vulnerability by Failure Condition

0.71*

0.28

.017*

Pathological Narcissism Inventory Traits
Narcissistic Grandiosity
Narcissistic Vulnerability

.003

Interactions between Variables
a

.003

b

.023*
2

Note. SE = Standard error of the unstandardized regression coefficient. **p < .01. a = ΔR in this row represents the change in Rsquared obtained by including the Narcissistic Grandiosity by RAT condition dummy codes as the final block in the model (i.e., after

156

Table 22 (cont’d)
2

the blocks containing the interactions between Narcissistic Vulnerability and the RAT condition dummy codes). b = ΔR in this row
represents the change in R-squared obtained by including the Narcissistic Vulnerability by RAT condition dummy codes as the final
block in the model (i.e., after the blocks containing the interactions between Narcissistic Grandiosity and the RAT condition dummy
codes).

157

Table 23
Simple Slopes for each Remote Associations Test (RAT) Condition for the Relation between the
Pathological Narcissism Inventory Scale of Narcissistic Vulnerability and Attraction towards the
Caring Target
Vulnerability
b

SE

β

-0.17

.16

-.11

.18

.24

.21

-.20

RAT Condition
Success
Failure
Control

0.38*
-0.32

Note. SE = Standard error of the unstandardized regression coefficient. *p < .05.

158

Table 24
Illustration of Asymmetric Block Design
Target
M1

W1

W2

W3

M1

X

X

X

M2
Perceiver

M2

M3

X

X

X

M3

X

X

X

W1

X

X

X

W2

X

X

X

W3

X

X

X

Note. M = Man. W = Woman.

159

Table 25
Social Relations Model Variance Partitioning for Dating Record Variables in Study 2
Relative Percentage of Variance in
Dating Record Variable Accounted for
by SRM component
Dating Record Variables

Perceiver

Target

Relationship

Total

plus error

Variance

Interpersonal Behavior
M-W Assured-Dominant

21.21*

16.23*

62.56

1.43

W-M Assured-Dominant

19.15**

30.86**

49.99

1.90

M-W Cold-hearted

47.22**

0.00

52.78

0.95

W-M Cold-hearted

30.37**

0.24

69.39

0.92

M-W Unassured-Submissive

20.30**

24.77**

54.93

1.64

W-M Unassured-Submissive

16.02**

30.58**

53.39

1.68

M-W Arrogant-Calculating

41.03**

2.46

56.51

1.00

W-M Arrogant-Calculating

47.48**

6.75*

45.76

1.16

M-W Warm-Agreeable

44.34**

14.65*

41.01

0.96

W-M Warm-Agreeable

33.95**

4.44

61.61

1.40

M-W Gregarious-Extraverted

16.63**

28.09**

55.28

1.01

W-M Gregarious-Extraverted

12.54*

29.81**

57.65

1.50

M-W Admiring

52.89**

6.56*

40.55

1.10

W-M Admiring

49.60**

7.56

42.84

1.23

M-W Desirable

21.98**

37.54**

40.47

1.17

W-M Desirable

17.48**

37.24**

45.27

1.55

M-W Caring

33.13**

12.95**

53.92

0.70

W-M Caring

30.67**

12.27**

57.06

1.16

M-W Romantic Attraction

18.37**

33.62**

48.01

1.58

W-M Romantic Attraction

22.97**

31.65**

45.38

1.99

24.84**

24.56**

50.59

1.39

Appeal

Rapport
M-W Connection

160

Table 25 (cont’d)
Relative Percentage of Variance in
Dating Record Variable Accounted for
by SRM component
Dating Record Variables

Perceiver

Target

Relationship

Total

plus error

Variance

Rapport
W-M Connection

21.70**

14.74*

63.57

1.66

M-W Disclose

48.70**

4.86

46.44

1.80

W-M Disclose

48.31**

8.07*

43.62

2.11

M-W Trust

27.83**

0.00

72.17

2.21

W-M Trust

47.23**

11.34**

41.42

2.37

M-W Perceived Responsiveness

38.44**

7.92*

53.64

1.14

W-M Perceived Responsiveness

32.08**

9.32*

58.60

1.40

Note. Estimates of parameters obtained from BLOCKO. M-W = Rating from men to women.
W-M = Rating from women to men. *p < .05. **p < .01.

161

Table 26
Reciprocity Correlations for Dating Record Variables in Study 2
Generalized Reciprocity Coefficients

Dyadic Reciprocity

Men

Women

Coefficients

r 

r 

r

-.25

-.23

-.05

nc

nc

.07

Unassured-Submissive

-.25

.36

-.06

Arrogant-Calculating

-.13

nc

.07

Warm-Agreeable

nc

.05

.13*

Gregarious-Extraverted

.08

-.07

.04

Admiring

nc

.28

.22*

Desirable

-.17

-.13

.18**

.01

-.00

.17**

-.01

-.14

.26**

.29

.35

.37**

Disclose

-.19

nc

.08

Trust

-.22

nc

.04

.43

-.02

Dating Record Variables
Interpersonal Behavior
Assured-Dominant
Cold-hearted

Appeal

Caring
Romantic Attraction
Rapport
Connection

Perceived Responsiveness

.23**

Note. Estimates of parameters obtained from BLOCKO. nc = estimate not computed due to a
lack of significant target variance. *p < .05. **p < .01.

162

Table 27
Zero-order correlations between narcissism variables in Study 2
1

2

3

4

5

1. NPI Leadership/Authority
2. NPI Grandiose Exhibitionism

.47**

3. NPI Entitlement/Exploitativeness

.35**

.31**

4. PNI Grandiosity

.34**

.37**

.32**

5. PNI Vulnerability

.10

.17*

.23**

.47**

Note. NPI = Narcissistic Personality Inventory. PNI = Pathological Narcissism Inventory. *p <
.05. **p < .01.

163

Table 28
Zero-order correlations between participants’ scores on the Narcissistic Personality Inventory (NPI) subscales, the Pathological
Narcissism Inventory (PNI) scales, and their Social Relations Model (SRM) Target effects for the Interpersonal Adjective ScaleRevised (IAS-R) Dating Record Variables
NPI Scales

PNI Scales

Leadership/

Effects
AssuredDominance
Arrogant-

Entitlement/

Authority
SRM Target

Grandiose
Exhibitionism

Exploitativeness
Men

Men

Women

Men

Women

.37*

.19

.52**

.22

.35

nc

.63**

-.27

-.26*

nc
.42*

Grandiosity

Vulnerability

Women

Men

Women

Men

Women

.05

.17

.15

-.14

-.03

-.08

nc

.12

nc

.01

nc

.06

nc

-.48*

-.26

-.01

-.29*

-.20

-.08

.03

-.06

-.26*

nc

-.21

nc

-.16

nc

-.17

nc

-.08

.06

.51*

-.04

.08

.04

.39*

-.17

-.01

-.14

Calculating
UnassuredSubmissive
WarmAgreeable
GregariousExtraverted
Note. Estimates of correlations obtained from BLOCKO and are disattenuated for error. nc = estimate not computed due to a lack of
significant target variance. *p < .05. **p < .01.

164

Table 29
Predicting Dates’ Perceptions of Participants’ Interpersonal Behavior from the Participants’ Narcissistic Personality Inventory (NPI)
Traits
Assured-

Arrogant-

Unassured-

Warm-

Gregarious-

Dominant

Calculating

Submissive

Agreeable

Extraverted

b

SE

b

SE

3.63**

.05

2.82**

.06

Gender

0.02

.07

-0.26**

.06

-0.02

.07

-0.13*

.05

LA

0.45*

.23

0.16

.15

-0.24

.24

-0.00

.16

0.45*

.21

GE

0.72**

.25

0.41*

.16

-0.68*

.26

-0.12

.18

0.49*

.23

Model Intercept

b
3.85**

SE

b

SE

b

SE

.06

4.88**

.06

5.17**

.06

First-Order Effects

EE

-0.07

-0.09

.05

-0.20

.25

-0.05

.16

-0.22

.26

0.05

.18

.23

LA by Gender

0.27

.23

0.01

.15

0.17

.24

0.27

.16

0.42*

.21

GE by Gender

0.33

.25

0.29

.16

-0.37

.26

0.06

.18

0.42

.23

EE by Gender

-0.12

.25

-0.13

.16

0.09

.26

0.17

.18

-0.13

.23

Interactions with Gender

Note. SE = Standard error of the unstandardized regression coefficient. LA = Leadership/Authority. GE = Grandiose Exhibitionism.
EE = Entitlement/Exploitativeness. *p < .05. **p < .01.

165

Table 30
Predicting Dates’ Perceptions of Participants’ Interpersonal Behavior from the Participants’ Pathological Narcissism Inventory
(PNI) Traits
Assured-

Arrogant-

Unassured-

Warm-

Gregarious-

Dominant

Calculating

Submissive

Agreeable

Extraverted

b

SE

b

SE

b

SE

b

SE

b

SE

3.61**

.06

2.83**

.06

3.88**

.07

4.88**

.06

5.15**

.05

Gender

0.04

.07

-0.26**

.06

-0.03

.07

-0.16**

.05

NG

0.10

.09

0.08

.06

-0.17

.10

0.09

.06

NV

-0.09

.08

-0.08

.05

0.06

.08

-0.05

.06

NG by Gender

0.16

.09

-0.06

.06

-0.16

.10

NV by Gender

-0.08

.08

0.06

.05

0.09

.08

Model Intercept
First-Order Effects

-0.09

.06

0.28**
†

-0.14

.09
.07

Interactions with Gender
0.15*
-0.02

.06
.06

0.30**
-0.08

Note. SE = Standard error of the unstandardized regression coefficient. NG = Narcissistic Grandiosity. NV = Narcissistic
Vulnerability. *p < .05. **p < .01. †p = .06.

166

.09
.07

Table 31
Zero-order correlations between participants’ scores on the Narcissistic Personality Inventory (NPI) subscales, the Pathological
Narcissism Inventory (PNI) scales, and their Social Relations Model (SRM) Target effects for the Appeal Dating Record Variables
NPI Scales

PNI Scales

Leadership/

Effects

Entitlement/

Authority
SRM Target

Grandiose
Exhibitionism

Exploitativeness

Women

Men

Men

Men

Admiring

nc

-.20

nc

Desirable

.36*

.05

.48**

Attraction

.35*

.05

.49**

Women
-.13

Women

Grandiosity

Vulnerability

Men

Women

Men

Women

nc

.03

nc

-.15

nc

.07

.27*

.10

.12

.40*

.01

.16

-.02

.16

.10

.11

.41**

.07

.14

.02

Note. Estimates of correlations obtained from BLOCKO and are disattenuated for error. nc = estimate not computed due to a lack of
significant target variance. *p < .05. **p < .01.

167

Table 32
Predicting Dates’ Perceptions of Participants’ Admiration and Desirability, and Dates’
Attraction to Participants, from the Participants’ Narcissistic Personality Inventory (NPI) Traits
Admiring

Attraction

SE

b

SE

b

SE

3.67**

.06

4.20**

.06

3.88**

.07

-0.25**

.06

-0.11

.06

-0.16*

.07

b
Model Intercept

Desirable

First-Order Effects
Gender
LA

0.08

.15

0.39

.24

0.35

.27

GE

0.06

.16

0.75**

.26

0.68*

.30

EE

0.08

.16

0.02

.26

0.04

.30

LA by Gender

0.31*

.15

0.46†

.24

0.37

.27

GE by Gender

0.32†

.16

0.13

.26

0.37

.30

.16

-0.09

.26

-0.23

.30

Interactions with Gender

EE by Gender

-0.11

Note. SE = Standard error of the unstandardized regression coefficient. LA =
Leadership/Authority. GE = Grandiose Exhibitionism. EE = Entitlement/Exploitativeness. *p <
.05. **p < .01. †p < .06.

168

Table 33
Predicting Dates’ Perceptions of Participants’ Admiration and Desirability, and Dates’
Attraction to Participants, from the Participants’ Pathological Narcissism Inventory (PNI)
Traits
Admiring

Desirable

Attraction

SE

b

SE

b

SE

3.67**

.07

4.19**

.06

3.87**

.07

Gender

-0.24**

.06

NG

-0.02

.06

NV

0.00

.05

0.12*

.06
.05

b
Model Intercept
First-Order Effects

-0.11
0.29**
-0.04

.07
.10

-0.17*
0.32**

.08
.11

.08

-0.07

.09

0.23*

.10

0.19

.11

0.03

.08

-0.01

.09

Interactions with Gender
NG by Gender
NV by Gender

-0.07

Note. SE = Standard error of the unstandardized regression coefficient. NG = Narcissistic
Grandiosity. NV = Narcissistic Vulnerability. *p < .05. **p < .01.

169

Table 34
Evaluating whether the relation between participants’ NPI traits and their dates’ perceptions of
them as Warm-Agreeable is Moderated by participants’ perception of their dates’ desirability
b
4.88**

Model Intercept

SE
.05

First-Order Effects
Gender

-0.14*

.05

Desirability

0.11**

.03

Leadership/Authority

0.04

.16

-0.11

.18

0.07

.18

Leadership/Authority by Gender

0.30

.16

Grandiose Exhibitionism by Gender

0.05

.18

Entitlement/Exploitativeness by Gender

0.20

.18

0.02

.03

-0.11

.13

Grandiose Exhibitionism by Desirability

0.23

.15

Entitlement/Exploitativeness by Desirability

0.04

.13

-0.17

.13

Grandiose Exhibitionism by Gender by Desirability

0.18

.15

Entitlement/Exploitativeness by Gender by Desirability

0.04

.13

Grandiose Exhibitionism
Entitlement/Exploitativeness
Two-Way Interactions between NPI traits and Gender

Two-Way Interaction between Gender and Desirability
Gender by Desirability
Two-Way Interactions between NPI traits and Desirability
Leadership/Authority by Desirability

Three-Way Interactions for NPI traits, Desirability, and Gender
Leadership/Authority by Gender by Desirability

Note. SE = Standard error of the unstandardized regression coefficient. *p < .05. **p < .01.

170

Table 35
Evaluating whether the relation between participants’ NPI traits and their dates’ perceptions of
them as Admiring is Moderated by participants’ perception of their dates’ desirability
b

SE

3.67**

.06

-0.26**

.06

Desirability

0.15**

.03

Leadership/Authority

0.11

.15

Grandiose Exhibitionism

0.24

.17

Entitlement/Exploitativeness

0.07

.16

Leadership/Authority by Gender

0.34*

.15

Grandiose Exhibitionism by Gender

0.34*

.17

Model Intercept
First-Order Effects
Gender

Two-Way Interactions between NPI traits and Gender

Entitlement/Exploitativeness by Gender

-0.04

.17

0.01

.03

-0.04

.11

Two-Way Interaction between Gender and Desirability
Gender by Desirability
Two-Way Interactions between NPI traits and Desirability
Leadership/Authority by Desirability
Grandiose Exhibitionism by Desirability

0.31*

.13

Entitlement/Exploitativeness by Desirability

0.09

.12

0.08

.11

-0.18

.13

0.03

.12

Three-Way Interactions for NPI traits, Desirability, and Gender
Leadership/Authority by Gender by Desirability
Grandiose Exhibitionism by Gender by Desirability
Entitlement/Exploitativeness by Gender by Desirability

Note. SE = Standard error of the unstandardized regression coefficient. *p < .05. **p < .01.

171

Table 36
Evaluating whether the relation between participants’ PNI traits and their dates’ perceptions of
them as Warm-Agreeable is Moderated by participants’ perception of their dates’ desirability
b
Model Intercept

SE

4.88**

.05

-0.16**

.06

Desirability

0.13**

.03

Narcissistic Grandiosity (NG)

0.10

.06

-0.05

.06

First-Order Effects
Gender

Narcissistic Vulnerability (NV)
Two-Way Interactions between PNI traits and Gender
NG by Gender

0.15*

NV by Gender

.06

-0.03

.06

0.02

.03

NG by Desirability

-0.04

.05

NV by Desirability

0.04

.04

-0.09†
0.03

.05

Two-Way Interaction between Gender and Desirability
Gender by Desirability
Two-Way Interactions between PNI traits and Desirability

Three-Way Interactions between PNI traits, Desirability, and Gender
NG by Gender by Desirability
NV by Gender by Desirability

.04

Note. SE = Standard error of the unstandardized regression coefficient. *p < .05. **p < .01. †p =
.051.

172

Table 37
Evaluating whether the relation between participants’ PNI traits and their dates’ perceptions of
them as Admiring is Moderated by participants’ perception of their dates’ desirability
b
Model Intercept

SE

3.67**

.06

-0.24**

.06

Desirability

0.13**

.03

Narcissistic Grandiosity (NG)

-0.00

.06

Narcissistic Vulnerability (NV)

-0.00

.05

First-Order Effects
Gender

Two-Way Interactions between PNI traits and Gender
NG by Gender

0.13*

NV by Gender

.06

-0.06

.05

0.03

.03

NG by Desirability

-0.02

.04

NV by Desirability

0.07

.04

NG by Gender by Desirability

-0.02

.04

NV by Gender by Desirability

0.03

.04

Two-Way Interaction between Gender and Desirability
Gender by Desirability
Two-Way Interactions between PNI traits and Desirability

Three-Way Interactions between PNI traits, Desirability, and Gender

Note. SE = Standard error of the unstandardized regression coefficient. *p < .05. **p < .01.

173

Table 38
Zero-order correlations between dates’ perceptions of the participants’ gregarious-extraversion,
assured-dominance, arrogance-calculation, desirability, and attractiveness
Variable

1

2

3

4

5

Dates’ Ratings of Participant
1. Perceived Gregarious-Extraversion
2. Perceived Assured-Dominance

.46**

3. Perceived Arrogance-Calculation

.15*

.31**

4. Perceived Desirability

.65**

.38**

.25**

5. Attraction to Participant

.54**

.31**

.29**

.81**

Note. df for each correlation = 173 (i.e., n – 2*number of groups – 1). *p < .05. **p < .01.

174

Table 39
Zero-order correlations between participants’ scores on the Narcissistic Personality Inventory (NPI) subscales, the Pathological
Narcissism Inventory (PNI) scales, and their Social Relations Model (SRM) Perceiver effects for the Rapport Dating Record Variables
NPI Scales

PNI Scales

Leadership/

Effects

Entitlement/

Authority
SRM Perceiver

Grandiose
Exhibitionism

Exploitativeness

Women

Men

Women

Men

-.02

.06

.08

.01

-.19

-.04

.12

Men

Disclosure

-.08

Trust

-.34**

Connection

.12

Grandiosity

Women

Men

Women

-.03

.09

.11

.11

.08

-.17

.22*

-.21

-.08

-.13

Note. *p < .05. **p < .01.

175

-.08

.18

-.26**
.23

Vulnerability

Men
.09
-.10
.07

Women
.18
-.38**
.00

Table 40
Zero-order correlations between participants’ scores on the Narcissistic Personality Inventory (NPI) subscales, the Pathological
Narcissism Inventory (PNI) scales, and their Social Relations Model (SRM) Target effects for the Rapport Dating Record Variables
NPI Scales

PNI Scales

Leadership/

Effects

Entitlement/

Authority
SRM Target

Grandiose
Exhibitionism

Exploitativeness

Women

Men

Women

Men

.38

nc
-.29

Men

Trust

.19

nc

Responsiveness

.34

.10

.57**

Grandiosity

Vulnerability

Women

Men

Women

Men

Women

.07

nc

.14

nc

-.08

nc

.23

-.10

.42*

-.09

.06

-.06

Note. Estimates of correlations obtained from BLOCKO and are disattenuated for error. nc = estimate not computed due to a lack of
significant target variance. *p < .05. **p < .01.

176

Table 41
Predicting Participants’ Disclosure, Trust, and Connection to their Dates in the Social
Interactions from the Participants’ Narcissistic Personality Inventory (NPI) Traits
Disclosure

Trust

a

Connection

b
Model Intercept

SE

b

SE

b

SE

4.12**

.07

5.50**

.07

3.49**

.07

First-Order Effects
Participant gender

-0.13

.08

-0.20*

.08

0.24**

.06

LA

-0.07

.32

-0.66*

.31

0.01

.24

GE

0.17

.36

0.23

.35

0.13

.26

EE

0.27

.36

0.58

.35

-0.22

.26

LA by Gender

0.03

.32

-0.24

.31

0.20

.23

GE by Gender

-0.05

.36

-0.05

.35

-0.01

.26

EE by Gender

-0.03

.36

-0.72*

.35

-0.03

.27

Interactions with Gender

Note. a = the women’s target variance for trust was fixed to zero in the random effects statement
due to minimal variance in this parameter. LA = Leadership/Authority. GE = Grandiose
Exhibitionism. EE = Entitlement/Exploitativeness. SE = Standard error of the unstandardized
regression coefficient. *p < .05. **p < .01.

177

Table 42
Predicting Dates’ Perceptions of Participants as Trusting and Responsive in the Social
Interactions from the Participants’ Narcissistic Personality Inventory (NPI) Traits
Dates’ Trust in
Participant
b
Model Intercept

a

SE

Perceived
Responsiveness
b
5.02**

SE

5.44**

.07

.06

0.19*

.07

-0.09

.05

First-Order Effects
Participant gender
LA

-0.04

.22

0.18

.16

GE

0.26

.24

0.04

.18

EE

0.04

.24

0.08

.18

-0.25

.22

-0.05

.16

Interactions with Gender
LA by Gender
GE by Gender
EE by Gender

0.70**
-0.00

.24

0.54**

.18

.24

0.11

.18

Note. a = the women’s perceiver variance for partners’ trust was fixed to zero in the random
effects statement due to minimal variance in this parameter. LA = Leadership/Authority. GE =
Grandiose Exhibitionism. EE = Entitlement/Exploitativeness. SE = Standard error of the
unstandardized regression coefficient. *p < .05. **p < .01.

178

Table 43
Predicting Participants’ Disclosure, Trust, and Connection to their Dates in the Social
Interactions from the Participants’ Pathological Narcissism Inventory (PNI) Traits
Disclosure

Trust

a

Connection

b

SE

b

SE

b

SE

4.10**

.07

5.48**

.08

3.52**

.07

Participant gender

-0.13

.08

-0.20*

.08

0.18**

.06

NG

0.14

.13

-0.03

.13

0.30**

.09

NV

0.14

.11

-0.17

.11

-0.09

.08

NG by Gender

0.08

.13

-0.07

.13

-0.11

.09

NV by Gender

0.01

.11

0.19

.11

0.04

.08

Model Intercept
First-Order Effects

Interactions with Gender

Note. a = the women’s target variance for trust was fixed to zero in the random effects statement
due to minimal variance in this parameter. NG = Narcissistic Grandiosity. NV = Narcissistic
Vulnerability. SE = Standard error of the unstandardized regression coefficient. *p < .05. **p <
.01.

179

Table 44
Predicting Dates’ Perceptions of Participants as Trusting and Responsive in the Social
Interactions from the Participants’ Pathological Narcissism Inventory (PNI) Traits
Dates’ Trust in
Participant

a

Perceived
Responsiveness

SE

b

SE

5.43**

.07

5.03**

.06

Participant gender

0.18*

.08

-0.10†

.05

NG

0.03

.09

0.09

.07

NV

-0.10

.08

-0.06

.06

NG by Gender

0.12

.09

0.11

.07

NV by Gender

-0.04

.08

-0.03

.06

b
Model Intercept
First-Order Effects

Interactions with Gender

Note. a = the women’s perceiver variance for partners’ trust was fixed to zero in the random
effects statement due to minimal variance in this parameter. NG = Narcissistic Grandiosity. NV
= Narcissistic Vulnerability. SE = Standard error of the unstandardized regression coefficient. *p
< .05. **p < .01. †p < .06.

180

Table 45
Predicting Participants’ Attraction towards Admiring Dates with the Participants’ Narcissistic
Personality Inventory (NPI) Traits
b
Model Intercept

SE

3.96**

.06

Participant gender

0.03

.07

Perceived Partner Admiration

0.56**

.04

First-Order Effects

Leadership/Authority

-0.13

.22

Grandiose Exhibitionism

-0.90**

.24

Entitlement/Exploitativeness

0.01

.26

Leadership/Authority by Gender

-0.14

.22

Grandiose Exhibitionism by Gender

-0.31

.24

Entitlement/Exploitativeness by Gender

-0.21

.26

-0.08*

.04

-0.13

.17

0.02

.18

-0.27

.18

-0.28

.16

Grandiose Exhibitionism by Gender by Admiration

0.11

.18

Entitlement/Exploitativeness by Gender by Admiration

0.06

.18

Two-Way Interactions between NPI traits and Gender

Two-Way Interaction between Gender and Admiration
Gender by Perceived Admiration
Two-Way Interactions between NPI traits and Admiration
Leadership/Authority by Perceived Admiration
Grandiose Exhibitionism by Perceived Admiration
Entitlement/Exploitativeness by Perceived Admiration
Three-Way Interactions for NPI traits, Gender, and Admiration
Leadership/Authority by Gender by Admiration

Note. SE = Standard error of the unstandardized regression coefficient. *p < .05. **p < .01.

181

Table 46
Predicting Participants’ Attraction towards Admiring Dates with the Participants’ Pathological
Narcissism Inventory (PNI) Traits
b
Model Intercept

3.92**

SE
.06

First-Order Effects
Gender

-0.02

.07

Admiration

0.51**

.04

Narcissistic Grandiosity (NG)

0.02

.09

-0.01

.08

NG by Gender

0.00

.09

NV by Gender

-0.03

.08

-0.14**

.04

Narcissistic Vulnerability (NV)
Two-Way Interactions between PNI traits and Gender

Two-Way Interaction between Gender and Admiration
Gender by Admiration
Two-Way Interactions between PNI traits and Admiration
NG by Admiration

0.08

.06

NV by Admiration

-0.14*

.06

NG by Gender by Admiration

0.04

.06

NV by Gender by Admiration

-0.03

.06

Three-Way Interactions for PNI traits, Admiration, and Gender

Note. SE = Standard error of the unstandardized regression coefficient. *p < .05. **p < .01.

182

Table 47
Predicting Participants’ Attraction towards Desirable Dates with the Participants’ Narcissistic
Personality Inventory (NPI) Traits
b

SE

3.91**

.04

Gender

0.07

.04

Desirability

0.91**

.02

Model Intercept
First-Order Effects

Leadership/Authority

-0.26

.15

Grandiose Exhibitionism

0.08

.17

Entitlement/Exploitativeness

0.04

.17

0.16

.15

Grandiose Exhibitionism by Gender

-0.07

.17

Entitlement/Exploitativeness by Gender

-0.16

.17

0.00

.02

-0.03

.10

Grandiose Exhibitionism by Desirability

0.02

.12

Entitlement/Exploitativeness by Desirability

0.17

.11

0.22*

.10

Two-Way Interactions between NPI traits and Gender
Leadership/Authority by Gender

Two-Way Interaction between Gender and Desirability
Gender by Perceived Desirability
Two-Way Interactions between NPI traits and Desirability
Leadership/Authority by Desirability

Three-Way Interactions for NPI traits, Desirability, and Gender
Leadership/Authority by Gender by Desirability
Grandiose Exhibitionism by Gender by Desirability
Entitlement/Exploitativeness by Gender by Desirability

-0.14

.12

0.03

.11

Note. The men’s target variance for attraction was fixed to zero in the random effects statement
due to minimal variance in this parameter. SE = Standard error of the unstandardized regression
coefficient. *p < .05. **p < .01.

183

Table 48
Predicting Participants’ Attraction towards Desirable Dates with the Participants’ Pathological
Narcissism Inventory (PNI) Traits
b

SE

3.89**

.04

Gender

0.05

.04

Desirability

0.92**

.02

Narcissistic Grandiosity (NG)

0.06

.06

-0.03

.05

NG by Gender

0.00

.06

NV by Gender

-0.05

.05

-0.00

.02

NG by Desirability

0.06

.04

NV by Desirability

-0.02

.03

NG by Gender by Desirability

-0.02

.04

NV by Gender by Desirability

-0.02

.03

Model Intercept
First-Order Effects

Narcissistic Vulnerability (NV)
Two-Way Interactions between PNI traits and Gender

Two-Way Interaction between Gender and Desirability
Gender by Perceived Desirability
Two-Way Interactions between PNI traits and Desirability

Three-Way Interactions for PNI traits, Desirability, and Gender

Note. The men’s target variance for attraction was fixed to zero in the random effects statement
due to minimal variance in this parameter. SE = Standard error of the unstandardized regression
coefficient. **p < .01.

184

Table 49
Predicting Participants’ Attraction towards Caring Dates with the Participants’ Narcissistic
Personality Inventory (NPI) Traits
b
Model Intercept

SE

3.92**

.06

Gender

0.08

.07

Caring

0.53**

.04

First-Order Effects

Leadership/Authority

-0.15

.22

Grandiose Exhibitionism

-0.14

.25

0.05

.26

0.13

.22

Grandiose Exhibitionism by Gender

-0.28

.25

Entitlement/Exploitativeness by Gender

-0.06

.25

-0.06

.04

-0.25

.18

0.30

.20

-0.08

.19

-0.04

.18

Entitlement/Exploitativeness
Two-Way Interactions between NPI traits and Gender
Leadership/Authority by Gender

Two-Way Interaction between Gender and Caring
Gender by Caring
Two-Way Interactions between NPI traits and Caring
Leadership/Authority by Caring
Grandiose Exhibitionism by Caring
Entitlement/Exploitativeness by Caring
Three-Way Interactions for NPI traits, Caring, and Gender
Leadership/Authority by Gender by Caring
Grandiose Exhibitionism by Gender by Caring

0.56**

Entitlement/Exploitativeness by Gender by Caring

-0.16

Note. SE = Standard error of the unstandardized regression coefficient. **p < .01.

185

.20
.19

Table 50
Predicting Participants’ Attraction towards Caring Dates with the Participants’ Pathological
Narcissism Inventory (PNI) Traits
b
Model Intercept

SE

3.92**

.06

Gender

0.02

.07

Caring

0.51**

.04

Narcissistic Grandiosity (NG)

0.22*

.09

First-Order Effects

Narcissistic Vulnerability (NV)

-0.06

.08

NG by Gender

-0.06

.09

NV by Gender

0.00

.08

-0.08

.04

NG by Caring

0.05

.07

NV by Caring

-0.04

.06

NG by Gender by Caring

0.09

.07

NV by Gender by Caring

-0.13*

.06

Two-Way Interactions between PNI traits and Gender

Two-Way Interaction between Gender and Caring
Gender by Caring
Two-Way Interactions between PNI traits and Caring

Three-Way Interactions for PNI traits, Caring, and Gender

Note. SE = Standard error of the unstandardized regression coefficient. *p < .05. **p < .01.

186

Figure 1. Higher order structure of Pathological Narcissism Inventory.

Narcissistic
Vulnerability

Narcissistic
Grandiosity

Grandiose
Fantasy

Exploitative

SelfSacrificing
Enhancement

Contingent
Self-Esteem

187

Devaluing

EntitlementRage

Hiding
the Self

Figure 2. Scree plot of eigenvalues from an exploratory factor analysis of items from the relationship initiation motives inventory.

188

Figure 3. Path analysis model for predicting the Relationship Initiation Motives Inventory subscales from the Narcissistic Personality
Inventory subscales.

Intimacy

Leadership/
Authority

e1

Coping/
Self-Affirmation

e2

Grandiose
Exhibitionism
Entitlement/
Exploitativeness

Hedonism

e3

Enhancement

e4

189

Figure 4. Path analysis model for predicting the Relationship Initiation Motives Inventory subscales from the Pathological Narcissism
Inventory subscales.

Intimacy

Grandiosity

Vulnerability

e1

Coping/
Self-Affirmation

e2

Hedonism

e3

Enhancement

e4

190

Figure 5. The Interpersonal Circumplex.

PA
Assured-Dominant
BC
ArrogantCalculating

NO
GregariousExtraverted

DE Cold-Hearted

Warm-Agreeable LM

UnassumingIngenious
JK

AloofIntroverted
FG
UnassuredSubmissive
HI

191

Figure 6. Diagram of a single-variable mediational model.

Panel A
e1
c

IV

DV

DV = i1 + cIV + e1
_______________________________________________________________
Panel B
e3

a

M

b
e2

IV

c′

DV

DV = i2 + c′ IV + bM + e2
M = i3 + aIV + e3
Note. IV = Independent variable. DV = Outcome variable. M = Mediating variable. Equations
taken from MacKinnon (2008, pp. 48-49).

192

Figure 7. Mediational Model for Grandiose Narcissism-Attraction Link.

e1
Participants’
Grandiose Narcissism

c

Dates’ Attraction to
Participant

e3

a

Dates’ Perception of
GregariousExtraversion

b
e2

Participants’
Grandiose Narcissism

c′

Dates’ Attraction to
Participant

193

APPENDICIES

194

APPENDIX 1: STUDY 1 MATERIALS
Description of Study to Participants
Participants MUST be SINGLE (that is, not currently involved in a romantic relationship) to
participate. Moreover, they MUST be at least 18 years old.
Materials for Study 1
Narcissistic Personality Inventory-40
Read each pair of statements and then choose the one that is closer to your own feelings and
beliefs. Indicate your answer by circling the letter "A" or "B" to the left of each item. Please do
not skip any items.
1.

A I have a natural talent for influencing people. [Leadership/Authority]
B I am not good at influencing people.

2.

A Modesty doesn't become me.
B I am essentially a modest person.

3.

A I would do almost anything on a dare.
B I tend to be a fairly cautious person.

4.

A When people compliment me I sometimes get embarrassed.
B I know that I am good because everybody keeps telling me so. [Grandiose
Exhibitionism]

5.

A The thought of ruling the world frightens the hell out of me.
B If I ruled the world it would be a much better place. [Leadership/Authority]

6.

A I can usually talk my way out of anything.
B I try to accept the consequences of my behavior.

7.

A I prefer to blend in with the crowd.
B I like to be the center of attention. [Grandiose Exhibitionism]

8.

A I will be a success.
B I am not too concerned about success.
195

9.

A I am no better or no worse than most people.
B I think I am a special person.

10.

A I am not sure if I would make a good leader.
B I see myself as a good leader. [Leadership/Authority]

11.

A I am assertive. [Leadership/Authority]
B I wish I were more assertive.

12.

A I like having authority over people. [Leadership/Authority]
B I don't mind following orders.

13.

A I find it easy to manipulate people. [Entitlement/Exploitativeness]
B I don't like it when I find myself manipulating people.

14.

A I insist upon getting the respect that is due me. [Entitlement/Exploitativeness]
B I usually get the respect that I deserve.

15.

A I don't particularly like to show off my body.
B I like to display my body. [Grandiose Exhibitionism]

16.

A I can read people like a book.
B People are sometimes hard to understand.

17.

A If I feel competent I am willing to take responsibility for making decisions.
B I like to take responsibility for making decisions.

18.

A I just want to be reasonably happy.
B I want to amount to something in the eyes of the world.

19.

A My body is nothing special.
B I like to look at my body. [Grandiose Exhibitionism]

20.

A I try not to be a show off.
B I am apt to show off if I get the chance. [Grandiose Exhibitionism]

21.

A I always know what I am doing.
196

B Sometimes I am not sure of what I am doing.
22.

A I sometimes depend on people to get things done.
B I rarely depend on anyone else to get things done.

23.

A Sometimes I tell good stories.
B Everybody likes to hear my stories.

24.

A I expect a great deal from other people. [Entitlement/Exploitativeness]
B I like to do things for other people.

25.

A I will never be satisfied until I get all that I deserve. [Entitlement/Exploitativeness]
B I take my satisfactions as they come.

26.

A Compliments embarrass me.
B I like to be complimented. [Grandiose Exhibitionism]

27.

A I have a strong will to power. [Leadership/Authority]
B Power for its own sake doesn't interest me.

28.

A I don't very much care about new fads and fashions.
B I like to start new fads and fashions. [Grandiose Exhibitionism]

29.

A I like to look at myself in the mirror. [Grandiose Exhibitionism]
B I am not particularly interested in looking at myself in the mirror.

30.

A I really like to be the center of attention. [Grandiose Exhibitionism]
B It makes me uncomfortable to be the center of attention.

31.

A I can live my life in any way I want to.
B People can't always live their lives in terms of what they want.

32.

A Being an authority doesn't mean that much to me.
B People always seem to recognize my authority. [Leadership/Authority]

33.

A I would prefer to be a leader. [Leadership/Authority]
B It makes little difference to me whether I am a leader or not.
197

34.

A I am going to be a great person. [Leadership/Authority]
B I hope I am going to be successful.

35.

A People sometimes believe what I tell them.
B I can make anybody believe anything I want them to.

36.

A I am a born leader. [Leadership/Authority]
B Leadership is a quality that takes a long time to develop.

37.

A I wish somebody would someday write my biography.
B I don't like people to pry into my life for any reason.

38.

A I get upset when people don't notice how I look when I go out in public. [Grandiose
Exhibitionism]
B I don't mind blending into the crowd when I go out in public.

39.

A I am more capable than other people.
B There is a lot that I can learn from other people.

40.

A I am much like everybody else.
B I am an extraordinary person. [Leadership/Authority]

Pathological Narcissism Inventory
Instructions: Below you will find 52 descriptive statements. Please consider each one and
indicate how well that statement describes you. There are no right or wrong answers. On your
scantron sheet, fill in only one answer. Simply indicate how well each statement describes you
as a person on the following 6-point scale:
0
Not at all
Like me

1
Moderately
Unlike me

2
A little
Unlike me

3
A little
Like me

4
Moderately
Like me

5
Very much
Like me

1. I often fantasize about being admired and respected. [Grandiose Fantasy]
2. My self-esteem fluctuates a lot. [Contingent Self-Esteem]

198

3. I sometimes feel ashamed about my expectations of others when they disappoint me.
[Devaluing]
4. I can usually talk my way out of anything. [Exploitativeness]
5. It’s hard for me to feel good about myself when I’m alone. [Contingent Self-Esteem]
6. I can make myself feel good by caring for others. [Self-Sacrificing Self-Enhancement]
7. I hate asking for help. [Hiding the Self]
8. When people don’t notice me, I start to feel bad about myself. [Contingent Self-Esteem]
9. I often hide my needs for fear that others will see me as needy and dependent. [Hiding
the Self]
10. I can make anyone believe anything I want them to. [Exploitativeness]
11. I get mad when people don’t notice all that I do for them. [Entitlement-Rage]
12. I get annoyed by people who are not interested in what I say or do. [Entitlement-Rage]
13. I wouldn’t disclose all my intimate thoughts and feelings to someone I didn’t admire.
[Hiding the Self]
14. I often fantasize about having a huge impact on the world around me. [Grandiose
Fantasy]
15. I find it easy to manipulate people. [Exploitativeness]
16. When others don’t notice me, I start to feel worthless. [Contingent Self-Esteem]
17. Sometimes I avoid people because I’m concerned that they’ll disappoint me. [Devaluing]
199

18. I typically get very angry when I’m unable to get what I want from others. [EntitlementRage]
19. I sometimes need important others in my life to reassure me of my self-worth.
[Contingent Self-Esteem]
20. When I do things for other people, I expect them to do things for me. [EntitlementRage]
21. When others don’t meet my expectations, I often feel ashamed about what I wanted.
[Devaluing]
22. I feel important when others rely on me. [Self-Sacrificing Self-Enhancement]
23. I can read people like a book. [Exploitativeness]
24. When others disappoint me, I often get angry at myself. [Devaluing]
25. Sacrificing for others makes me the better person. [Self-Sacrificing Self-Enhancement]
26. I often fantasize about accomplishing things that are probably beyond my means.
[Grandiose Fantasy]
27. Sometimes I avoid people because I’m afraid they won’t do what I want them to do.
[Devaluing]
28. It’s hard to show others the weaknesses I feel inside. [Hiding the Self]
29. I get angry when criticized. [Entitlement-Rage]
30. It’s hard to feel good about myself unless I know other people admire me. [Contingent
Self-Esteem]

200

31. I often fantasize about being rewarded for my efforts. [Grandiose Fantasy]
32. I am preoccupied with thoughts and concerns that most people are not interested in me.
[Contingent Self-Esteem]
33. I like to have friends who rely on me because it makes me feel important. [SelfSacrificing Self-Enhancement]
34. Sometimes I avoid people because I’m concerned they won’t acknowledge what I do for
them. [Devaluing]
35. Everybody likes to hear my stories. [Exploitativeness]
36. It’s hard for me to feel good about myself unless I know other people like me.
[Contingent Self-Esteem]
37. It irritates me when people don’t notice how good a person I am. [Entitlement-Rage]
38. I will never be satisfied until I get all that I deserve. [Entitlement-Rage]
39. I try to show what a good person I am through my sacrifices. [Self-Sacrificing SelfEnhancement]
40. I am disappointed when people don’t notice me. [Contingent Self-Esteem]
41. I often find myself envying others’ accomplishments. [Contingent Self-Esteem]
42. I often fantasize about performing heroic deeds. [Grandiose Fantasy]
43. I help others in order to prove I’m a good person. [Self-Sacrificing Self-Enhancement]
44. It’s important to show people I can do it on my own even if I have some doubts inside.
[Hiding the Self]

201

45. I often fantasize about being recognized for my accomplishments. [Grandiose Fantasy]
46. I can’t stand relying on other people because it makes me feel weak. [Hiding the Self]
47. When others don’t respond to me the way that I would like them to, it is hard for me to
still feel ok with myself. [Contingent Self-Esteem]
48. I need others to acknowledge me. [Contingent Self-Esteem]
49. I want to amount to something in the eyes of the world. [Grandiose Fantasy]
50. When others get a glimpse of my needs, I feel anxious and ashamed. [Hiding the Self]
51. Sometimes it’s easier to be alone than to face not getting everything I want from other
people. [Devaluing]
52. I can get pretty angry when others disagree with me. [Entitlement-Rage]

Experience in Close Relationships Scale (ECR)
Instructions: The following statements concern how you generally feel in close relationships
(e.g., with romantic partners, close friends, or family members). Respond to each statement by
indicating how much you agree or disagree with it. Write the number in the space provided,
using the following rating scale:

1
Disagree
Strongly

2

3

Disagree
Disagree
Slightly

4

5

Neutral/

Agree
Slightly

Mixed

___ 1. I prefer not to show others how I feel deep down. [Avoidance]
___ 2. I worry about being rejected or abandoned.
202

6

7
Agree

Agree
Strongly

___ 3. I am very comfortable being close to other people. [Avoidance]
___ 4. I worry a lot about my relationships.
___ 5. Just when someone starts to get close to me I find myself pulling away. [Avoidance]
___ 6. I worry that others won’t care about me as much as I care about them.
___ 7. I get uncomfortable when someone wants to be very close to me. [Avoidance]
___ 8. I worry a fair amount about losing my close relationship partners.
___ 9. I don’t feel comfortable opening up to others. [Avoidance]
___ 10. I often wish that close relationship partners’ feelings for me were as strong as my
feelings for them.
___ 11. I want to get close to others, but I keep pulling back. [Avoidance]
___ 12. I want to get very close to others, and this sometimes scares them away.
___ 13. I am nervous when another person gets too close to me. [Avoidance]
___ 14. I worry about being alone.
___ 15. I feel comfortable sharing my private thoughts and feelings with others. [Avoidance]
___ 16. My desire to be very close sometimes scares people away.
___ 17. I try to avoid getting too close to others. [Avoidance]
___ 18. I need a lot of reassurance that close relationship partners really care about me.
___ 19. I find it relatively easy to get close to others. [Avoidance]
___ 20. Sometimes I feel that I try to force others to show more feeling, more commitment to our
relationship than they otherwise would.
___ 21. I find it difficult to allow myself to depend on close relationship partners. [Avoidance]
___ 22. I do not often worry about being abandoned.
___ 23. I prefer not to be too close to others. [Avoidance]
___ 24. If I can’t get a relationship partner to show interest in me, I get upset or angry.
___ 25. I tell my close relationship partners just about everything. [Avoidance]
___ 26. I find that my partners don’t want to get as close as I would like.
203

___ 27. I usually discuss my problems and concerns with close others. [Avoidance]
___ 28. When I don’t have close others around, I feel somewhat anxious and insecure.
___ 29. I feel comfortable depending on others. [Avoidance]
___ 30. I get frustrated when my close relationship partners are not around as much as I would
like.
___ 31. I don’t mind asking close others for comfort, advice, or help. [Avoidance]
___ 32. I get frustrated if relationship partners are not available when I need them.
___ 33. It helps to turn to close others in times of need. [Avoidance]
___ 34. When other people disapprove of me, I feel really bad about myself.
___ 35. I turn to close relationship partners for many things, including comfort and reassurance.
[Avoidance]
___ 36. I resent it when my relationship partners spend time away from me.
Demographics
1) What is your gender? (1=Male; 2=Female)
2) How old are you? (1 = Less than 18 years old; 2 = 18 years old; 3 = 19 years old; 4 = 20
years old; 5 = 21 years old or older)
3) How would you describe your ethnicity? (1 = Black or African American; 2 = White (Not
Hispanic/Not Latino); 3 = Hispanic or Latino; 4 = American Indian; 5 = Asian; 6 =
Native Hawaiian or Pacific Islander; 7 = Multiracial (Having parents of more than one
race)
4) What is your present religious orientation? (1=Protestant/Christian; 2=Jewish;
3=Catholic; 4=Islamic; 5=No affiliation; 6=Other)
5) What is your current relationship status? (1 = Single (not dating); 2 = Dating casually; 3
= Dating exclusively; 4 = Engaged; 5 = Married; 6 = Separated; 7 = Living with
significant other; 8 = Divorced; 9 = Widowed)
6) Have you ever been divorced? (1=Yes; 2=No)
7) How would you describe your political orientation? (1=Very Conservative;
2=Conservative; 3=Slightly Conservative; 4=Moderate (partly Conservative and partly
Liberal); 5=Slightly Liberal; 6=Liberal; 7=Very Liberal).
8) How would you describe your sexual orientation? (1=Heterosexual, 2=Bisexual,
3=Homosexual)
9) I answered these questions honestly: (1=Yes; 2=No)
204

Relationship Initiation Motives Measure: (Adapted from Cooper, Shapiro, & Powers, 1998)
Instructions: Below is a series of statements that describe reasons that people may have for
beginning romantic relationships. For each statement, please use the following scale to indicate
how much you agree with each of the statements.
Response Scale:
1 = Strongly Disagree
2 = Somewhat Disagree
3 = Neither agree nor disagree
4 = Somewhat Agree
5 = Strongly Disagree
I often begin romantic relationships with new partners …
1) Because I want to be intimate with the other person [Old Intimacy Motive]
2) Because I feel “horny” [Hedonism Motive]
3) To prove to myself that the other person thinks I’m attractive [Self-Affirmation
Motive]
4) To cope with upset feelings [Coping Motive]
5) To express my affection to the other person [Old Intimacy Motive]
6) Because it feels good [Hedonism Motive]
7) Because it makes me feel like I’m a more interesting person [Self-Affirmation
Motive]
8) To help me deal with disappointment in my life [Coping Motive]
9) To make an emotional connection with the other person [Old Intimacy Motive]
10) Just for the excitement of it [Hedonism Motive]
11) Because it makes me feel more self-confident [Self-Affirmation Motive]
12) Because it helps me feel better when I’m lonely [Coping Motive]
13) To become closer with the other person [Old Intimacy Motive]
14) Just for the thrill of it [Hedonism Motive]
15) To reassure myself that I am desirable [Self-Affirmation Motive]
16) Because it helps me feel better when I’m feeling low [Coping Motive]
17) To feel emotionally close to the other person [Old Intimacy Motive]
18) To satisfy my sexual needs [Hedonism Motive]
19) To help me feel better about myself [Self-Affirmation Motive]
20) To cheer myself up [Coping Motive]
21) To be admired [Enhancement Motive]
22) To impress others [Enhancement Motive]
23) To receive compliments [Enhancement Motive]
205

24) Because they remind me I’m a great person [Enhancement Motive]
25) To have other person treat me like the center of his/her world [Enhancement
Motive]
26) To better understand the other person [New Intimacy Motive]
27) Because I want to open up to the other person [New Intimacy Motive]
28) To feel close with the other person [New Intimacy Motive]
29) To have another person truly understand me [New Intimacy Motive]
30) Because I want to feel connected to the other person [New Intimacy Motive]
The Remote Associations Test (RAT) (from Bowden & Jung-Beeman, 2003)
Condition

RAT item

Answer

(10 easy items)

Cream/skate/water
Cane/daddy/plum
Political/surprise/line
Night/wrist/stop
Duck/fold/dollar
Rocking/wheel/high
Dew/comb/bee
Fountain/baking/pop
Preserve/ranger/tropical
Measure/worm/video

Ice
Sugar
Party
Watch
Bill
Chair
Honey
Soda
Forest
Tape

(7 difficult items)

Land/hand/house
Hungry/order/belt
Forward/flush/razor
Way/ground/weather
Cast/side/jump
Reading/service/stick
Over/plant/horse

Farm
Money
Straight
Fair
Broad
Lip
Power

(7 moderate difficulty items)

Dust/cereal/fish
Palm/shoe/house
Wheel/hand/shopping
Mill/tooth/dust
Fly/clip/wall
Wagon/break/radio
Dress/dial/flower

Bowl
Tree
Cart
Saw
Paper
Station
Sun

206

The State Self-Esteem Scale (from Heatherton & Polivy, 1991)
Instructions: This is a questionnaire designed to measure what you are thinking at this moment.
There is, of course, no right answer for any statement. The best answer is what you feel is true of
yourself at this moment. Be sure to answer all of the items, even if you are not certain of the best
answer. Again, answer these questions as they are true for you RIGHT NOW.
Each item is scored on a 5-point scale (1 = not at all, 2 = a little bit, 3 = somewhat, 4 = very
much, and 5 = extremely).
1) I feel confident about my abilities
2) I am worried about whether I am regarded as a success or failure (Reverse)
3) I feel satisfied with the way my body looks right now
4) I feel frustrated or rattled about my performance (Reverse)
5) I feel that I am having trouble understanding things that I read (Reverse)
6) I feel that others respect and admire me
7) I am dissatisfied with my weight (Reverse)
8) I feel self-conscious (Reverse)
9) I feel as smart as others
10) I feel displeased with myself (Reverse)
11) I feel good about myself
12) I am pleased with my appearance right now
13) I am worried about what other people think of me (Reverse)
14) I feel confident that I understand things
15) I feel inferior to others at this moment (Reverse)
16) I feel unattractive (Reverse)
17) I feel concerned about the impression I am making (Reverse)
18) I feel that I have less scholastic ability right now than others (Reverse)
19) I feel like I’m not doing well (Reverse)
20) I am worried about looking foolish (Reverse)

207

Romantic Targets Stimuli (from Campbell, 1999, dissertation)
Instructions: Please read the following passage carefully. You will be asked some questions
related to the passage afterward.
Imagine the following situation: You are at a party with some people you know and some
people you don’t know. A friend of yours introduces you to three people of the opposite
sex who are single—we’ll call them by their initials: J.G., C.F., and B.W. Later, you run
into your friend in the kitchen and she tells you about these people in more detail. Your
friend describes the three single people you met— J.G., C.F., and B.W.—as follows…
Participants are then presented with the three following person descriptions in random order:
Person Description 1
J.G. told me that he/she thought you were terrific! J.G. thought you were the best looking person
at the party and that you were very charming. J.G. also thought you were one of the most
intelligent people he/she had ever met.
Person Description 2
C.F. is really a great person. C.F. is very popular and good-looking. He/she is also one of the top
students at the school and is rated as one of the best soccer players in the nation.
Person Description 3
B.W. is a really caring person. B.W. told me that he/she’s looking for a person to date with
whom he/she could be close, intimate friends. B.W. is one of the most sensitive, sharing,
compassionate, and friendly people I know.
Instructions: Please answer the following questions based on what your friend in the kitchen
told you about J.G.
1) How attractive do you find this person?
1 (not at all) 2 3 4
5
6
7 (very)
2) How desirable would you find this person as a dating partner?
1 (not at all) 2 3 4
5
6
7 (very)
3) How much would you actually like to date this person?
1 (not at all) 2 3 4
5
6
7 (very)
4) How would you feel about yourself if you were dating this person?
1 (very bad) 2 3 4
5
6
7 (very good)
5) How do you think your friends would feel about you if you were dating this person?
1 (disapproving of me) 2 3 4
5
6
7 (approving of me)

208

Instructions: Please answer the following questions based on what your friend in the kitchen
told you about C.F.
1) How attractive do you find this person?
1 (not at all) 2 3 4
5
6
7 (very)
2) How desirable would you find this person as a dating partner?
1 (not at all) 2 3 4
5
6
7 (very)
3) How much would you actually like to date this person?
1 (not at all) 2 3 4
5
6
7 (very)
4) How would you feel about yourself if you were dating this person?
1 (very bad) 2 3 4
5
6
7 (very good)
5) How do you think your friends would feel about you if you were dating this person?
1 (disapproving of me) 2 3 4
5
6
7 (approving of me)
Instructions: Please answer the following questions based on what your friend in the kitchen
told you about B.W.
1) How attractive do you find this person?
1 (not at all) 2 3 4
5
6
7 (very)
2) How desirable would you find this person as a dating partner?
1 (not at all) 2 3 4
5
6
7 (very)
3) How much would you actually like to date this person?
1 (not at all) 2 3 4
5
6
7 (very)
4) How would you feel about yourself if you were dating this person?
1 (very bad) 2 3 4
5
6
7 (very good)
5) How do you think your friends would feel about you if you were dating this person?
1 (disapproving of me) 2 3 4
5
6
7 (approving of me)
Instructions: Please answer the following question based on what you have heard about B.W.,
C.F., and J.G.
1) Which of the people would you most like to meet? (please choose one)
a. J.G.
b. C.F.
c. B.W.

209

APPENDIX 2: STUDY 2 MATERIALS
Narcissistic Personality Inventory-40 (same scale as Study 1)
Pathological Narcissism Inventory (same scale as Study 1)
Demographics (same questions as Study 1)
Ideal Partner-Scale (developed for current study)
Instructions: Please use the following scale to describe how well the following characteristics
describe your ideal romantic partner
Response Scale:
1 = Extremely Inaccurate
2
3
4 = Somewhat Accurate
5
6
7 = Extremely Accurate
1) Admires me
2) Is cunning
3) Is interested in me
4) Is good looking
5) Is unsympathetic
6) Loves my company
7) Is charming
8) Is outgoing
9) Is unaggressive
10) Is popular
11) Is friendly
12) Is caring
13) Is tenderhearted
14) Is dominant
15) Is warmthless
16) Is sly
17) Is assertive
18) Is shy
19) Is compassionate
20) Is narcissistic
210

21) Is gentlehearted

211

Interaction Record
Please circle your answer to the following questions on the person with whom you just
interacted.
Definitely
No
1. Do you like this person?

Definitely
Yes

Maybe

1

2

3

4

5

6

7

1

2

3

4

5

6

7

1

2

3

4

5

6

7

4. Do you think this person is
desirable?

1

2

3

4

5

6

7

5. Do you think this person is attracted
to you?

1

2

3

4

5

6

7

1

2

3

4

5

6

7

1

2

3

4

5

6

7

8. Did you bond with your interaction
partner?

1

2

3

4

5

6

7

9. Were there sparks between you and
your partner?

1

2

3

4

5

6

7

10. Were you skeptical of your
interaction partner?

1

2

3

4

5

6

7

2. Are you attracted to this person?
3. Would you like to date this person?

6. Did you feel a real connection?
7. Was there ‘sexual’ chemistry?

Very
Little
11. I disclosed…

1

2
212

A Great
Deal

Somewhat
3

4

5

6

7

12. My partner disclosed…

1

2

3

4

5

6

7

13. How much did your partner
understand you?

1

2

3

4

5

6

7

14. Did the other person see you as
acceptable?

1

2

3

4

5

6

7

Instructions: Please rate how this person behaved with you today using the rating scales
provided below.
Submissive
1
2

3

4

5

6

7

8

9

Dominant
10
11

Submissive behavior involves being passive and looking to others to take control or lead.
Dominant behavior involves being assertive, controlling, aggressive, and in charge.
Cold
1

2

3

4

5

6

7

8

9

10

Warm
11

Cold behavior involves being uninterested in others, unfriendly, and even quarrelsome.
Warm behavior involves being kind, friendly, and close to others.

213

Interaction Record
Please circle your answer to the following statements on the person with whom you just
interacted.
My Interaction Partner…
1. admires me
2. is cunning
3. is interested in me
4. is good looking
5. is unsympathetic
6. loves my company
7. is charming
8. is outgoing
9. is unaggressive
10. is popular
11. is friendly
12. is caring
13. is tenderhearted
14. is dominant

Somewhat
Accurate

Extremely
Inaccurate

Extremely
Accurate

1

2

3

4

5

6

7

1

2

3

4

5

6

7

1

2

3

4

5

6

7

1

2

3

4

5

6

7

1

2

3

4

5

6

7

1

2

3

4

5

6

7

1

2

3

4

5

6

7

1

2

3

4

5

6

7

1

2

3

4

5

6

7

1

2

3

4

5

6

7

1

2

3

4

5

6

7

1

2

3

4

5

6

7

1

2

3

4

5

6

7

1

2

3

4

5

6

7

214

15. is warmthless
16. is sly
17. is assertive
18. is shy
19. is compassionate
20. is narcissistic
21. is gentlehearted

1

2

3

4

5

6

7

1

2

3

4

5

6

7

1

2

3

4

5

6

7

1

2

3

4

5

6

7

1

2

3

4

5

6

7

1

2

3

4

5

6

7

1

2

3

4

5

6

7

215

APPENDIX 3: SCRIPT USED IN STUDY 2
Script used in Study 2 (note that the following script borrows heavily from the script used by Eli
Finkel):
Hello, and welcome to the Personality in Social Interactions Study. My name is ____,
and I’m one of the research assistants in the RASP Lab. We’re very happy to see you
guys here.
We are interested in how the romantic attraction process works. In the field of close
relationships, researchers have become increasingly interested in the processes that occur
in short-meetings between persons—situations like speed-dating, for example. For this
second part of the study, we have therefore chosen to use a set-up that is in many ways
similar to speed-dating and we ask you to treat each of the interactions you have tonight
as an actual speed-date. Please note, however, that we will not be providing you with
anyone’s phone number or other contact information after the event.
You will notice that we are asking you to complete a number of questionnaires as part of
the study. If you guys take the questionnaires seriously, this study will be very
informative. As such, I hope you view yourself the way I view you: not as a research
subject, but rather as a collaborator on a major research project. We simply couldn’t
conduct this research without your active collaboration.
OK, here’s a brief overview of what will happen here this evening. You will meet every
opposite-sex speed-dater here today for 5 minutes each. After each speed-date, you will
be given about 5 minutes to fill out a brief questionnaire about it. We will warn you when

216

there is a minute left for each date and then ring a bell after each date to signal that time
is up and it’s time to move on. The entire event will take approximately 2 hours.
You will notice that all of you are wearing name-tags with unique numbers. The
numbers will be how you refer to one another on the questionnaires. Women [or men]
will remain seated at the tables situated around the room; men [or women] will rotate
clockwise after each round until you’ve had a chance to date everybody here.
Now let’s talk about the questionnaires. You will see that all the questionnaires are the
same—each is two pages long and they ask you to answer some questions about the
interaction that you just had. First, put in the number of the name-tag of the person you
just interacted with at the top. Then proceed to fill out the questions. Please be as honest
as possible on these questionnaires—no one besides the research team will ever see what
you put down here.
Are there any questions about how the process will work?
As you can see, this room is set up with a group of desks in the center and a group of
desks along with sides of the room. In a moment, I will be asking each of the women to
take a seat at one of the desks along the sides of the room. I will then be asking each of
the men to take a seat with one of the women. After the five minutes is up, I will ask the
men to take a seat back at the center of the room. At this point, I will ask you to complete
the interaction record on the person that you were just with. After everyone has
completed the interaction record, I will ask the men in the center of the room to rotate
clockwise to the next woman along the side of the room. You will continue this process
until each woman has interacted with each man. Does anyone have any questions on this
217

process?

218

REFERENCES

219

REFERENCES
Ackerman, R. A., & Donnellan, M. B. (2011). The Narcissistic Personality Inventory
Entitlement/Exploitativeness subscale: Further evidence of its psychometric properties,
validity, and research utility. Unpublished Manuscript.
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, 67-87.
Aron, A., Melinat, E., Aron, E. N., Vallone, R. D., & Bator, R. J. (1997). The experimental
generation of interpersonal closeness: A procedure and some preliminary findings.
Personality and Social Psychology Bulletin, 23, 363-377.
Back, M. D., Schmukle, S. C., & Egloff, B. (2010). Why are narcissists so charming at first
sight? Decoding the narcissism-popularity link at zero-acquaintance. Journal of
Personality and Social Psychology, 98, 132-145.
Bakan, D. (1966). The duality of human existence. Chicago: Rand McNally.
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social
psychological research: Conceptual, strategic, and statistical considerations. Journal of
Personality and Social Psychology, 51, 1173-1182.
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 and
Adolescent Psychology, 32 (1), 139-152.
Bauer, D. J., Preacher, K. J., & Gil, K. M. (2006). Conceptualizing and testing random indirect
effects and moderated mediation in multilevel models: New procedures and
recommendations. Psychological Methods, 11, 142-163.
Bishop, J., & Lane, R. C. (2002). The dynamics and dangers of entitlement. Psychoanalytic
Psychology, 19, 739-758.
Bowden, E. M., & Jung-Beeman, M. (2003). Normative data for 144 compound remote associate
problems. Behavior Research Methods, Instruments, & Computers, 35, 634-639.
Bowlby, J. (1988). A Secure Base: Parent-Child Attachment and Healthy Human Development.
New York: Basic Books.
Bredow, C. A., Cate, R. M., & Huston, T. L. (2008). Have we met before? A conceptual model
of first romantic encounters. In S. Sprecher, A. Wenzel, & J. Harvey (Eds.), Handbook of
Relationship Initiation. New York: Psychology Press.
Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult romantic
220

attachment: An integrative overview. In J. A. Simpson & W. S. Rholes (Eds.),
Attachment theory and close relationships (pp. 46-76). New York: Guilford Press.
Bresin, K., & Gordon, K. H. (2011). Characterizing pathological narcissism in terms of the
HEXACO model of personality. Journal of Psychopathology Behavioral Assessment, 33,
228-235.
Brown, R. P., Budzek, K., & Tamborski, M. (2009). On the meaning and measure of narcissism.
Personality and Social Psychology Bulletin, 35, 951-964.
Brown, M. B., & Forsythe, A. B. (1974a). Robust tests for equality of variance. Journal of the
American Statistical Association, 69, 364-367.
Brown, M. B., & Forsythe, A. B. (1974b). The small sample behavior of some statistics which
test the equality of several means. Technometrics, 16, 129-132.
Brunswick, E. (1947). Systematic and representative design of psychological experiments.
Berkeley: University of California Press.
Buss, D. M., & Shackelford, T. K. (1997). Susceptibility to infidelity in the first year of
marriage. Journal of Research in Personality, 31 (2), 193-221.
Cain, N. M., Pincus, A. L., & Ansell, E. B. (2008). Narcissism at the crossroads: Phenotypic
description of pathological narcissism across clinical theory, social/personality
psychology, and psychiatric diagnosis. Clinical Psychology Review, 28, 638-656.
Campbell, W. K. (1998). Narcissism and romantic attraction. Dissertation Abstracts
International, 58, 12B, p. 6866.
Campbell, W. K. (1999). Narcissism and romantic attraction. Journal of Personality and Social
Psychology, 77, 1254-1270.
Campbell, W. K., Brunell, A. B., & Finkel, E. J. (2006). Narcissism, interpersonal selfregulation, and romantic relationships: An agency model approach. In E. J. Finkel & K.
D. Vohs (Eds.), Intrapersonal processes, interpersonal relationships. New York:
Guilford.
Campbell, W. K., & Foster, C. A. (2002). Narcissism and commitment in romantic relationships:
An investment model analysis. Personality and Social Psychology Bulletin, 28, 484-495.
Campbell, W. K., Foster, C. A., & Finkel, E. J. (2002). Does self-love lead to love for others? A
story of narcissistic game playing. Journal of Personality and Social Psychology, 83,
340-354.
Carver, C. S., & White, T. L. (1994). Behavioral inhibition, behavioral activation, and affective
responses to impending reward and punishment: The BIS/BAS scales. Journal of
221

Personality and Social Psychology, 67, 319-333.
Clark, C. L., Shaver, P. R., & Abrahams, M. F. (1999). Strategic behaviors in romantic
relationship initiation. Personality and Social Psychology Bulletin, 25, 707-720.
Cooper, M. L., Shapiro, C. M., & Powers, A. M. (1998). Motivations for sex and risky sexual
behavior among adolescents and young adults: A functional perspective. Journal of
Personality and Social Psychology, 75, 1528-1558.
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.
Crocker, J., Thompson, L. L., McGraw, K. M., & Ingerman, C. (1987). Downward comparison,
prejudice, and evaluations of others: Effects of self-esteem and threat. Journal of
Personality and Social Psychology, 52, 907-916.
Dickinson, K. A., & Pincus, A. L. (2003). Interpersonal analysis of grandiose and vulnerable
narcissism. Journal of Personality Disorders, 17, 188-207.
del Rosario, P. M., & White, R. M. (2005). The Narcissistic Personality Inventory: Test-retest
stability and internal consistency. Personality and Individual Differences, 39, 1075-1081.
Eagly, A. H., & Steffen, V. J. (1984). Gender stereotypes stem from the distribution of women
and men into social roles. Journal of Personality and Social Psychology, 46, 735-754.
Eastwick, P. W., & Finkel, E. J. (2008). Speed-dating: A powerful and flexible paradigm for
studying romantic relationship initiation. In S. Sprecher, A. Wenzel, & J. Harvey (Eds.),
Handbook of Relationship Initiation, New York: Taylor & Francis Group.
Ellis, H. (1898). Auto-eroticism: A psychological study. Alienist and Neurologist, 19, 260-299.
Emmons, R. A. (1984). Factor analysis and construct validity of the narcissistic personality
inventory. Journal of Personality Assessment, 48 (3), 291-300.
Emmons, R. A. (1987). Narcissism: Theory and measurement. Journal of Personality and Social
Psychology, 52 (1), 11-17.
Emmons, R. A. (1989). Exploring the relations between motives and traits: The case of
narcissism. In D. M. Buss & N. Cantor (Eds). Personality Psychology: Recent Trends
and Emerging Directions. New York: Springer-Verlag Publishing.
Fabrigar, L. R., Wegener, D. T., MacCallum, R. C., & Strahan, E. J. (1999). Evaluating the use
of exploratory factor analysis in psychological research. Psychological Methods, 4, 272299.
Finkel, E. J., & Eastwick, P. W. (2009). Arbitrary social norms influence sex differences in
222

romantic selectivity. Psychological Science, 20, 1290-1295.
Finkel, E. J., Eastwick, P. W., & Matthews, J. (2007). Speed-dating as an invaluable tool for
studying romantic attraction: A methodological primer. Personal Relationships, 14, 149166.
Foster, J. D., & Trimm IV, R. F. (2008). On being eager and uninhibited: Narcissism and
approach-avoidance motivation. Personality and Social Psychology Bulletin, 34 (7),
1004-1017.
Freud, S. (1914). On narcissism: An introduction. In A. P. Morrison (Ed.), Essential Papers on
Narcissism (1986). New York: New York University Press.
Friedman, J. N. W., Oltmanns, T. F., Gleason, M. E. J., & Turkheimer, E. (2006). Mixed
impressions: Reactions of strangers to people with pathological personality traits. Journal
of Research in Personality, 40, 395-410.
Friedman, J. N. W., Oltmanns, T. F., & Turkheimer, E. (2007). Interpersonal perception and
personality disorders: Utilization of a thin slice approach. Journal of Research in
Personality, 41, 667-688.
Gray, J. A. (1987). The psychology of fear and stress. Cambridge, England: Cambridge
University Press.
Havlicek, J., Roberts, S. C., & Flegr, J. (2005). Women’s preference for dominant males’ odour:
Effects of menstrual cycle and relationship status. Biology Letters, 1, 256-259.
Heatherton, T. F., & Polivy, J. (1991). Development and validation of a scale for measuring state
self-esteem. Journal of Personality and Social Psychology, 60, 895-910.
Heatherton, T. F., & Vohs, K. D. (2000). Interpersonal evaluations following threats to self: Role
of self-esteem. Journal of Personality and Social Psychology, 78, 725-736.
Hepper, E. G., Gramzow, R. H., & Sedikides, C. (2010). Individual differences in selfenhancement and self-protection strategies: An integrative analysis. Journal of
Personality, 78, 781-814.
Hill, P. L., & Roberts, B. W. (in press). Narcissism, well-being, and observer-rated personality
across the lifespan. Social Psychological and Personality Science.
Jarry, J. L., & Kossert, A. L. (2007). Self-esteem threat combined with exposure to thin media
images leads to body image compensatory self-enhancement. Body Image, 4, 39-50.
Kenny, D. A. (1998). BLOCKO: Version 6 [Unpublished computer program]. University of
Connecticut, Storrs.

223

Kenny, D. A., Kashy, D. A., & Cook, W. L. (2006). Dyadic Data Analysis. New York: The
Guilford Press.
Kenny, D. A., Korchmaros, J. D., & Bolger, N. (2003). Lower level mediation in multilevel
models. Psychological Methods, 8, 115-128.
Kenrick, D. T., Trost, M. R., & Sheets, V. L. (1996). Power, harassment, and trophy mates: The
feminist advantage of an evolutionary perspective. In D. M. Buss & N. M. Malamuth,
Sex, power, conflict: Evolutionary and feminist perspectives (pp. 29-53). New York:
Oxford University Press.
Kernberg, O. F. (1975). Borderline Conditions and Pathological Narcissism. New York: Jason
Aronson.
Kernberg, P. F. (1998). Developmental Aspects of Normal and Pathological Narcissism. In E. F.
Ronningstam (Ed.), Disorders of Narcissism: Diagnostic, Clinical, and Empirical
Implications (1998). Northvale, NJ: Jason Aronson Inc.
Kernberg, O. F. (1998). Pathological Narcissism and Narcissistic Personality Disorder:
Theoretical Background and Diagnostic Classification. In E. F. Ronningstam (Ed.),
Disorders of Narcissism: Diagnostic, Clinical, and Empirical Implications (1998).
Northvale, NJ: Jason Aronson Inc.
Kohut, H. (1966). Forms and transformations of narcissism. In Self Psychology and the
Humanities: Reflections on a New Psychoanalytic Approach (1985). Edited by C. B.
Strozier. New York: W. W. Norton & Company, Inc.
Krull, J. L., & MacKinnon, D. P. (1999). Multilevel mediation modeling in group-based
intervention studies. Evaluation Review, 23, 418-444.
Krull, J. L., & MacKinnon, D. P. (2001). Multilevel modeling of individual and group level
mediated effects. Multivariate Behavioral Research, 36, 249-277.
Laurenceau, J. P., Barrett, L. F., & Pietromonaco, P. R. (1998). Intimacy as an interpersonal
process: The importance of self-disclosure, partner disclosure, and perceived partner
responsiveness in interpersonal exchanges. Journal of Personality and Social
Psychology, 74, 1238-1251.
Levy, K. N., Reynoso, J. S., Wasserman, R. H., & Clarkin, J. F. (2006). Narcissistic personality
disorder. In W. O’Donohue & S. O. Lilienfield (Eds.). Sage Handbook of Personality
Disorders (pp. 233-277). New York: Sage Press.
MacKinnon, D. P. (2008). Introduction to Statistical Mediation Analysis. New York: Lawrence
Erlbaum Associates.
Mednick, S. A. (1962). The associative basis of the creative process. Psychological Review, 69,
224

220-232.
Miller, J. D., & Campbell, W. K. (2008). Comparing clinical and social-personality
conceptualizations of narcissism. Journal of Personality, 76 (3), 449-476.
Millon, T. (1996). Disorders of Personality: DSM-IV and Beyond. New York: Wiley.
Mongeau, P. A., Serewicz, M. C., & Therrien, L. F. (2004). Goals for cross-sex first dates:
Identification, measurement, and the influence of contextual factors. Communication
Monographs, 71, 121-147.
Oltmanns, T. F., Friedman, J. N. W., Fiedler, E. R., & Turkheimer, E. (2004). Perceptions of
people with personality disorders based on thin slices of behavior. Journal of Research in
Personality, 38, 216-229.
Paulhus, D. L. (1998). Interpersonal and intrapsychic adaptiveness of trait self-enhancement: A
mixed blessing? Journal of Personality and Social Psychology, 74, 1197-1208.
Pincus, A. L., Ansell, E. B., Pimentel, C. A., Cain, N. M., Wright, A. G. C., & Levy, K. N.
(2009). Initial construction and validation of the pathological narcissism inventory.
Psychological Assessment, 21 (3), 365-379.
Pincus, A. L., & Lukowitsky, M. R. (2010). Pathological narcissism and narcissistic personality
disorder. Annual Review of Clinical Psychology, 6, 421-446.
Raskin, R. N., & Hall, C. S. (1979). A narcissistic personality inventory. Psychological Reports,
45 (2), 590.
Raskin, R. N., & Terry, H. (1988). A principal-components analysis of the Narcissistic
Personality Inventory and further evidence of its construct validity. Journal of
Personality and Social Psychology, 54 (5), 890-902.
Rhodewalt, F. & Morf, C. C. (1995). Self and interpersonal correlates of the Narcissistic
Personality Inventory. Journal of Research in Personality, 29, 1-23.
Sedikides, C., & Luke, M. (2008). On when self-enhancement and self-criticism function
adaptively and maladaptively. In E. C. Chang (Ed.), Self-criticism and self-enhancement:
Theory, research, and clinical implications. Washington, D.C.: American Psychological
Association.
Sobel, M. E. (1982). Asymptotic confidence intervals for indirect effects in structural equation
models. Sociological Methodology, 13, 290-312.
Stone, M. H. (1998). Normal Narcissism: An Etiological and Ethological Perspective. In E. F.
Ronningstam (Ed.), Disorders of Narcissism: Diagnostic, Clinical, and Empirical
Implications (1998). Northvale, NJ: Jason Aronson Inc.
225

Tenney, E. R., & Spellman, B. A. (2011). Complex social consequences of self-knowledge.
Social Psychological and Personality Science, 2, 343-350.
Tritt, S. M., Ryder, A. G., Ring, A. J., & Pincus, A. L. (2010). Pathological narcissism and the
depressive temperament. Journal of Affective Disorders, 122, 280-284.
Vazire, S., Naumann, L. P., Rentfrow, P. J., & Gosling, S. D. (2008). Portrait of a narcissist:
Manifestations of narcissism in physical appearance. Journal of Research in Personality,
42, 1439-1447.
Vonk, R. (2002). Self-serving interpretations of flattery: Why ingratiation works. Journal of
Personality and Social Psychology, 82, 515-526.
Watson, P. J., & Biderman, M. D. (1993). Narcissistic Personality Inventory factors, splitting and
self-consciousness. Journal of Personality Assessment, 61, 41-57.
Watson, P. J., Little, T., Sawrie, S. M., & Biderman, M. D. (1992). Measures of the narcissistic
personality: Complexity of relationships with self-esteem and empathy. Journal of
Personality Disorders, 6 (4), 434-449.
Watson, P. J., & Morris, R. J. (1991). Narcissism, empathy and social desirability. Personality
and Individual Differences, 12 (6), 575-579.
Wiggins, J. S., Trapnell, P., & Phillips, N. (1988). Psychometric and geometric characteristics of
the Revised Interpersonal Adjective Scales (IAS-R). Multivariate Behavioral Research,
23, 517-530.
Wright, A. G. C., Lukowitsky, M. R., Pincus, A. L., & Conroy, D. E. (2010). The higher order
factor structure and gender invariance of the Pathological Narcissism Inventory.
Assessment, 17, 467-483.
Zitek, E. M., Jordan, A. H., Monin, B., & Leach, F. R. (2010). Victim entitlement to behave
selfishly. Journal of Personality and Social Psychology, 98, 245-255.

226