An Examination of the Perceptions of Social Network ... - APA PsycNET

Personality Disorders: Theory, Research, and Treatment 2014, Vol. 5, No. 2, 137–145

© 2013 American Psychological Association 1949-2715/14/$12.00 DOI: 10.1037/per0000024

An Examination of the Perceptions of Social Network Characteristics Associated With Grandiose and Vulnerable Narcissism Joanna Lamkin

Allan Clifton

University of Georgia

Vassar College

W. Keith Campbell and Joshua D. Miller This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

University of Georgia Two dimensions of narcissism exist, grandiose and vulnerable, which are thought to be associated with distinctly different patterns of interpersonal behavior. Social network analysis is a way of quantifying and analyzing interpersonal interactions that may prove useful for characterizing the networks associated with these narcissism dimensions. In the current study, participants (N ⫽ 148) completed scales assessing both narcissism dimensions and a measure of the five-factor model of personality. Egocentric network information about participants’ 30 closest friends and family members (i.e., “alters”) was also obtained. Both narcissism dimensions were characterized by negative perceptions of the individuals who comprise one’s social networks, and many of these relations were mediated by individuals’ higher levels of antagonism. Grandiose narcissism also interacted with alter centrality (i.e., importance to the network) such that individuals low on grandiose narcissism were less likely to perceive central alters in a negative light and were more attuned to central alters than were individuals high on grandiose narcissism. Overall, both narcissism dimensions were associated with perceiving one’s overall social environment negatively because of the high levels of antagonism that characterize both narcissism dimensions. Individuals high on grandiose narcissism, however, appear to be more insensitive to the relative importance of individuals in their social networks. Keywords: narcissism, personality, social network analysis, interpersonal behavior

agency and low communion; whereas vulnerable narcissism is primarily related to low communion (e.g., Miller, Price, Gentile, Lynam, & Campbell, 2012; Pincus, Ansell, Pimentel, Cain, Wright, & Levy, 2009). These differences may have an important effect on the interpersonal behaviors and social networks associated with these two narcissism dimensions. For instance, grandiose narcissism is associated with being liked by others, at least initially and in certain circumstances (e.g., Back, Schmukle, & Egloff, 2010; Küfner, Nestler, & Back, in press; Paulhus, 1998). This is most likely because of a combination of personality traits (e.g., confidence, assertiveness, and lack of social anxiety; Miller et al., 2011), attractiveness (Holtzman & Strube, 2010), style of dress (e.g., well put-together; wearing of stylish clothing), self-presentation, and speaking style (Back et al., 2010) that are associated with grandiose narcissism. However, the initial positive impressions that others sometimes hold of grandiose individuals tend to disappear over time as they become better acquainted with these individuals and come to see narcissistic individuals as less agreeable, less emotionally stable, and more maladjusted (Paulhus, 1998). Similarly, Narcissistic Personality Disorder (NPD) is associated with causing substantial distress to significant others (Miller, Campbell, & Pilkonis, 2007), which is likely because of the use of “domineering, vindictive, and intrusive” interpersonal behaviors (Ogrodniczuk, Piper, Joyce, Steinberg, & Duggal, 2009, p. 837). At this time, the literature is less well-developed regarding the interpersonal manifestations of vulnerable narcissism, although individu-

It is now well-established that there are at least two relatively distinct dimensions of narcissism, which are most commonly referred to as grandiose and vulnerable narcissism (Miller, Hoffman, Gaughan, Gentile, Maples, & Campbell, 2011; Pincus & Lukowitsky, 2010; Wink, 1991). Both dimensions are associated with antagonistic interpersonal traits such as suspiciousness, manipulativeness, immodesty, noncompliance, and aggression, although to varying degrees (Miller, Dir, Gentile, Wilson, Pryor, & Campbell, 2010; Miller et al., 2011). However, grandiose and vulnerable narcissism do differ in certain aspects of interpersonal functioning. For example, the two differ significantly in their relations with the personality domain of extraversion/introversion (grandiose: high extraversion; vulnerable: high introversion) (Campbell & Miller, 2013) and attachment styles (i.e., grandiose: secure attachment style; vulnerable: fearful attachment styles) (Dickinson & Pincus, 2003; Miller et al., 2010). From the perspective of the interpersonal circumplex, grandiose narcissism is related to both high

This article was published Online First December 23, 2013. Joanna Lamkin, Department of Psychology, University of Georgia; Allan Clifton, Department of Psychology, Vassar College; W. Keith Campbell, and Joshua D. Miller, Department of Psychology, University of Georgia. Correspondence concerning this article should be addressed to Joshua D. Miller, PhD, Department of Psychology, University of Georgia, Athens, GA 30602-3013. E-mail: [email protected] 137

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

138

LAMKIN, CLIFTON, CAMPBELL, AND MILLER

als high on this dimension are not initially viewed as likable or attractive by strangers (Miller et al., 2011), informants see them as being both neurotic and introverted (Miller et al., 2010), and they report being hostile, distrustful, and noncompliant (Campbell & Miller, 2013). A method exists for quantifying the nature of interpersonal relationships that might prove useful for elucidating the interpersonal functioning associated with grandiose and vulnerable narcissism: Social Network Analysis (SNA). A social network consists of a group of individuals connected through some commonality. SNA calculates the pattern of these connections, revealing structures that are otherwise hidden (see Wasserman & Faust, 1994, for an in-depth overview). There are two major types of social networks: egocentric and complete (also known as sociocentric or bounded networks; van Duijn & Vermunt, 2006). In complete networks, individuals are connected by membership or identification (e.g., a business or sorority), and each member of the group reports on his or her ties with every other member of the group. In egocentric networks, a focal individual lists a specified number of individuals (referred to as “alters”) in his or her personal social world. The focal individual then provides information about the nature of all of the relationships within that network (e.g., Clifton, 2011). For both types of networks, the connection between pairs of individuals is quantified on some kind of relationship, such as friendship or level of acquaintanceship (see Koehly & Shivy, 1998). Although SNA has a rich interdisciplinary history (see Martino & Spoto, 2006) and has been used in several studies to investigate personality correlates of network structures (e.g., Burt, Jannotta, & Mahoney, 1998; Kalish & Robins, 2006), it has rarely been used in the study of personality pathology. In one such study, egocentric SNA was used to analyze the network tendencies of individuals with borderline personality disorder (Clifton, Pilkonis, & McCarty, 2007). These individuals were more likely to have former romantic partners and a smaller concentration of positive relationships in their social networks when compared to individuals without a personality disorder. A sociocentric study compared network position with presence of personality disorder in groups of military recruits (Clifton, Turkheimer, & Oltmanns, 2009). NPD was linked to greater out-degree (i.e., individuals reported having a greater number of closer relationships that were not verified by others) and an overestimation of one’s social connections. One of the most common ways of quantifying position within a network relies on the concept of centrality. Centrality in personal networks is a measure of importance within the focal participant’s network (e.g., McCarty & Wutich, 2005, Wasserman & Faust, 1994), indicating whether an alter is a central (rather than peripheral) member of the network. Several methods have been developed to compute centrality, the simplest of which is degree centrality (Martino & Spoto, 2006). Degree centrality calculates the number of direct connections between an alter and each other alter in the network, weighted by the strength of these connections. The greater the number of connections between an alter and the rest of the network, the more highly embedded, or central, the alter is to the focal participant’s social world. Because central alters are important to the overall network (Wasserman & Faust, 1994), centrality reflects prominence (Martino & Spoto, 2006). Given that narcissism is frequently characterized by interpersonal conflict (Miller, Campbell, & Pilkonis, 2007), alter centrality likely man-

ifests complex relationships with alter characteristics (i.e., “important” alters may be perceived in a certain way that differs from the way nonprominent alters are seen). For example, narcissistic individuals could see someone influential as a threat to the self or someone to be envied, or conversely as a high-status, prominent person who provides an opportunity for self-enhancement (see Campbell, 1999). In the current study, we used egocentric SNA as a way to examine and compare the social network perceptions related to grandiose and vulnerable narcissism. We first examined the correlations between grandiose and vulnerable narcissism and the social network ratings of alter characteristics and relationship dynamics (averaged across 30 alters). Prior research has shown that disagreeable individuals tend to perceive others as being disagreeable (Wood, Harms, & Vazire, 2010). First, we expected that both grandiose and vulnerable narcissism would be associated with negative perceptions of individuals within one’s social network. Second, given that both grandiose and vulnerable narcissism dimensions are multidimensional constructs (e.g., grandiose narcissism is composed primarily of interpersonal antagonism [or low communality] and high extraversion [or agency]) we next tested whether there are indirect effects between the narcissism dimensions and negative alter perceptions through the participants’ own antagonistic traits. It is important to note that these analyses are not based on the idea that narcissism leads to trait disagreeableness, which in turn leads to perceptions of alters. Instead, these analyses simply test the hypothesis that the relations between narcissism and negative alter perceptions are due, in part, to participants’ own disagreeableness. Finally, we examined participants’ perceptions of their alters as a function of their own narcissism and the position of each alter within the participant’s social network (i.e., centrality). We used multilevel modeling to predict participant ratings of each alter from participant narcissism, degree centrality of each alter within the network, and the interaction between the two. The focus of these final analyses is to test whether alter centrality moderates the relations between the narcissism dimensions and alter perceptions.

Method Participants Participants were 148 undergraduate (53% women; 86% White; Mean age ⫽ 19.2; SD ⫽ 1.5) students recruited from a research participant pool at a large Southeastern university. Participants gave written informed consent, completed questionnaires, and received research credit. IRB approval was obtained for all aspects of the study.

Materials Narcissistic Personality Inventory. The Narcissistic Personality Inventory (NPI; Raskin & Terry, 1988) is a frequently used, 40-item assessment of grandiose narcissism (␣ ⫽ .86; Mean ⫽ 16.62; SD ⫽ 7.02). Total NPI score was used as the primary measure of grandiose narcissism. Pathological Narcissism Inventory. The Pathological Narcissism Inventory (PNI; Pincus, Ansell, Pimentel, Cain, Wright, & Levy, 2009) is a 52-item self-report measure of narcissistic traits.


NARCISSISM AND SOCIAL NETWORKS

The four subscales that measure vulnerable narcissism were used here: Contingent Self-esteem (PNI CSE; Mean ⫽ 37.01; SD ⫽ 12.74), Hiding the self (PNI HS; Mean ⫽ 25.63; SD ⫽ 6.27), Devaluing (PNI Dev; Mean ⫽ 18.54; SD ⫽ 7.58), and Entitlement rage (PNI ER; Mean ⫽ 23.37; SD ⫽ 8.12). Alphas for the PNI-V subscales ranged from .71 to .91. Revised NEO Personality Inventory. The Revised NEO Personality Inventory (NEO PI-R; Costa & McCrae, 1992) is a 240item self-report measure of the FFM. The higher-order domains of the NEO PI-R are neuroticism, extraversion, openness, agreeableness, and conscientiousness. Each domain consists of six more specific facets. Alphas for the present study ranged from .89 to .92. Social network report. Participants were asked to list 30 individuals (alters) perceived as most important to them within the past year. Participants rated each alter’s gender, race, student status if applicable (e.g., freshman), origin of relationship (school, family, or other), and length of relationship. Participants also rated whether they had been roommates, or romantically or sexually involved with each alter. Participants rated the personality of each alter as well as relationship characteristics (from “not at all” to “very much”). Participants rated each alter on constructs reflecting different aspects of personality. The Ten Item Personality Inventory (TIPI; Gosling, Rentfrow, & Swann, 2003) was used to rate alters on the Big Five. Additional items were used to rate each alter on self-centeredness, attractiveness, social status, leadership, narcissism, intelligence, likability, and kindness. Perceptions of relationship characteristics were also rated for frequency of arguing with each alter, frequency of envying each alter, frequency of comparing self to each alter, perceived familiarity with each alter, and perceived closeness to each alter on the same. Finally, participants also rated how close each alter was to every other alter using a ⫺3 (actively dislike each other) to 3 (very close) scale.

Procedure Data were collected in groups no larger than four in rooms with adequate privacy (e.g., barriers separating each cubicle). EgoNet (McCarty, 2004), a computer software program, was used to obtain social network and alter information. All data collected in these settings were self-reported, including descriptions of the social networks. UCINet (Borgatti, Everett, & Freeman, 2002) computer software was used to extract the data.

Results Perceived Alter Characteristics and Relationship Dynamics A p value of ⱕ.01 was used in all analyses to control for Type I error. The two narcissism measures were not significantly related (r ⫽ .03, ns). Grandiose and vulnerable narcissism scores were correlated with the average ratings of the 30 alters in terms of individual characteristics (e.g., Big Five personality domains) and relationship dynamics (e.g., frequency with which participant argued with alters; Table 1). Both narcissism dimensions were significantly negatively correlated with perceiving alters as likable, kind, and agreeable. Both dimensions were also significantly related to a tendency to argue with one’s alters. Grandiose narcissism

139

Table 1 Correlations Between Narcissism and Mean Perceived Alter Characteristics

Mean alter characteristics Self-centered Attractive Social status Leadership Narcissism Intelligence Likeability Kindness Mean alter personality Neuroticism Extraversion Openness to experience Agreeableness Conscientiousness Mean relationship dynamics Frequency of arguing Envying the alter Comparing self to alter Closeness Familiarity

Grandiose

Vulnerable

.33ⴱ ⫺.05 ⫺.12 ⫺.04 .29ⴱ ⫺.17 ⫺.22ⴱ ⫺.27ⴱ

.19 ⫺.14 ⫺.05 ⫺.08 .19 ⫺.17 ⫺.24ⴱ ⫺.24ⴱ

.18 ⫺.09 ⫺.05 ⫺.24ⴱ ⫺.16

.32ⴱ ⫺.24ⴱ ⫺.14 ⫺.34ⴱ ⫺.12

.24ⴱ .05a .13 ⫺.01 ⫺.09

.28ⴱ .34ⴱb .24ⴱ ⫺.12 ⫺.06

Note. Correlations in the same row with different superscripts are statistically significantly different. Grandiose ⫽ NPI score; Vulnerable ⫽ PNI Vulnerable scores. ⴱ p ⬍ .01.

was also positively linked to perceptions of alters as self-centered and narcissistic, whereas vulnerable narcissism was related to perceiving alters as more neurotic and introverted and comparing oneself to and envying one’s alters. Only the correlations manifested by the two narcissism dimensions and envy were significantly different. A second-order correlation of these two sets of correlations (grandiose and vulnerable narcissism with the 18 alter descriptions) supports their general similarity (r ⫽ .86, p ⬍ .001).

Does Agreeableness Explain Why Grandiose and Vulnerable Narcissism Are Both Related to Negative Perceptions of the Individuals in One’s Social Network? We next tested whether there were significant indirect effects of self-report agreeableness on the relations between the narcissism dimensions and the alter ratings. To test this question, mediation analyses were conducted using the bootstrapping procedures described by Preacher and Hayes (2004) to provide the indirect effects (product of the A and B paths), which were estimated after 5,000 replications as suggested by Hayes (2009). Significance was determined using 99% confidence intervals (CIs). Diverging from the classic Baron and Kenny (1996) approach, we followed recent suggestions (e.g., Hayes, 2009; MacKinnon et al., 2000) that all relations (X ¡ Y) be tested for indirect effects, not just those that manifest significant total effects. The unstandardized regression coefficients for all paths are presented in Tables 2 (grandiose narcissism) and 3 (vulnerable narcissism). For grandiose narcissism, there were significant indirect effects of self-reported agreeableness for 14 of the 18 relations. For example, the relation between grandiose narcissism and perceptions of alters as self-


140

Table 2 Regression Coefficients From Mediation Analyses for Grandiose Narcissism


Indirect effect (AB)

Dependent variables Alter self-centeredness Alter Attractiveness Alter social status Alter leadership Alter narcissism Alter intelligence Alter likeability Alter kindness Alter neuroticism Alter extraversion Alter openness Alter agreeableness Alter conscientiousness Frequency of arguing Envying the alter Comparing self to alter Closeness Familiarity

Path A

Path B

Path C

Path C’

B

SE

B

SE

B

SE

B

SE

B

SE

.12ⴱ ⫺.12ⴱ ⫺.07ⴱ ⫺.10ⴱ .11ⴱ ⫺.08ⴱ ⫺.08ⴱ ⫺.09ⴱ .11ⴱ ⫺.11ⴱ ⫺.05ⴱ ⫺.20ⴱ ⫺.06ⴱ .07ⴱ .01 ⫺.02 ⫺.08 ⫺.03

.03 .04 .03 .03 .03 .02 .02 .02 .03 .04 .02 .04 .02 .02 .03 .04 .03 .03

⫺.24ⴱ ⫺.24ⴱ ⫺.24ⴱ ⫺.24ⴱ ⫺.24ⴱ ⫺.24ⴱ ⫺.24ⴱ ⫺.24ⴱ ⫺.24ⴱ ⫺.24ⴱ ⫺.24ⴱ ⫺.24ⴱ ⫺.24ⴱ ⫺.24ⴱ ⫺.24ⴱ ⫺.24ⴱ ⫺.24ⴱ ⫺.24ⴱ

.04 .04 .04 .04 .04 .04 .04 .04 .04 .04 .04 .04 .04 .04 .04 .04 .04 .04

⫺.51ⴱ .52ⴱ .31 .43ⴱ ⫺.44ⴱ .32ⴱ .32ⴱ .38ⴱ ⫺.46ⴱ .46ⴱ .34ⴱ .85ⴱ .26ⴱ ⫺.29ⴱ ⫺.02 .08 .33 .13

.13 .13 .12 .11 .14 .09 .09 .10 .11 .17 .11 .17 .09 .09 .13 .13 .13 .12

.23ⴱ ⫺.04 ⫺.08 ⫺.02 .22ⴱ ⫺.08 ⫺.10ⴱ ⫺.14ⴱ .10 ⫺.08 ⫺.03 ⫺.21ⴱ ⫺.08 .11ⴱ .03 .08 ⫺.01 ⫺.05

.05 .06 .05 .05 .06 .04 .04 .04 .05 .07 .04 .07 .04 .04 .05 .05 .05 .05

.11 .09 ⫺.00 .08 .11 ⫺.00 ⫺.03 ⫺.05 ⫺.01 .03 .05 ⫺.01 ⫺.01 .04 .02 .10 .07 ⫺.02

.06 .06 .06 .05 .06 .04 .04 .05 .05 .08 .05 .08 .04 .04 .06 .06 .06 .06

Note. Path A ⫽ Effect of grandiose narcissism on agreeableness; Path B ⫽ Effect of agreeableness on DV (e.g., alter self-centeredness); Path C ⫽ Effect of grandiose narcissism on DV (e.g., alter self-centeredness); Path C’ ⫽ Effect of grandiose narcissism on the DV (e.g., alter self-centeredness) controlling for the indirect path. ⴱ p ⬍ .01.

centered was accounted for by participants’ self-report levels of agreeableness. For vulnerable narcissism, there were significant indirect effect of self-reported agreeableness for 13 of 18 alter ratings. For example, the relation between vulnerable narcissism and perceptions of alters as disagreeable was mediated by selfreported level of agreeableness.

Social Network Analyses Analyses were conducted to examine how participants’ perceptions of alters were statistically predicted by target narcissism, centrality (i.e., the embeddedness of the alter within the network), and the interaction between narcissism scores and centrality. The primary findings of interest here were the interactions. That is, we wanted to test whether narcissistic individuals judged their alters or relationships with alters differently depending on how central the alter was to the participants’ social networks. To account for the nonindependence inherent in multiple ratings of alters by each participant, we used mixed linear models with a “one-with-many” design (Kenny, Kashy, & Cook, 2006). All predictor variables were centered for the analyses, and separate analyses were conducted for vulnerable and grandiose narcissism. The results of these analyses are shown in Tables 4 (grandiose narcissism) and 5 (vulnerable narcissism). Replicating and extending the results reported in Table 1, controlling for centrality and the interaction term between narcissism and centrality, both variants of narcissism were significantly related to perceiving others as self-centered, narcissistic, less intelligent, less likable, less kind, less open to experience, and less conscientious.1 Both variants of narcissism were also related to arguing with alters and comparing oneself to one’s alters. Gran-

diose narcissism was uniquely related to perceiving alters as possessing less social status and being less familiar with alters. Vulnerable narcissism was uniquely positively related to envying alters, feeling less close to alters, and perceiving alters as less attractive.2 Across both narcissism dimensions (and controlling for narcissism and the relevant interaction terms), centrality was positively associated with perceptions of attractiveness, social status, leadership, intelligence, likability, kindness, extraversion, agreeableness, and conscientiousness. Again, across both narcissism dimensions, targets reported arguing with, envying, comparing themselves to, feeling close to, and familiar with central members of their networks. Controlling for grandiose narcissism and its interaction term, centrality was also negatively related to perceived alter self-centeredness and narcissism. In cases in which a significant interaction was found, simple slope analyses were conducted for centrality at varying levels of narcissism (⫾1 SD; see Aiken & West, 1991). The levels of narcissism were defined as low (⫺1 SD) and high (1 SD). Of 18 interactions between grandiose narcissism and centrality, 5 were 1 Both grandiose and vulnerable narcissism manifested more significant relations with the social network characteristics when examined in the multilevel modeling (i.e., Tables 4 and 5) than in the bivariate correlations (i.e., Table 1) because of differences in power. In the first set of analyses (i.e., Table 1), grandiose and vulnerable narcissism were correlated with the average alter ratings and thus had many fewer degrees of freedom (df ⫽ 146) than those in the multilevel modeling (df 1, 4434). 2 The inclusion of self-esteem as a covariate did not alter the pattern of results for grandiose and vulnerable narcissism when self-esteem was added as a main effect.

NARCISSISM AND SOCIAL NETWORKS

141

Table 3 Regression Coefficients From Mediation Analyses for Vulnerable Narcissism


Indirect effect (AB)

Dependent variables Alter self-centeredness Alter Attractiveness Alter social status Alter leadership Alter narcissism Alter intelligence Alter likeability Alter kindness Alter neuroticism Alter extraversion Alter openness Alter agreeableness Alter conscientiousness Frequency of arguing Envying the alter Comparing self to alter Closeness Familiarity

Path A

Path B

Path C

Path C’

B

SE

B

SE

B

SE

B

SE

B

SE

.08ⴱ ⫺.05ⴱ ⫺.04ⴱ ⫺.05ⴱ .07ⴱ ⫺.04ⴱ ⫺.04ⴱ ⫺.05ⴱ .05ⴱ ⫺.05 ⫺.04ⴱ ⫺.10ⴱ ⫺.04ⴱ .04ⴱ ⫺.01 ⫺.01 ⫺.03 ⫺.02

.03 .02 .02 .02 .03 .02 .02 .02 .02 .03 .02 .03 .01 .02 .02 .02 .02 .02

⫺.14ⴱ ⫺.14ⴱ ⫺.14ⴱ ⫺.14ⴱ ⫺.14ⴱ ⫺.14ⴱ ⫺.14ⴱ ⫺.14ⴱ ⫺.14ⴱ ⫺.14ⴱ ⫺.14ⴱ ⫺.14ⴱ ⫺.14ⴱ ⫺.14ⴱ ⫺.14ⴱ ⫺.14ⴱ ⫺.14ⴱ ⫺.14ⴱ

.04 .04 .04 .04 .04 .04 .04 .04 .04 .04 .04 .04 .04 .04 .04 .04 .04 .04

⫺.59ⴱ .40ⴱ .32ⴱ .34ⴱ ⫺.52ⴱ .30ⴱ .31ⴱ .39ⴱ ⫺.37ⴱ .33 .25ⴱ .74ⴱ .26ⴱ ⫺.28ⴱ .09 .07 .23 .15

.11 .12 .11 .10 .13 .08 .08 .09 .09 .15 .10 .15 .08 .08 .11 .11 .12 .11

.13 ⫺.10 ⫺.03 ⫺.05 .14 ⫺.08 ⫺.11ⴱ ⫺.12ⴱ .18ⴱ ⫺.20ⴱ ⫺.08 ⫺.30ⴱ ⫺.06 .13ⴱ .22ⴱ .15ⴱ ⫺.07 ⫺.04

.06 .06 .05 .05 .06 .04 .04 .04 .04 .07 .04 .07 .04 .04 .05 .05 .05 .05

.05 ⫺.04 .02 .00 .07 ⫺.04 ⫺.07 ⫺.07 .13ⴱ ⫺.15 ⫺.04 ⫺.20ⴱ ⫺.02 .09 .23ⴱ .16ⴱ ⫺.04 ⫺.02

.05 .06 .05 .05 .06 .04 .04 .04 .04 .07 .05 .07 .04 .04 .05 .05 .05 .05

Note. Path A ⫽ Effect of vulnerable narcissism on agreeableness; Path B ⫽ Effect of agreeableness on DV (e.g., alter self-centeredness); Path C ⫽ Effect of vulnerable narcissism on DV (e.g., alter self-centeredness); Path C’ ⫽ Effect of vulnerable narcissism on the DV (e.g., alter self-centeredness) controlling for the indirect path. ⴱ p ⬍ .01.

significant (see Figure 1); no significant interactions were found between vulnerable narcissism and centrality. Alter narcissism. Participants low on grandiose narcissism rated more central alters as less narcissistic; the slope of this line was significantly different from zero. For participants high on grandiose narcissism, there was no significant relation between centrality and ratings of alter narcissism. Envying the alter. Participants low on grandiose narcissism were more likely to endorse envying alters who were more central to the network; the slope of this line was significantly different from zero. For participants high on grandiose narcissism, there was no significant relation between centrality and participants’ envy of alters. Comparing self to the alter. Participants who were either low or high on grandiose narcissism were more likely to compare themselves to central alters. The slopes of these lines were significantly different from one another and from zero. Familiarity of alter. Participants who were either low or high on grandiose narcissism were more likely to rate central alters as being familiar. The two slopes differed from one another and from zero. Alter extraversion. Participants low on grandiose narcissism rated more central alters as being more extraverted; the slope of this line was significantly different from zero. For participants high on grandiose narcissism, there was no significant relation between centrality and ratings of alter extraversion.

Discussion Narcissism and NPD are associated with substantial interpersonal impairment, particularly in the domain of causing distress to significant others (e.g., romantic partners, family members,

friends; Miller et al., 2007), due in part to the antagonistic and noncommunal interpersonal approach associated with both constructs (e.g., Miller et al., 2011). Much less is known, unfortunately, about how vulnerable narcissism is related to interpersonal impairment as researchers have only recently begun to examine systematically the role of both narcissism dimensions in interpersonal behavior. One would expect, however, that vulnerable narcissism would also be related to interpersonal dysfunction as it is associated with traits related to interpersonal impairment such as antagonism, neuroticism, and introversion (e.g., Campbell & Miller, 2013), a cold and noncommunal interpersonal style (Miller et al., 2012; Pincus et al., 2009), and anxious and avoidant attachment styles (e.g., Miller et al., 2010). In the current study, we used social network analyses in an attempt to better understand the social worlds in which individuals with elevated levels of grandiose and vulnerable narcissism are embedded. As expected (e.g., Srivastava, Guglielmo, & Beer, 2010; Wood et al., 2010), individuals with high scores on either vulnerable or grandiose narcissism held a number of negative perceptions of the individuals who comprise their social environment and the participants’ relationship dynamics with these alters. For instance, both grandiose and vulnerable narcissism were correlated with perceptions of their alters as being less kind, likable, and agreeable. Overall, the perceptions of one’s social network were quite similar across the two narcissism dimensions, as evidenced by the very strong correlation between the two sets of correlations (r ⫽ .86). These current findings are consistent with recent research that has found that one’s own traits are related to one’s perceptions of others. For instance, Srivastava et al. (2010) demonstrated that


142

Table 4 Mixed Model Analyses of Grandiose Narcissism, Degree Centrality, and Alter Characteristics


Grandiose narcissism

Grand-narⴱcentrality

Centrality

Alter characteristic

B

SE

t(1,4434)

B

SE

t(1,4434)

B

SE

t(1,4434)

Self-centeredness Attractiveness Social status Leadership Narcissism Intelligence Likeability Kindness Argue Envy Comparison Closeness Familiarity Neuroticism Extraversion Openness Agreeableness Conscientiousness

.229 ⫺.037 ⫺.079 ⫺.025 .223 ⫺.076 ⫺.103 ⫺.134 .104 .025 .076 ⫺.016 ⫺.069 .092 ⫺.042 ⫺.048 ⫺.106 ⫺.049

.02 .02 .02 .02 .02 .01 .01 .01 .01 .02 .02 .02 .02 .01 .01 .01 .01 .01

12.18ⴱ ⫺2.13 ⫺5.22ⴱ ⫺1.47 12.57ⴱ ⫺5.69ⴱ ⫺8.09ⴱ ⫺10.08ⴱ 6.99ⴱ 1.63 4.77ⴱ ⫺1.00 ⫺3.36ⴱ 6.44ⴱ ⫺2.86ⴱ ⫺3.56ⴱ ⫺7.51ⴱ ⫺3.38ⴱ

⫺.004 .007 .012 .009 ⫺.003 .005 .009 .009 .009 .006 .008 .232 .024 ⫺.001 .005 .001 .004 .004

.00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00

⫺3.10ⴱ 6.47ⴱ 12.32ⴱ 8.13ⴱ ⫺2.89ⴱ 5.70ⴱ 10.37ⴱ 10.82ⴱ 8.74ⴱ 6.15ⴱ 7.38ⴱ 21.09ⴱ 17.05ⴱ ⫺0.90 5.17ⴱ 0.83 4.60ⴱ 3.99ⴱ

.001 .000 ⫺.000 ⫺.000 .004 ⫺.001 ⫺.000 ⫺.000 ⫺.000 ⫺.003 ⫺.004 ⫺.001 ⫺.009 .000 ⫺.003 .002 .001 ⫺.002

.00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00

1.18 0.27 ⫺0.52 ⫺0.84 4.02ⴱ ⫺0.76 ⫺0.49 ⫺0.57 ⫺1.09 ⫺3.75ⴱ ⫺4.74ⴱ ⫺1.32 ⫺7.08ⴱ 0.04 ⫺3.06ⴱ 2.16 1.29 ⫺1.76

Note. Grand-narⴱcentrality ⫽ interaction term for grandiose narcissism and centrality. p ⬍ .01.

ⴱ

and vulnerable narcissism were due, in part, to participants’ own disagreeableness. Despite the narcissism dimensions being unrelated in the current study, they manifested similar (and mostly negatively valenced) relations with alter perceptions because of the role of disagreeableness that is central to both grandiose and vulnerable narcissism. The current study also allowed us to examine how social network characteristics (i.e., centrality) affected perceptions of the alters and the participants’ relations with alters. As expected, individuals who were perceived as being central to their social

self-reported disagreeableness was significantly correlated with perceptions of others as being disagreeable. Similarly, Wood et al. (2010) found that self-reported disagreeableness was correlated with perceiving others (e.g., dormitory floor mates) as less agreeable, less conscientious, and more conventional, neurotic, and introverted. These researchers also found that self-reported personality disorder scores, including NPD, were negatively related to rating others in a generally positive light. Evidence of these negative perceiver effects was found in the current study. Generally negative perceptions of alters that were associated with grandiose

Table 5 Mixed Model Analyses of Vulnerable Narcissism, Degree Centrality, and Alter Characteristics Vulnerable narcissism

Vul-narⴱcentrality

Centrality

Alter characteristic

B

SE

t(1,4434)

B

SE

t(1,4434)

B

SE

t(1,4434)

Self-centeredness Attractiveness Social status Leadership Narcissism Intelligence Likeability Kindness Argue Envy Comparison Closeness Familiarity Neuroticism Extraversion Openness Agreeableness Conscientiousness

.125 ⫺.089 ⫺.013 ⫺.035 .142 ⫺.066 ⫺.099 ⫺.102 .140 .226 .159 ⫺.044 ⫺.011 .143 ⫺.092 ⫺.104 ⫺.147 ⫺.099

.02 .02 .02 .02 .02 .01 .01 .01 .01 .02 .02 .02 .02 .01 .01 .01 .01 .01

6.55ⴱⴱ ⫺5.11ⴱⴱ ⫺0.83 ⫺2.06 7.92ⴱⴱ ⫺4.92ⴱⴱ ⫺7.75ⴱⴱ ⫺7.65ⴱⴱ 9.43ⴱⴱ 14.87ⴱⴱ 10.06ⴱⴱ ⫺2.74ⴱ ⫺0.52 9.99ⴱⴱ ⫺6.29ⴱⴱ ⫺7.80ⴱⴱ ⫺10.43ⴱⴱ ⫺6.85ⴱⴱ

⫺.003 .007 .012 .009 ⫺.002 .004 .008 .008 .009 .007 .008 .022 .023 .000 .004 .000 .003 .003

.00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00

⫺2.25 5.81ⴱⴱ 12.04ⴱⴱ 7.85ⴱⴱ ⫺1.55 4.57ⴱⴱ 9.16ⴱⴱ 9.45ⴱⴱ 9.13ⴱⴱ 6.84ⴱⴱ 7.71ⴱⴱ 19.98ⴱⴱ 15.99ⴱⴱ 0.22 3.92ⴱⴱ 0.24 3.63ⴱⴱ 3.15ⴱ

.001 .001 ⫺.001 ⫺.001 .002 .002 .001 .001 .001 ⫺.000 ⫺.001 .002 ⫺.004 ⫺.001 .001 .001 .000 ⫺.001

.00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00

0.77 1.04 ⫺0.54 ⫺0.63 1.24 1.88 1.28 1.57 1.26 ⫺0.15 ⫺1.25 1.95 ⫺2.55 ⫺0.90 0.67 1.29 0.91 ⫺0.64

Note. Vul-narⴱcentrality ⫽ interaction term for vulnerable narcissism and centrality. For t tests, p-values are: ⴱⴱ ⫽ p ⬍ .001. ⴱ ⫽ p ⬍ .01.

7

6

6

5 4 3 2 1 0 -55

-35

-15

5

25

45

5 4 3 2 1 0

65

-55

-35

-15

5

25

45

65

Degree Centrality 7

6

6

Alter Extraversion

Familiarity of Alter


Degree Centrality 7

5 4 3 2 1 0

143

Comparing Self to Alter

7

Envying the Alter

Alter Narcissism

NARCISSISM AND SOCIAL NETWORKS 7 6 5 4 3 2 1 0 -55

-35

-15

5

25

45

65

Degree Centrality

5 4 3 2 1 0

-55

-35

-15

5

25

45

Degree Centrality

65

-55

-35

-15

5

25

45

65

Degree Centrality

Figure 1. Using degree centrality, participant grandiose narcissism, and their interaction to predict perceived alter characteristics. Note: A dotted line indicates low grandiose narcissism (⬍1 SD); a continuous line indicates high grandiose narcissism (⬎1 SD).

networks were generally viewed in a favorable light. Across two sets of analyses (controlling for the narcissism dimensions), central individuals were seen as more attractive, intelligent, likable, kind, extraverted, agreeable, and conscientious, better leaders, and as having greater social status. We were surprised to learn that centrality was associated with both positive (i.e., closeness; familiarity) and negative (i.e., arguing, envying, comparing oneself to) relationship dynamics. It is possible that individuals tend to admire and respect those who are most central to their social environment and that these central alters serve as standard-bearers to whom individuals can compare themselves; such comparisons may come with some cost, however, in the way of an increased likelihood of experiencing envy and contentiousness in relationships with these central alters. Of more direct relevance to the study of narcissism, centrality interacted with ratings of grandiose narcissism to statistically predict various alter characteristics. Essentially, the relationships between centrality and low grandiose narcissism closely reflected the general pattern that was seen for centrality when narcissism was controlled (i.e., seeing alters positively but also experiencing envy and comparisons). Uniquely, individuals who were low on grandiose narcissism rated more central alters as less narcissistic. These results suggest that less grandiose individuals may be more attuned to the characteristics of the individuals who comprise their social network such that central individuals—those who matter most to the network—are perceived as having more positive attributes (e.g., more likable; less narcissistic) and are sources of

comparison and envy. Conversely, centrality was either unrelated or related more weakly to these characteristics for grandiose individuals. As a whole, these findings suggest that grandiose individuals may be less discriminating in how they relate to those in their networks. Whereas individuals low on grandiosity view more central members of their networks more positively, grandiose individuals are equally critical of both central and noncentral members. This may be reflective of the arrogance and lack of empathy associated with grandiose narcissism (American Psychological Association, 2000). It is also consistent with previous social network analyses of personality pathology, which found that patients with borderline personality disorder were less discriminating in their positive relationships than were nonpersonality disordered patients (Clifton, Pilkonis, & McCarty, 2007). These analyses are exploratory and further research is needed to more fully understand these issues.

Limitations and Future Directions The current study is the first to use egocentric social network analysis to examine the nature of interpersonal functioning associated with both grandiose and vulnerable narcissism. A significant limitation of such an approach is that, without data from informants, it is impossible to separate idiosyncratic perceiver effects from consensual perceiver effects (e.g., Kenny, 2004; Kenny, Albright, Malloy, & Kashy, 1994). That is, we cannot discern



144

whether narcissistic individuals simply perceive the members of their networks as more dislikeable, unkind, and disagreeable, or whether these perceptions are, in fact, accurate. Therefore, it will be important to extend this line of research by using sociocentric SNA in which closed or bounded networks are used (e.g., dorms, schools) such that every member of the network rates themselves and their closeness to every other member of the network. A sociocentric approach would also allow for an examination of how grandiose and vulnerable narcissism dimensions affect position in the overall network. In the current egocentric design, the respondent is considered separately from the network because he or she is by definition connected to all members of the network (cf. McCarty & Wutich, 2005). Therefore, although egocentric methodology can tell us about the network position of alters, it cannot provide a measure of the respondent’s position within his or her larger social world. Additional research using a sociocentric approach may help explicate the interpersonal differences between grandiose and vulnerable narcissism. It is also important to note that the general similarities in perceptions of the social networks found for grandiose and vulnerable narcissism does not necessarily mean their networks are alike in all ways. It is possible that the networks may differ in ways not examined here such as size, adaptivity, or presence or absence of discernible cliques. Future studies should address the ways in the social networks of individuals high in grandiose or vulnerable narcissism converge and diverge. In addition, it will be important to test these findings using a more diverse sample with regard to relevant demographic variables and psychopathology. In terms of clinical implications, the Diagnostic and Statistical Manual for Mental Disorders-Fifth Edition (DSM-5) will include two strategies for assessing personality disorders. The primary strategy will simply be a reiteration of the DSM–IV system. A new system, however, will also be introduced that uses both basic personality pathology traits and functional impairment to diagnose PDs. For this latter approach, interpersonal impairment is one of the two domains of impairment that will be central to the diagnosis of PD. Given the growing importance placed on interpersonal impairment to the diagnosis of PD, we believe SNA-based techniques may prove fruitful in elucidating the nature of interpersonal functioning associated with these narcissism dimensions.

References Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Thousand Oaks, CA: Sage. American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: Author. 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. doi:10.1037/a0016338 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, 5, 1173–1182. Borgatti, S. P., Everett, M. G., & Freeman, L. C. (2002). Ucinet for Windows. [Computer software and manual]. Harvard, MA: Analytic Technologies. Burt, R. S., Jannotta, J. E., & Mahoney, J. T. (1998). Personality correlates of structural holes. Social Networks, 20, 63– 87. doi:10.1016/S03788733(97)00005-1

Campbell, W. (1999). Narcissism and romantic attraction. Journal of Personality and Social Psychology, 77, 1254 –1270. doi:10.1037/00223514.77.6.1254 Campbell, W. K., & Miller, J. D. (2013). Narcissistic Personality Disorder (NPD) and the Five-Factor Model: Delineating NPD, grandiose narcissism, and vulnerable narcissism. In T. A. Widiger & P. T. Costa (Eds.), Personality Disorders and the Five-Factor Model of Personality (3rd ed., pp. 133–145). Washington, DC: APA. doi:10.1037/13939-009 Clifton, A. (2011). Narcissism and social networks. In W. K. Campbell & J. D. Miller (Eds.), Handbook of narcissism and narcissistic personality disorder: Theoretical approaches, empirical findings, and treatments (pp. 360 –370). Hoboken, NJ: Wiley. Clifton, A., Pilkonis, P. A., & McCarty, C. (2007). Social network in borderline personality disorder. Journal of Personality Disorders, 21, 434 – 441. doi:10.1521/pedi.2007.21.4.434 Clifton, A., Turkheimer, E., & Oltmanns, T. F. (2009). Personality disorder in social networks: Network position as a marker of interpersonal dysfunction. Social Networks, 31, 26 –32. doi:10.1016/j.socnet.2008.08 .003 Costa, P. T., & McCrae, R. R. (1992). Revised NEO Personality Inventory (NEO-PIR) and NEO Five-Factor Inventory (NEO-FFI) Professional Manual. Lutz, FL: PAR, Inc. Dickinson, K. A., & Pincus, A. (2003). Interpersonal analysis of grandiose and vulnerable narcissism. Journal of Personality Disorders, 17, 188 – 207. doi:10.1521/pedi.17.3.188.22146 Gosling, S. D., Rentfrow, P. J., & Swann, W. B., Jr. (2003). A very brief measure of the big five personality domains. Journal of Research in Personality, 37, 504 –528. doi:10.1016/S0092-6566(03)00046-1 Hayes, A. F. (2009). Beyond Baron and Kenny: Statistical mediation analysis in the new millennium. Communication Monographs, 76, 408 – 420. doi:10.1080/03637750903310360 Holtzman, N. S., & Strube, M. J. (2010). Narcissism and attractiveness. Journal of Research in Personality, 44, 133–136. doi:10.1016/j.jrp.2009 .10.004 Kalish, Y., & Robins, G. (2006). Psychological predispositions and network structure: The relationship between individual predispositions, structural holes and network closure. Social Networks, 28, 56 – 84. doi:10.1016/j.socnet.2005.04.004 Kenny, D. A. (2004). PERSON: A general model of interpersonal perception. Personality and Social Psychology Review, 8, 265–280. doi: 10.1207/s15327957pspr0803_3 Kenny, D. A., Albright, L., Malloy, T. E., & Kashy, D. A. (1994). Consensus in interpersonal perception: Acquaintance and the big five. Psychological Bulletin, 116, 245–258. doi:10.1037/0033-2909.116.2 .245 Kenny, D. A., Kashy, D. A., & Cook, W. L. (2006). Dyadic data analysis. New York, NY: Guilford Press. Koehly, L. M., & Shivy, V. A. (1998). Social network analysis: A new methodology for counseling research. Journal of Counseling Psychology, 45, 3–17. doi:10.1037/0022-0167.45.1.3 Küfner, A. C. P., Nestler, S., & Back, M. D. (in press). The two pathways to being an (un-) popular narcissist. Journal of Personality. MacKinnon, D. P., Krull, J. L., & Lockwood, C. M. (2000). Equivalence of the mediation, confounding, and suppression effect. Prevention Science, 1, 173–181. doi:10.1023/A:1026595011371 Martino, F., & Spoto, A. (2006). Social Network Analysis: A brief theoretical review and further perspectives in the study of Information Technology. PsychNology Journal, 4, 53– 86. McCarty, C., & Wutich, A. (2005). Conceptual and empirical arguments for including or excluding ego from structural analyses of personal networks. Connections: Journal of the International Network for Social Network Analysis, 26, 80 – 86. Miller, J. D., Campbell, W. K., & Pilkonis, P. A. (2007). Narcissistic Personality Disorder: Relations with distress and functional impairment.


NARCISSISM AND SOCIAL NETWORKS Comprehensive Psychiatry, 48, 170 –177. doi:10.1016/j.comppsych .2006.10.003 Miller, J. D., Dir, A., Gentile, B., Wilson, L., Pryor, L. R., & Campbell, W. K. (2010). Searching for a vulnerable dark triad: Comparing factor 2 psychopathy, vulnerable narcissism, and borderline personality disorder. Journal of Personality, 78, 1529 –1564. doi:10.1111/j.1467-6494.2010 .00660.x Miller, J. D., Hoffman, B. J., Gaughan, E. T., Gentile, B., Maples, J., & Campbell, W. K. (2011). Grandiose and vulnerable narcissism: A nomological network analysis. Journal of Personality, 79, 1013–1042. doi:10.1111/j.1467-6494.2010.00711.x Miller, J. D., Price, J., Gentile, B., Lynam, D. R., & Campbell, W. K. (2012). Grandiose and vulnerable narcissism from the perspective of the interpersonal circumplex. Personality and Individual Differences, 53, 507–512. doi:10.1016/j.paid.2012.04.026 Ogrodniczuk, J. S., Piper, W. E., Joyce, A. S., Steinberg, P. I., & Duggal, S. (2009). Interpersonal problems associated with narcissism among psychiatric outpatients. Journal of Psychiatric Research, 43, 837– 842. doi:10.1016/j.jpsychires.2008.12.005 Paulhus, D. L. (1998). Interpersonal and intrapsychic adaptiveness of trait self-enhancement: A mixed blessing? Journal of Personality and Social Psychology, 74, 1197–1208. doi:10.1037/0022-3514.74.5.1197 Pincus, A. L., Ansell, E. B., Pimentel, C. A., Cain, N. M., Wright, A., & Levy, K. N. (2009). Initial construction and validation of the Pathological Narcissism Inventory. Psychological Assessment, 21, 365–379. doi: 10.1037/a0016530

145

Pincus, A. L., & Lukowitsky, M. R. (2010). Pathological narcissism and narcissistic personalit disorder. Annual Review of Clinical Psychology, 6, 421– 446. doi:10.1146/annurev.clinpsy.121208.131215 Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behavior Research Methods, Instruments, & Computers, 36, 717–731. Raskin, R., & Terry, H. (1988). A principle-components analysis of the Narcissistic Personality Inventory and further evidence of its construct validity. Journal of Personality and Social Psychology, 54, 890 –902. doi:10.1037/0022-3514.54.5.890 Srivastava, S., Guglielmo, S., & Beer, J. S. (2010). Perceiving others’ personalities: Examining the dimensionality, assumed similarity to the self, and stability of perceiver effects. Journal of Personality and Social Psychology, 98, 520 –534. doi:10.1037/a0017057 van Duijn, M. J., & Vermunt, J. K. (2006). What is special about social network analysis? Methodology: European Journal of Research Methods for the Behavioral and Social Sciences, 2, 2– 6. doi:10.1027/16142241.2.1.2 Wasserman, S., & Faust, K. (1994). Social network analysis: Methods and applications. New York, NY: Cambridge University Press. Wink, P. (1991). Two faces of narcissism. Journal of Personality and Social Psychology, 61, 590 –597. doi:10.1037/0022-3514.61.4.590 Wood, D., Harms, P., & Vazire, S. (2010). Perceiver effects as projective tests: What your perceptions of others say about you. Journal of Personality and Social Psychology, 99, 174 –190. doi:10.1037/a0019390