Are Psychopathic Traits Associated with Core Social Networks? An Exploratory Study in University Students

Abstract

In a sample of 480 university students, we examined associations between self-ratings of psychopathic traits, made using the Comprehensive Assessment of Psychopathic Personality (CAPP), the Psychopathic Personality Inventory: Short Form (PPI: SF), and self-ratings of the structure of their core social networks (i.e., best friends, intimates). Results indicated that higher self-ratings of domains (CAPP) and subscales (PPI: SF) related to interpersonal dominance, manipulation, poor attachment, and emotional regulation were associated with less connected core networks. We interpret the dominance and manipulation domain and subscale findings as preliminary evidence of a deliberate strategy to provide a more influential position within one’s social network. As for the associations with the attachment and emotional regulation domain and subscale findings, we suggest this could be reflective of deficits or a lack of desire both in establishing and maintaining long-term relationships.

Keywords

CAPP interpersonal relationships PPI SNA social networks

Psychopathy (also referred to as antisocial, dissocial, and psychopathic personality disorder) has been the focus of considerable research over the past 50 years. Consequently, there is an impressive body of research supporting its validity in comparison to other mental disorders in general and more specifically in comparison to other personality disorders (for reviews, see Hare and Neumann 2008; Hart and Storey 2013). Although there is still active debate concerning how best to conceptualize and assess the disorder, psychopathy has long been characterized by interpersonal dysfunctions. In fact, interpersonal dysfunctions are considered the hallmark of psychopathy (Blackburn 2006; McCord 1982). Moreover, the various symptoms used to describe psychopathic individuals, such as affectively cold, interpersonally deceptive, and antisocial (Ali and Chamorro-Premuzic 2010; Hare and Neumann 2010), emphasize the lack of social relatedness and emotional distance typical of this disorder (Conradi et al. 2016).

Although psychopathy has been studied primarily in correctional and forensic mental health settings, researchers increasingly have studied settings where subclinical forms of the disorder are more typical—that is, people with some psychopathic symptoms—such as in community settings. Community-based samples provide an ideal setting to study interpersonal relationships because individuals with nonclinical psychopathy typically have fewer problems with overt antisocial and criminal behaviors but tend to demonstrate similar core affective and interpersonal traits (e.g., callousness, guiltlessness, fearlessness, superficial charm, and grandiosity; Kastner, Sellbom, and Lilienfeld 2012). Research with nonclinical samples has found evidence for diverse expressions of psychopathic traits across the population (Skeem et al. 2003), and clinical experts have argued that research on community samples is necessary for findings to generalize to more individuals (e.g., Lilienfeld 1998). Thus, the current study utilizes a community sample of university undergraduate students to explore nonclinical psychopathy through the lens of interpersonal relations.

Despite the longstanding relationship between psychopathy and interpersonal dysfunctions, there has been very little research that has examined self-reported psychopathic traits and how they are associated with differences in the individual-level network structure of interpersonal relations. Rather, most studies have assessed interpersonal relations through measures of relationship quality (e.g., intimate relationship satisfaction, Ali and Chamorro-Premuzic 2009; friendship satisfaction, Uziebolo et al. 2010) or adult attachment styles in partner relationships (e.g., Conradi et al. 2016). Although some studies have used peer ratings (e.g., Muñoz, Kerr, and Besić 2008; Rauthmann 2011), these have been used to assess the accuracy of the perceptions of social interactions among individuals with psychopathic traits rather than examine the structure of these interactions. Lacking from the extant literature are studies jointly measuring the individual perspectives of those with psychopathic traits as well as how they perceive the position of others in their networks. This is an important step forward because at the most basic level, personality has been found to play a substantial role in the development and maintenance of social ties (Zhu, Woo, Porter, and Brzezinski 2013). Considering the importance of interpersonal dysfunction to the construct of psychopathy, it seems likely that psychopathic traits will impact individual perceptions of how one structures and maintains their social networks. Accordingly, we approach our understanding of psychopathy using the social network perspective, which sees interpersonal functions as a product of a complex pattern of relationships that constitute a person’s social network. Additionally, this perspective allows for people’s relationships within an entire network to be interpreted rather than just their relationship with a single person.

In this article, we explore psychopathy in the context of the general population, focusing on whether social networks differ among individuals who demonstrate psychopathic tendencies within a university setting.

Network Structure and Personality Traits

Social network analysis (SNA) is the most suitable approach to assess the role of social relations on behavioral or psychological outcomes like psychopathy. The fundamental assumption underlying SNA is that individuals are interdependent and the connections between individuals have consequences for behavior (Knoke and Yang 2008). Traditionally, SNA was most concerned with the structure and effects of relations between people, groups, or organizations, although there is a growing body of literature that has used SNA to examine psychological attributes (for a full review, see Selden and Goodie 2017). To the best of our knowledge, the only studies to use SNA on psychopathic traits explored the relationship between symptoms in a network (e.g., McCuish et al. 2019) rather than using SNA to explore associations between individual differences in psychopathic traits and social network structure. When SNA is used to focus on individual (i.e., ego) networks, the entire set of social relations of a single individual as well as the social ties among the contacts (the alters) of this individual are considered. Thus, by examining the ego networks of individuals, we can gain a better appreciation of people’s perceptions of their place within their social world.

Studies that have used an ego network approach to understand personality have shown that network characteristics are associated with various personality traits (e.g., greater independence, nonconformity, and a need for change, Burt, Jannotta, and Mahoney 1998; high levels of self-monitoring, Mehra, Kilduff, and Brass 2001). Using a sample of 124 egocentric networks, Kalish and Robins (2006) explored how individual differences might predispose actors to structure their social environment by seeking network closure (i.e., a highly cohesive network of “strong ties”) or by sustaining structural holes—that is, maintaining ties with alters who do not necessarily connect with anyone else in their network. They showed that people characterized by a high level of extraversion tend to keep their close social partners close and actively seek to introduce them to other people. Neurotic people, in contrast, do not keep their social partners so close and tend to have smaller social networks (Kalish and Robins 2006). Although these SNA studies provide evidence to suggest that a person’s network position may act as a reflection of personality characteristics, the extent that this translates to personality disorders has received much less attention.

Two studies have used SNA to examine differences in ego network structure among individuals with personality disorders in a way that is particularly relevant for our purposes. In Clifton, Pilkonis, and McCarty (2007), the authors compared the ego-centered networks of psychiatric patients with borderline personality disorder (BPD) to patients without personality disorders. They found that patients with BPD exhibited marked disturbances in their support seeking such that they sought closeness and support from inappropriate members of their social networks. In Clifton, Turkheimer, and Oltmanns’s (2009) study using the complete networks of 21 groups of military recruits, individuals’ network positions were examined for associations with personality disorders as diagnosed by the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV). They found that numerous pathological personality traits were significant predictors of network position. More specifically, these general associations were consistent with DSM-IV descriptions such that Cluster B disorders (e.g., narcissistic, histrionic, and antisocial) were associated with increased social connections and more central positions in the network, whereas Cluster A and C disorders (e.g., avoidant, schizoid) were negatively associated with these characteristics (Clifton et al. 2009).

These few studies highlight the impact that maladaptive personality traits can have on the way people manage and structure their social networks. Together, findings suggest that a person’s network structure can be examined to better understand personality traits. Thus, a social network approach to understanding the relationship between psychopathic traits and social networks may be an important step toward quantifying and understanding the nature of interpersonal dysfunction in psychopathy.

Current Study

The current study utilizes SNA to examine the personal networks of individuals, focusing on those who score higher on psychopathic traits. We chose to observe self-reported psychopathic traits within core networks (i.e., best friends or intimates: those people from whom one would seek advice, support, or help in times of need) where they are likely to have their strongest effect. That is because it is within these networks that people will generally invest more time and use more of their social capabilities than in less central networks (Kardos et al. 2017). We use effective size to capture network structure, which considers the number of alters that an ego has as well as average number of ties that each alter has to other alters in their network (Burt 1992). High effective size indicates few direct connections among alters. Having disconnected ties can put the ego in aunique—and potentially advantageous—position to broker interactions among network members who otherwise do not interact (Burt 1992). This suggests the ego may be attempting to maximize profit or advantage by acting as “power brokers” by controlling connections and the flow of information among other individuals in the network. Considering that several psychopathic traits relate to behaviors such as interpersonal dominance, manipulation, and egocentricity, we chose to use this measure over other network measures (e.g., density) because it better captures the possibility of strategic brokerage.¹

To asses self-reported psychopathic traits in our community sample, we use the Comprehensive Assessment of Psychopathic Personality (CAPP; Cooke et al. 2004, 2012) and the Psychopathic Personality Inventory-Revised: Short-Form (PPI:SF; Lilienfeld and Hess 2001) because multiple self-report measures should be used to capture the multidimensional nature of psychopathy (Tsang et al. 2018). By using two well-established measures that capture overlapping but different operations of the disorder, we can increase our confidence in our findings that we are capturing the disorder and not constraining our results to how either the PPI:SF or CAPP measure the disorder specifically. These measures were specifically designed to tap into the interpersonal traits of psychopathy and appropriate for use in community settings, making them well suited for use in this study.

Based on the relationship between psychopathy and interpersonal functioning, we hypothesized several associations between psychopathic traits and network structure. Because the CAPP is a tool designed to capture the global interpersonal dysfunction of psychopathy and each domain contains symptoms that would typically be associated with poor interpersonal relationships, we first hypothesized that higher overall scores on the CAPP will be associated with larger effective size. We, however, do not form specific hypotheses regarding the PPI total score or factor scores at this early stage of inquiry, although our results may be able to inform further work from these associations. This is because we acknowledge that PPI comprises traits that have the potential to have an inverse relationship with effective size. For example, the higher-order dimension of Fearless Dominance has typically been viewed as a marker of boldness (Patrick, Fowles, and Krueger 2009) and is associated with traits such as social visibility, gregariousness, and popularity, which could prove to be beneficial in forming and maintaining social ties.

Next, we hypothesized that specific domains/subscales including symptoms that relate to poor interpersonal attachment, emotion, and empathy deficits—specifically, CAPP Attachment and Emotion domains as well as PPI Coldheartedness—will be associated with networks larger in effective size, reflecting interpersonal dysfunction and the inability or lack of desire to maintain close social bonds or emotional attachment with others. As noted previously, a larger effective size would also place the ego in a position of brokerage and offers them the power to control information and the flow of resources within the network (Burt 1992). Therefore, we also hypothesized that domains/subscales including symptoms that relate to manipulation, egocentricity, and interpersonal dominance—specifically, CAPP Dominance, PPI Factor 2, and Machiavellian Egocentricity—will be associated with greater effective size, reflecting their desire to obtain more advantageous position within their network.

Lastly, in an effort to further explore the potential relationships between psychopathic traits and interpersonal dynamics, we examined two variables that account for egos’ perceptions of their core networks. This included how close these egos felt to their core network and also how close they thought the alters in their network would feel to them. Based on the extensive literature on psychopathy and interpersonal dysfunction, we expected to find that egos higher in psychopathic traits would feel less close to those in their network and also believe that others felt less close to them.

Methods

Participants

Participants were 480 undergraduates from a large Canadian university. They ranged in age from 17 to 33 years (M = 19.50, SD = 2.10), and 41 percent were male. With respect to ethnocultural heritage, 39 percent identified as East Asian descent, 31.5 percent as European descent, 21 percent as South Asian descent, and 9 percent as Other descent. All were fluent in English.

Procedure

Overview

Students enrolled in psychology courses were offered the opportunity to participate in the study in exchange for course credit. Those who provided informed consent completed the study online via survey software hosted on a secure web server. In the first part of the survey, participants provided information regarding demographic variables—age, gender, and ethnicity—that are commonly controlled in social network research (Burt 1992; Kalish and Robins 2006). In the second part of the survey, they completed a battery of self-report measures, including self-ratings of symptoms of psychopathic personality disorder. Finally, participants completed a survey about their core social networks. The study took between 60 and 90 minutes to complete. The recruitment and other study procedures were in accordance with the American Psychological Association ethics and approved by the university’s ethics review board.

Social network survey

Participants (egos) were asked to name up to six people (alters) whom they considered to be the most significant or important people in their life (e.g., partner, family members, friends, other significant people). The number of alters was set at six to respect the time constraints of the survey and limit respondent burden. Although we acknowledge that constraining the network size to six individuals does not allow for a complete sample of core networks, a meta-analysis on the different approaches to measuring ego-centered social support networks showed that limitations to network size are most predominant with network composition (Hlebec and Kogovšek 2013). Moreover, six alters is large enough to produce variations in network structure amenable to analysis (Merluzzi and Burt 2013), regardless of composition. This number is also consistent with the literature on self-reported ties of close personal or support networks (e.g., Doeven-Eggens et al. 2008; Dunbar and Spoors 1995; Roberts and Dunbar 2011; Roberts et al. 2008). By asking participants to list six alters, we aimed to avoid potential problems with the distribution of the dependent variable (effective size) where alter numbers are very small and connections between alters would yield unreliable results (e.g., two persons; Kogovšek, Mrzel, and Hlebec 2010). Additionally, participants created pseudonyms for alters to facilitate their identification while maintaining anonymity and were asked to provide the age and gender for each alter as well as any known connections between each possible pair of alters (e.g., “Has Person A ever interacted with Person B?”).

Comprehensive Assessment of Psychopathic Personality

The Comprehensive Assessment of Psychopathic Personality (CAPP; Cooke et al. 2004, 2012) is a conceptual model of psychopathic personality disorder that also focuses on the range of personality traits that capture psychopathy rather than focusing solely on antisocial and criminal behavior. To develop the CAPP, Cooke et al. (2004) reviewed research literature and surveyed the perspectives of clinical experts to develop a 33-symptom concept map of psychopathic personality disorder. The 33 symptoms were rationally allocated into six domains: Attachment, Behavioral, Cognitive, Dominance, Emotion, and Self (see Figure 1; Cooke et al. 2012). Since the inception of the CAPP, several studies have demonstrated the construct validity of the CAPP across multiple settings and populations (Flórez et al. 2018; Hoff et al. 2012; Kreis and Cooke 2011), and the CAPP domains relating to interpersonal style (e.g., Attachment and Dominance) are found to be particularly prototypical of psychopathic personality (Kreis and Cooke 2011). The model comprises 33 symptoms expressed in the form of adjectives or brief adjectival phrases, each defined in turn by another three adjectives or brief adjectival phrases (see Figure 1 for a complete list of symptoms).

Figure 1.

A Concept Map of Psychopathy: The Comprehensive Assessment of Psychopathic Personality (CAPP; Cooke et al. 2004, 2012).

In this study, participants rated themselves on how characteristic each symptom was using a four-point scale (1 = not at all like me, 4 = very like me). The self-ratings on the 33 symptoms were summed to derive domain scores, with higher scores indicating higher levels of psychopathic traits.

Psychopathic Personality Inventory-Revised: Short Form

Beyond the CAPP, a common and well-validated tool to assess psychopathy in community samples has been the Psychopathic Personality Inventory (PPI; Lilienfeld and Andrews 1996) and its revised version (PPI-R). The PPI and PPI-R are arguably the most widely used questionnaire measures of psychopathy specifically constructed for nonclinical and noncriminal samples (Kastner et al. 2012). The PPI-R-SF (Lilienfeld and Hess 2001) is a modified shorter version of the PPI-R and consists of 56 items answered on a four-point Likert scale. Participants rate how false or true they felt the 56 item descriptions are to them (1 = false, 4 = true). PPI-R-SF scores are summed to derive Factor 1 and Factor 2 scores as well as subscale scores for each of the eight domains, with higher scores indicating higher levels of psychopathic traits. The PPI:SF is particularly useful in multimeasure batteries in research settings because of its brevity and efficiency compared to the full-length version (Kastner et al. 2012).

Measures

Effective size

Effective size is calculated based on the total number of reported alters in a network minus the average number of reported connections each alter has to the other alters (Burt 1992). Thus, a high effective size indicates low redundancy between alters relative to network size. Effective size includes decimal data and was initially considered as a continuous variable (M = 2.13, SD = 1.11, range = 1–6, skew = .83, kurtosis = –.05). The distribution was heavily positively skewed (i.e., most of the sample had well-connected core networks, and few had disconnected core networks), and after unsuccessful efforts to normalize the data,² we chose to recode effective size to a categorical variable. This provided us the best option to preserve the natural distribution of the data as well as isolate the minority of individuals who were high in effective size. We used three categories that reflected the natural breakdown of the variable, with most individuals scoring in the low to midrange effective size range and a small minority with scores that indicate a large amount of nonredundant ties (i.e., high effective size): 1 = low effective size (range = 1–1.9) for 38.5 percent of the sample, 2 = medium effective size (range = 2–2.9) for 46 percent of the sample, and 3 = high effective size (range = 3–6) for 14.5 percent of the sample.

CAPP

Each domain represents an independent variable, coded continuously: (1) Attachment, reflecting disturbances in interpersonal attachment or affiliation; (2) Behavioral, reflecting problems with the organization of goal-directed activities; (3) Cognitive, reflecting problems with mental flexibility and adaptability; (4) Dominance, reflecting difficulties with interpersonal agency and assertiveness; (5) Emotional, reflecting problems with affective responses and mood regulation; and (6) Self, reflecting problems with identity or individuality.

PPI:SF

Each factor and subscale represent an independent variable, coded continuously: Factor 1: Fearless Dominance, described as bold and dominant interpersonal style; Factor 2: Impulsive Antisociality, described as disinhibited, self-centered, and ruthless. Next are the eight subscales: (1) Machiavellian Egocentricity, described as manipulative and egocentric in interactions with others; (2) Social Potency, described as charming, influential, and able to manipulate others; (3) Fearlessness, described as eager to take risks, no harm or anxiety concerns; (4) Impulsive Nonconformity, described as a reckless disregard for social norms, unconventional; (5) Carefree Nonplanfulness, described as lack of forethought, fails to learn from consequences; (6) Blame externalization, described as rationalizes behavior and blames others; (7) Stress Immunity, absence of arousal in stressful situations; and lastly, (8) Coldheartedness, described as guiltless, callous, and unemotional.

Closeness rating

Participants were asked to rate how close they felt to each alter on a Likert scale variable (range = 1–5). These scores were then collapsed to (1 = not close/not very close, 2 = somewhat close, 3 = close/very close). Then responses to each alter were summed and divided by the total network size to indicate the average total “closeness rating” for each ego network.

Perceived closeness rating

Additionally, participants were asked to rate whether they thought alters in their network would reciprocate their feelings of closeness on a Likert scale variable (1 = much less than they would, 2 = less than, 3 = equal to, 4 = more than , 5 = much more than they would). The scores were then collapsed to three categories (0 = not reciprocated, 1 = equal to, 2 = greater than they would reciprocate). The ego’s response to each alter was summed and divided by the total network size to indicate an average “perceived closeness rating” for each ego network.

Control variables

The following two network variables and demographic variable (in addition to age and gender) were used as controls because they could all reasonably have an impact on network structure. The homophily principle suggests that similarity breeds connection, or in other words, that contact between similar people occurs at a higher rate than among dissimilar people (McPherson, Smith-Lovin, and Cook 2001). Thus, variations in network size can sometimes be explained by differences in gender (e.g., Dunbar and Spoors 1995) or age (e.g., Roberts et al. 2008). For instance, we could expect that the closer in age that one’s contacts are, the more likely they are to interact in similar social circles. Considering that the current sample also comprises a large number of foreign-born participants (n = 24 percent), we also included this as a control given that foreign-born individuals may have more disparate networks that do not interact given geographical and cultural differences.

Age homophily

This variable refers to the age similarity between egos and their alters in a network. It was coded as the inverse of the mean absolute age difference (in years) between all possible ego-alter dyads. High scores reflect high age homophily.

Gender homophily

This variable refers to the tendency of egos to report networks comprising alters of the same gender. It was calculated as the proportion of ego-alter dyads in a network that were same-gender. High scores reflect high gender homophily.

Foreign born

This variable refers to whether the participant was born outside their country of residence. It was coded dichotomously (0 = yes, 1 = no).

Analytical Strategy

Data were analyzed first at the bivariate level using a series of one-way analysis of variance (ANOVA) with effective size (dependent variable) against network and demographic controls and self-reported psychopathy ratings. Then, only self-reported psychopathy domains (CAPP) or subscales (PPI:SF) that demonstrated significant associations with effective size were included in multivariate analysis.³ Factor 2 Antisocial Impulsivity was also included in multivariate analysis because there were significant differences observed in post hoc testing (see Table 3). The first series of multinomial regressions examines effective size in relation to CAPP scores. As there was high multicollinearity (r = < .600) between CAPP domains and between CAPP domains and total score (r = < .800), we ran separate models to reduce model noise, introducing each variable in succession. We used the same bivariate selection and analytical strategy for a second series of multinomial regressions using PPI:SF. We also tested for interaction effects between domains; however, no such interactions were observed and thus were not reported. We then explored the relationship between the closeness and perceived closeness variables through bivariate associations with the significant CAPP and PPI variables used in the multivariate analysis.

Results

Self-Ratings of Psychopathic Traits

Table A1 in the online appendix⁴ presents the distribution of CAPP total and domain scores. The scores were, on average, low in absolute terms, as would be expected for a sample made of relatively healthy young adults living in the community, but there was some variability among participants. Table A1 also presents Cronbach’s α for the CAPP total scores. As is evident from Table A1, domain scores had large positive correlations with total scores, .79 < r < .86, and moderate to large positive correlations with each other, .50 < r < .69.

Table A2 in the online appendix presents the distribution of PPI factor and subscale scores. The scores were also relatively low, on average, but there was also some variability among participants. Table A2 also presents Cronbach’s α for the PPI factor and subscale scores. As is evident from Table A2, subscale scores had moderate to large positive correlations with total scores, .40 < r < .81.

Network Characteristics

The mean network size was large in absolute terms, M = 5.83 (SD = .71). Most participants were easily able to name six significant others, the maximum allowed by the survey.⁵ Thus, we chose effective size as our basic index of network structure. Table 1 summarizes the bivariate associations between effective size and other demographic and network characteristics, both CAPP domain and total scores as well as PPI:SF factor and subscale scores.

Table 1.

Bivariate Analysis of Effective Size and Demographics, Network Characteristics, and CAPP and PPI:SF Scores

	Lown = 185 (38.5%)	Mediumn = 227 (46.3)	Highn = 68 (14.2%)
Variables	% (n)/ M (SD)	% (n)/ M (SD)	% (n)/ M (SD)	χ²/ANOVA
Age	19.58 (1.87)	19.47 (2.29)	19.39 (2.07)	F = .27(2), p = .76
Gender				χ² = 18.68 (2), p = .00***
Female	93 (19.4)	158 (32.9)	34 (7.1)
Male	92 (19.2)	69 (14.4)	34 (17.4)
Foreign-born				χ² = 3.60(2), p = .17
Yes	148 (40.8)	170 (46.8)	45 (9.4)
No	40 (8.4)	92 (31.9)	100 (34.7)
Age homophily	14.71 (7.57)^c	13.15 (7.08)	11.11 (6.67)^a	F = 4.67(2), p = .01**
Gender homophily	.31 (.34)^b	.43 (.32)^a,c	.30 (.32)^b	F = 7.54(2), p = .00^d***
CAPP
Self	5.76 (3.83)	6.24 (3.63)	6.75 (3.09)	F = 2.08(2), p = .13
Emotion	3.09 (2.44)	3.35 (2.48)	4.16 (2.52)	F = 4.70(2), p = .01*
Dominance	2.65 (2.98)^c	2.55 (2.58)^c	4.07 (3.27)^a,b	F = 7.88(2), p = .00***
Cognitive	3.67 (2.99)	4.03 (2.62)	4.61 (2.70)	F = 2.91(2), p = .06
Behavioral	3.70 (2.90)	3.78 (2.61)	4.48 (2.97)	F = 2.10(2), p = .12
Attachment	1.69 (2.06)^c	1.57 (1.80)^c	2.69 (2.55)^a,b	F = 8.35(2), p = .00***
Total score	20.45 (14.53)^c	21.20 (12.54)^c	26.76 (14.22)^a,b	F = 5.67(2), p = .04**
PPI:SF
Coldheartedness	14.45 (3.70)^c	14.32 (3.83)^c	15.89 (4.32)^a,b	F = 4.55(2), p = .01**
Machiavellian Egocentricity	13.67 (3.56)^c	13.67 (3.74)^c	15.08 (3.72)^a,b	F = 4.07(2), p = .02*
Social Influence	16.75 (4.27)	16.53 (4.27)	16.97 (3.95)	F = .35(2), p = .70
Fearlessness	15.90 (4.91)	16.03 (4.68)	15.88 (4.76)	F = .03(2), p = .97
Rebellious Nonconformity	12.71 (3.81)	12.99 (3.99)	13.37 (4.10)	F = .73(2), p = .48
Blame Externalization	13.48 (4.52)	14.14 (4.35)	14.71 (4.42)	F = 2.27(2), p = .10
Carefree Nonplanfulness	12.92 (3.38)	13.23 (3.21)	13.32 (3.35)	F = .59(2), p = .55
Stress Immunity	16.02 (4.40)	15.56 (4.52)	16.29 (4.53)	F = .87(2), p = .42
Factor 1	48.69 (9.20)	48.12 (9.32)	49.14 (9.54)	F = .39(2), p = .68
Factor 2	52.89 (10.62)^c	54.02 (11.13)	56.49 (9.66)^a	F = 2.80(2), p = .06

Note: N = 480. ANOVA = analysis of variance; CAPP = Cognitive Assessment of Psychopathic Personality; PPI:SF = Psychopathic Personality Inventory-Revised: Short-Form.

Indicates significantly different from Group 1.

Indicates significantly different from Group 2.

Indicates significantly different from Group 3.

Kruskal-Wallis Test.

p < .05. **p < .01. ***p < .001.

With respect to other network characteristics, age homophily had a significant association with network redundancy. Post hoc testing revealed that networks with greater redundancy (i.e., low effective size) were less similar in age. Given that we are studying the network of significant others, this result may have simply captured the typical family dynamics of the participants whereby parents and siblings have wide age differences but interact regularly. Similarly, gender homophily had a significant association with network redundancy. More specifically, post hoc testing indicated that gender homophily was highest in networks with a moderate amount of redundancy, suggesting that individuals in these networks maintain the most similarity in terms of gender, compared to high and low effective size networks.

Finally, with respect to psychopathic traits, there were several significant associations with both the CAPP and the PPI:SF. More specifically, higher total scores on the CAPP and in the Dominance, Attachments, and Emotion domains were significantly associated with a high effective size. In other words, high scores in these domains as well as total score were associated with less redundant ego networks (i.e., fewer ties between alters). For the PPI:SF, higher scores on the Coldheartedness and Machiavellian Egocentricity subscales as well as Factor 2: Impulsive Antisociality were associated with high effective size. Taken together, findings support the general hypothesis that psychopathic traits are associated with network structure.

Regression Analyses

Multinomial logistic regression analyses were conducted to examine the associations between psychopathy and effective size for both the CAPP (Table 2) and the PPI:SF (Table 3).⁶ For the first multinomial regression series on effective size and the CAPP (Table 2), Model 1 included demographic and network controls as well as total score on the CAPP.⁷ Total CAPP score was significantly lower for the low effective size networks (B = –.033, p = .002) compared to high effective size networks; that is, individuals who had more psychopathic traits were more likely to have core social networks associated with less redundancy (i.e., more disconnected). Age homophily was significantly greater for low effective size networks (B = .069, p = .002), indicating that a greater difference in age between respondents and their contacts was associated with more connected networks. Next, Model 2 included the CAPP Dominance domain in addition to demographic and network controls. Dominance domain scores were lower for networks with low (B = –.157, p = .001) and medium (B = –.148, p = .002) effective sizes compared to networks with a high effective size. Findings for age homophily remain stable across this model. This indicates that higher scores on Dominance (e.g., manipulative and domineering interpersonal style) are associated with more disconnected networks even though these networks typically consist of more members of a similar age. Then, Model 3 included the Attachment domain in addition to demographic and network controls. Attachment domain scores were lower for networks with low (B = –.209, p = .002) and medium (B = –.216, p = .001) effective sizes compared to networks with a high effective size. Age homophily was also significantly greater for low (B = .070, p = .002) effective size networks compared to high effective size networks. This indicates that higher scores on the Attachment domain were associated with more disconnected networks. Lastly, Model 4 included the Emotion domain in addition to demographic and network controls. Emotion domain scores were lower for networks with low effective size networks (B = –.176, p = .003) compared to networks with high effective size. This indicates that higher scores on emotion are associated with less redundant networks. Age homophily was also significantly greater for low effective size networks (B = .070, p = .002) compared to networks with high effective size. Taken together, results indicate that self-reported psychopathic traits on the CAPP are associated with differences in effective size even after controlling for relevant demographic and network effects.

Table 2.

Multinomial Regression Predicting Effective Size versus Control Variables and the CAPP

	Model 1		Model 2		Model 3		Model 4
	Lowβ (95% CI)	Mediumβ (95% CI)	Lowβ (95% CI)	Mediumβ (95% CI)	Lowβ (95% CI)	Mediumβ (95% CI)	Lowβ (95% CI)	Mediumβ (95% CI)
Controls
Age	1.30 (.21–7.80)	1.07 (.92–1.24)	1.08 (.92–1.25)	1.07 (.92–1.24)	1.02 (.91–1.23)	1.05 (.90–1.22)	1.01 (.92–1.25)	1.07 (.92–1.24)
Gender (female)	.77 (.24–2.47)	1.62 (.53–4.91)	.83 (.26–2.72)	1.69 (.54–5.22)	.82 (.26–2.72)	1.66 (.53–5.16)	.76 (.23–2.49)	1.61 (.52–5.01)
Foreign born	1.51 (.79–2.88)	1.34 (.71–2.50)	1.62 (.84–3.12)	1.36 (.73–2.56)	1.59 (.83–3.06)	1.34 (.71–2.53)	1.68 (.89–3.22)	1.41 (.77–2.63)
Age homophily	1.51** (.79–2.88)	1.05* (1.00-1.09)	1.07** (1.03–1.12)	1.05* (1.00–1.01)	1.07 (1.03–1.12)	1.05* (1.00–1.09)	1.07** (1.03–1.12)	1.05* (1.00–1.09)
Gender homophily	1.34 (.26–7.97)	1.44 (.27–7.78)	1.20 (.20–7.22)	1.35 (.25–7.38)	1.26 (.21–7.52)	1.40 (.26–7.71)	1.56 (.26–9.53)	1.69 (.26–9.27)
CAPP
Total score	.97** (.95–.99)	.97* (.96–1.00)
Dominance			.86** (.78–.94)	.86** (.79–.95)
Attachment					.81** (.71–.92)	.81** (.71–.92)
Emotion							.84** (.75–.94)	.89* (.80–.99)
Nagelkerke	.10		.12		.12		.10
Model fit (χ²)	42.34***		44.64***		45.10***		41.86***

Note: N = 480. CAPP = Cognitive Assessment of Psychopathic Personality.

p < .05. **p < .01. ***p < .001.

Table 3.

Multinomial Regression Predicting Effective Size versus Control Variables and the PPI:SF

	Model 1		Model 2		Model 3
	Lowβ (95% CI)	Mediumβ (95% CI)	Lowβ (95% CI)	Mediumβ (95% CI)	Lowβ (95% CI)	Mediumβ (95% CI)
Controls
Age	1.08 (.93–1.25)	1.07 (.92–1.24)	1.09 (.93–1.26)	1.08 (.93–1.25)	1.07 (.93–1.25)	1.07 (.92–1.24)
Gender (male)	−.91 (.28–2.93)	1.87 (.61–5.72)	.63 (.17–2.13)	1.40 (.44–4.50)	.85 (.27–2.67)	1.72 (.57–5.17)
Foreign born	1.70 (.90–3.25)	1.42 (.77–2.64)	1.64 (.86–3.13)	1.40 (.75–2.62)	1.62 (.85–3.10)	1.38 (.74–2.57)
Age homophily	1.07** (1.02–1.12)	1.05 (1.00–1.09)	1.07** (1.02–1.12)	1.05 (1.00–1.01)	1.07* (1.03–1.12)	1.04 (1.00–1.09)
Gender homophily	1.34 (.22–7.83)	1.52 (.28–8.18)	1.57 (.26–9.64)	1.70 (.31–9.52)	1.31 (.23–7.60)	1.46 (.28–7.70)
PPI:SF
Factor 2	.97* (.94–.99)	.98 (.95–1.01)
Coldheartedness			.90** (.83–.97)	.93 (.86–1.00)
Machiavellian Egocentricity					.90** (.83–.98)	.91* (.84–1.24)
Nagelkerke	.09		.09		.93
Model fit (χ²)	38.71***		39.85***		39.82***

Note: N = 480. PPI:SF = Psychopathic Personality Inventory-Revised: Short-Form.

p < .05. **p < .01. ***p < .001.

For the next series of multinomial regressions for effective size and the PPI:SF (Table 3), Model 1 included demographic and network controls as well as Factor 2 Antisocial Impulsivity scores. PPI Factor 2 scores were lower for low effective size networks (B = –.034, p = .019) compared to high effective size networks; that is, individuals who had higher ratings on the Antisocial Impulsivity factor were more likely to have significant other networks associated with less redundancy.⁸ Age homophily was also significantly greater for low effective size networks (B = .066, p = .003), indicating that a greater difference in age between respondents and their contacts was associated with the most connected networks reported by participants (i.e., low effective size). Next, Model 2 included Coldheartedness in addition to demographic and network controls. Coldheartedness scores were lower for networks with low effective sizes (B = –.108, p = .008) compared to networks with a high effective size. Findings for age homophily remained stable through this model. This indicates that Coldheartedness significantly differentiates between low and high redundancy networks even after considering age homophily. Lastly, Model 3 included Machiavellian Egocentricity in addition to demographic and network controls. Machiavellian Egocentricity scores were lower for networks with low (B = –.105, p = .010) effective sizes compared to networks with a high effective size. Age homophily was also significantly greater for low (B = .066, p = .003) effective size networks compared to high effective size networks. This indicates that higher scores on Machiavellian Egocentricity were associated with more disconnected networks. Taken together, results indicate that self-reported psychopathic traits on the PPI:SF are associated with differences in effective size even after controlling for relevant demographic and network effects.

Finally, we examined bivariate associations between closeness and perceived closeness ratings with respect to psychopathic traits. We chose to compare the CAPP domains (Attachment, Emotion, and Dominance) and Total Score, as well as the PPI:SF factor 2 and subscales (Coldheartedness and Machiavellian Egocentricity) that were significantly associated with effective size to indirectly provide context into how individuals higher in psychopathic traits perceived their more disconnected social networks. We found several significant associations between higher scores on these psychopathy subscales and domains and the ego’s closeness and perceived closeness ratings (Table 4). Taken together, findings support the general hypothesis that psychopathic traits are associated with feeling less close to others in one’s core networks and the belief that others would feel less close to them.

Table 4.

Bivariate Associations with Closeness and Perceived Closeness against CAPP and PPI and Domain/Subscale Scores

	Closeness rating			F	Perceived Closeness Rating
	Not close/not very close^a	Somewhat close^b	Close/very close^c		Less than^a	Equal to^b	Greater than^c
	M (SD)	(M (SD)	M (SD)		M (SD)	M (SD)	M (SD)	F
CAPP
Attachment	2.70 (2.49)^c	1.80 (2.04)^c	1.56 (1.92)^a,b	6.04**	2.16 (2.33)	1.41 (1.77)^c	2.01 (2.18)^b	5.82 **
Emotion	4.45 (2.56)^c	3.50 (2.39)	3.37 (2.49)^a	7.06**	3.92 (2.47)^b	2.90 (2.52)^a,c	3.65 (2.41)^b	6.80**
Dominance	3.98 (2.95)^c	3.13 (3.07)^c	2.26 (2.57)^a,b	9.51***	3.17 (2.87)	2.18 (2.79)^c	3.31 (2.88)^b	8.90***
Total Score	28.36 (13.82)^c	23.18 (13.74)^c	18.96 (13.02)^a,b	11.62***	24.89 (12.94)^b	18.19 (13.62)^a,c	24.19 (13.29)^b	12.57***
PPI
Coldheartedness	15.54 (4.02)	14.65 (3.75)	14.35 (3.96)	1.83	14.13 (4.10)	14.73 (3.82)	14.61 (3.88)	.61
Machiavellian Egocentricity	14.85 (3.19)	12.36 (3.63)	13.40 (3.78)	4.67	14.58 (3.93)	13.41 (3.62)	14.20 (3.65)	3.72
Factor 2	58.38 (9.12)^c	54.66 (10.54)	53.94 (10.78)^a	6.95**	56.31 (11.14)^b	51.46 (10.37)^a,c	55.65 (10.61)^b	10.08***

Note: CAPP = Cognitive Assessment of Psychopathic Personality; PPI = Psychopathic Personality Inventory. ^a,b,cSignificant difference between groups (p < .10).

p < .05. **p < .01. ***p < .001.

Discussion

This study used network data to examine how psychopathic traits are associated with social relationship structure, specifically, relationships with core networks among a sample of university students. We found evidence to suggest that people with higher levels of overall psychopathic traits, as measured by the Cognitive Assessment of Psychopathic Personality (CAPP), perceived the structure of their social networks differently. More specifically, higher CAPP scores were associated with nonredundant networks compared to those with more connected core networks. Our findings also provide evidence to support the possibility that people with higher levels of psychopathic traits structure their core social networks with in ways to allow themselves to reap benefits from their relationships. Given that both the CAPP domains and the Psychopathic Personality Inventory (PPI) subscales associated with interpersonal dominance, egocentricity, and manipulation were associated with low redundancy in core networks, this relative disconnection of alters may be a deliberate strategy on behalf of the ego. This finding is reflective of the core traits of psychopathic personality disorder as evidenced in the theoretical literature and assessment of the disorder. From the first conceptions of the disorder (Cleckly 1941) to current diagnostic and assessment standards (American Psychiatric Association 2013; World Health Organization 1992), a marker of psychopathic personality disorder is someone who is getting more then they give in social and economic relationships through deceit, cunning, manipulation, and exhibiting chronic and lifelong patterns of irresponsibility in relationships and parasitic lifestyles.

In studies on organizational and management settings, sparse networks have been found to afford brokers greater influence and power over alters by permitting them to control the flow of information in the network and as a consequence, the images they present to alters (Burt 1992). For example, low redundancy networks provide brokers benefits such as greater access to information as well as better job opportunities, work performance, and status attainment (Mehra et al. 2001). Thus, as suggested in our findings, it is possible that individuals high in psychopathic traits are more likely to structure their network in a way that allows them to obtain greater perceived benefits. This would be in line with Foulkes and colleagues’ (2014) study that concluded social interactions of people high on psychopathy are likely motivated by instrumental gain, such as monetary or status attainment, and not by the need for closeness and affiliation. Interestingly, we also found that higher scores in the dominance domain were associated with a greater likelihood of the ego reporting feeling less close to alters.

Although we consider the findings in the context of the current literature to support the inference that this is a deliberate strategy used by individuals high in psychopathic traits, an alternative explanation is that the larger effective size is related to problematic traits affecting their ability or their lack of desire to establish or maintain a close, connected network. Some support for this can be gleaned from our findings that both the CAPP and Psychopathic Personality Inventory-Revised: Short-Form (PPI:SF) domains and subscales related to poor attachment and emotional deficits (e.g., callous/unemotional traits) were also associated with low redundancy in the core network. For example, in a study that used the PPI:SF on a subclinical community sample, Uziebolo and colleagues (2010) found that the Coldheartedness factor related to diminished empathic abilities and had a tendency to value friendships less. Similarly, we also found that higher scores in the CAPP total scores and Emotion domains as well as the Factor 2 (Antisocial Impulsivity) in the PPI:SF were associated with a greater likelihood that the ego reported feeling not very close to the alters within their core networks. More insight into the ego’s motivation for maintaining smaller networks would help to clarify whether this is the result of an unwillingness on behalf of the ego to maintain close ties and emotional bonds with others, or whether it is the ego’s behavior that deters others from them, or a combination of both. For example, both higher CAPP total and Factor 2 scores were also associated with a greater likelihood that the ego believed alters in the network would not reciprocate feelings of closeness in the same way that they would themselves. It may also be that individuals high on psychopathic traits deliberately chose to maintain smaller networks because they lack the desire for close social bonds, leading them to maintain only the relationships they feel they can reap the most benefits from (e.g., monetary or status attainment), not those that offer emotional connection or attachment.

We also must consider the possibility that other factors could be at play. For instance, we found that increases in age homophily were associated with high effective size networks. As family members are more likely to know one another, this could, in part, explain some of the differences found between effective size groups. We found CAPP and PPI scale scores to be significantly associated with larger effective sizes even when controlling for age homophily, suggesting that higher effective size could not be explained solely by whether participants named a predominantly familial or friendship network.

It has also been argued that smaller and less dense networks were reflective of individuals who simply want to avoid the stress of managing a large network of friends and family members (Kleim 2011). However, given the fact we restricted the number of alters to six and there was not a great variation in the absolute network size, this was unlikely a factor.

Finally, given that this was a sample of university students with a substantial percentage who were foreign born, it is also possible that core networks among international students may involve less ties (e.g., between their family and new friends). However, inclusion of foreign-born status as a control variable was not found to be significant.

We believe the current study findings provide important insights into the relationship between individuals high in psychopathic traits and their perceptions of their core networks; however, there are important data constraints to discuss in light of these findings. First, ego networks were constrained to six people and to core networks only. By constraining network size to six alters, we do not have a complete sample of all possible core network members and thus are only able to interpret our results within the confines of this particular study. Given the initial study design consisted of several self-reported psychological instruments—of which this study focused only on two related to psychopathy—allowing for unconstrained networks was unfortunately not an option given the time limitations of the survey and the risk of respondent fatigue. We acknowledge that this is a limitation of the present study, and we encourage future research to explore the relationship between psychopathic traits using unconstrained core networks to determine if similar patterns are observed.

Second, and importantly, we must acknowledge limitations to the network survey design. When asking respondents to evaluate ties between alters, we created a broad measure: “Has Person A ever interacted with Person B?” In doing so, the relationship between alters could be overstated. In the future, we would also include a tie question that allows us to capture a more nuanced understanding of how connected alter ties truly are (e.g., “Is it likely that Person and Person B would interact independently of you?”). It is also noteworthy to mention the limited data on the composition of the core networks outside of age and gender homophily. This has implications for the ability to draw conclusions about the disconnection among alters in those who scored higher on psychopathic traits. Our study design would have benefitted from considering the number and types of relationships within these core networks (e.g., romantic partners, family, and friends) as well as other possible factors (e.g., whether one is an only child, has moved many times) that would aid in the interpretation of whether the perceived disconnection of alter network members was intentional of the ego (e.g., due to emotional deficits, callousness, to reap benefits, to avoid stress) or unintentional (e.g., because they reported having fewer family members in the core networks).

A third limitation includes the measurement of psychopathic traits and social networks via self-ratings. Because the criteria used to rate personality disorders tend to be highly evaluative, this can lead to defensiveness and cognitive distortions in self-report (Clifton et al. 2009). Thus, sole reliance on self-ratings in research on psychopathy is problematic given the nature of personality disorders in general and psychopathic personality disorder in particular (e.g., Hart and Storey 2013). Additionally, we did not have validity measures to ensure participants were responding to the measures in a careful, meaningful, consistent, and valid manner. Fortunately, the Cronbach’s α and the correlations between the CAPP and PPI scale scores were as expected and consistent with those reported in the general literature, suggesting invalid responding is not a major concern. By focusing on ego networks, we considered only the ego’s perspective. Future research should expand network measures to take into consideration the perspective of the alters, considering that other studies have identified biased perceptions of relationships on the part of individuals high in psychopathic traits (e.g., Muñoz et al. 2008).

Fourth, the study relied on a sample of university students. This allowed us to recruit a large sample of cooperative and literate participants, but university students of course tend to be younger than the general population and advantaged in terms of education, income, and intelligence. They may also have unique network characteristics such as the distribution of relationships with family, friends, and romantic partners (Furman and Buhrmester 1992), which could influence the distribution of alter connections relative to other age groups.

Lastly, we used one network measure, effective size, to measure social network structure in a small core network. This study was arguably the first step toward a potentially more extensive research program examining the role of personality traits in shaping social networks. Future research should use more varied network measures to capture network structure and test these measures on different types of social networks to expand our understanding of the effects of psychopathic traits on individual differences in network structure.

Notwithstanding the aforementioned limitations, we believe our findings support the use of social network analysis (SNA) when examining social relationships and psychopathic traits because it can help to reveal structural differences that may otherwise go unnoticed. Future studies should continue to use SNA to explore the relationship between brokerage and psychopathy as well as the emotional detachment and callous/unemotional traits and how this impacts network structure from both the egos’ and alters’ perspectives. This requires both the use of repeated measures designs and qualitative, contextual data that capture the ways in which respondents’ approach and manage their interpersonal relationships. We hope our findings contribute to the development of a research agenda that can address the theoretical mechanisms that link psychopathic traits to personal network structures in both general and criminal populations.

Supplemental Material

Reale_et_al._Supplemental_Material – Supplemental material for Are Psychopathic Traits Associated with Core Social Networks? An Exploratory Study in University Students

Supplemental material, Reale_et_al._Supplemental_Material for Are Psychopathic Traits Associated with Core Social Networks? An Exploratory Study in University Students by Kylie S. Reale, Martin Bouchard, Yan L. Lim, Alana N. Cook and Stephen D. Hart in Social Psychology Quarterly

Footnotes

ORCID iD

Kylie S. Reale

1

Exploratory analyses showed that the results for measures such as density were relatively similar, though lacking the clarity of those analyses tapping into strategic network behavior.

2

Natural logarithm transformation did not adequately improve the distribution of effective size. Exploratory analyses with alternative measures (e.g., density) revealed similar patterns.

3

We ran a model that included all CAPP domains as well as a model with all the PPI subscales to address the possibility that certain subdimensions exhibit suppressor effects. There were no differences between significant variables found in these models compared to those presented; therefore, the most parsimonious model is presented.

4

The supplemental appendix is available with the of the paper.

5

Almost all participants (n = 453, 94.5 percent) listed six core network members, and there were no alters who reported fewer than two ties. As a robustness check, we conducted all multivariate models with and without participants who reported fewer than six alters; no differences between models were observed.

6

Due to the number of analyses performed in order to reduce the risk of Type 1 error, in-text reporting of multivariate analyses includes only values that are significant at less than or equal to p < .010.

7

To acknowledge the possibility that the control factors and covariates may inadvertently remove variance relevant to psychopathy, we tested models with and without control variables. There were no differences in statistically significant psychopathy domains or subscales between models with or without controls included.

8

These results are not discussed further as they did not meet the p < .01 threshold.

Bios

Kylie S. Reale is a PhD candidate at the School of Criminology at Simon Fraser University. Kylie’s research interests are in the area of homicide and sexual violence, particularly in situational approaches to crime, offender strategies to avoid detection, and offender profiling. She received the Doctoral Award from the Social Sciences and Humanities Research Council in 2019.

Martin Bouchard is a professor of criminology at Simon Fraser University, where he leads the Crime and Illicit Networks Laboratory. His research focuses on the dynamics of illicit markets as well as the role of social networks in a variety of criminal phenomena, including street gangs and organized crime. He is the 2018–2019 recipient of the Western Society of Criminology (WSC) Fellows Award for individuals associated with the Western region who have made important contributions to the field of criminology.

Yan L. Lim is a PhD student in clinical forensic psychology at Simon Fraser University. Yan’s research interests are in the area of forensic psychology, particularly in violence risk assessment, management and assessment, and management of psychopathic personality disorder.

Alana N. Cook, PhD, is a registered psychologist in private practice in British Columbia, Canada, and adjunct faculty in the Department of Psychology at Simon Fraser University. Her research interests are at the intersection of psychology and law, specifically on psychological and forensic assessment and intervention, assessment and management of psychopathic personality disorder, and national and international policy. She received the Christopher Webster Young Scholar Award from the International Association of Forensic Mental Health Services in 2014.

Stephen D. Hart is a professor of psychology and director of the Terrorism, Risk, and Security Studies graduate program at Simon Fraser University as well as visiting professor of psychology at the University of Bergen. His research interests are in the area of clinical-forensic psychology, particularly violence risk and psychopathic personality disorder. He was awarded the 2019 Distinguished Contribution Award from the American Academy of Forensic Psychology.

References

Ali

Farrah

Chamorro-Premuzic

Tomas

. 2010. “The Dark Side of Love and Life Satisfaction: Associations with Intimate Relationships, Psychopathy and Machiavellianism.”Personality and Individual Differences 48(2):228–33.

American Psychiatric Association, DSM-5 Task Force. 2013. Diagnostic and Statistical Manual of Mental Disorders: DSM-5 (5th ed.) Arlington, VA: American Psychiatric Association.

Blackburn

Ronald

. 2006. “Other Theoretical Models of Psychopathy.” Pp. 35–57 in The Handbook of Psychopathy, edited by Patrick

C. J.

New York, NY: Guildford.

Burt

Ronald S.

1992. Structural Holes: The Social Structure of Competition. Cambridge, MA: Harvard.

Burt

Ronald S.

Jannotta

Joseph E.

Mahoney

James T.

1998. “Personality Correlates of Structural Holes.”Social Networks 20:63–87.

Cleckley

Hervey M.

1941. The Mask of Sanity. St. Louis, MO: Mosby.

Clifton

Allan

Pilkonis

Paul A.

McCarty

Christopher

. 2007. “Social Networks in Borderline Personality Disorder.”Journal of Personality Disorders 21(4):434–41.

Clifton

Allan

Turkheimer

Eric

Oltmanns

Thomas F.

2009. “Personality Disorder in Social Networks: Network Position As a Marker of Interpersonal Dysfunction.”Social Networks 31(1):26–32.

Conradi

Henk Jan

Boertien

Sanne Dithe

Cavus

Hal

Verschuere

Bruno

. 2016. “Examining Psychopathy from an Attachment Perspective: The Role of Fear of Rejection and Abandonment.”The Journal of Forensic Psychiatry & Psychology 1:92–109.

10.

Cooke

David J.

Hart

Stephen D.

Logan

Caroline

Michie

Christine

. 2004. “Comprehensive Assessment of Psychopathic Personality-Institutional Rating Scale (CAPP-IRS).” Unpublished manuscript. Department of Psychology, Glasgow Caledonian University.

11.

Cooke

David J.

Hart

Stephen D.

Logan

Caroline

Michie

Christine

. 2012. “Explicating the Construct of Psychopathy: Development and Validation of a Conceptual Model, the Comprehensive Assessment of Psychopathic Personality (CAPP).”International Journal of Forensic Mental Health 11(4):242–52.

12.

Doeven-Eggens

Lilian.

De Fruyt

Filip

Hendriks

Jolijn A. A.

Bosker

Roel J.

Van der Werf

Margaretha P. C.

2008. “Personality and Personal Network Type.”Personality and Individual Differences 45(7):689–93.

13.

Dunbar

Robin I. M.

Spoors

Matt

. 1995. “Social Networks, Support Cliques, and Kinship.”Human Nature 6(3):273–90.

14.

Flórez

Gerardo

Ferrer

Ventura

García

Luis S.

Crespo

Maria R.

Pérez

Manual

Saíz

Pilar A.

Cooke

David J.

2018. “Clinician Ratings of the Comprehensive Assessment of Psychopathic Personality (CAPP) in a Representative Sample of Spanish Prison Inmates: New Validity Evidence.”PLoS ONE 13(4):e0195483.

15.

Foulkes

Lucy

Seara-Cardoso

Ana

Neumann

Craig

Rogers

John

Viding

Essi

. 2014. “Looking after Number One: Associations between Psychopathic Traits and Measures of Social Motivation and Functioning in a Community Sample of Males.”Journal of Psychopathology and Behavioral Assessment 36(1):22–29.

16.

Furman

Wyndol

Buhrmester

Duane

. 1992. “Age and Sex Differences in Perceptions of Networks of Personal Relationships.”Child Development 63(1):103–15.

17.

Hare

Robert D.

Neumann

Craig S.

2008. “Psychopathy as a Clinical and Empirical Construct.”Annual Review of Clinical Psychology 4:217–46.

18.

Hare

Robert D.

Neumann

Craig S.

2010. “The Role of Antisociality in the Psychopathy Construct: Comment on Skeem & Cooke.”Psychological Assessment 22(2):446–54.

19.

Hart

Stephen D.

Storey

Jen S.

2013. “Clinical and Forensic Issues in the Assessment of Psychopathy.” Pp. 556–78 in The Handbook of Forensic Psychology. 4th ed., edited by Weiner

Otto

R. K.

New York, NY: Wiley.

20.

Hlebec

Valentina

Kogovšek

Tina

. 2013. “Different Approaches to Measure Ego-Centered Social Support Networks: A Meta-analysis.”Quality & Quantity 47(6):3435–55.

21.

Hoff

Helge A.

Rypdal

Knut

Mykletun

Arnstein

Cooke

David J.

2012. “A Prototypicality Validation of the Comprehensive Assessment of Psychopathic Personality Model (CAPP).”Journal of Personality Disorders 26(3):414–27.

22.

Kalish

Yuval

Robins

Garry

. 2006. “Psychological Predispositions and Network Structure: The Relationship between Individual Predispositions, Structural Holes and Network Closure.”Social Networks 28(1):56–84.

23.

Kardos

Peter

Leidner

Bernhard

Pléh

Csaba

Soltész

Peter

Unoka

Zsolt

. 2017. “Empathic People Have More Friends: Empathic Abilities Predict Social Network Size and Position in Social Network Predicts Empathic Efforts.”Social Networks 50:1–5.

24.

Kastner

Rebecca M.

Sellbom

Martin

Lilienfeld

Scott O.

2012. “A Comparison of the Psychometric Properties of the Psychopathic Personality Inventory Full-Length and Short-Form Versions.”Psychological Assessment 24(1):261–67.

25.

Kleim

Sylvia

. 2011. Social Networks and Family Formation Processes: Young Adults’ Decision Making about Parenthood. Wiesbaden: VS Verlag für Sozialwissenschaften.

26.

Knoke

David

Yang

Song

. 2008. Social Network Analysis. Thousand Oaks, CA: Sage.

27.

Kogovšek

Tina

Mrzel

Maja

Hlebec

Valentina

. 2010. “Please Name the First Two People You Would Ask for Help: The Effect of Limitation of the Number of Alters on Network Composition.”Metodoloski zvezki 7(2):95–106.

28.

Kreis

Mette K.

Cooke

David J.

2011. “Capturing the Psychopathic Female: A Prototypicality Analysis of the Comprehensive Assessment of Psychopathic Personality (CAPP) across Gender.”Behavioral Sciences & the Law 29(5):634–48.

29.

Lilienfeld

Scott O.

1998. “Methodological Advancements and Developments in the Assessment of Psychopathy.”Behaviour Research and Therapy 36:99–125.

30.

Lilienfeld

Scott O.

Andrews

Brian P.

1996. “Development and Preliminary Validation of a Self-Report Measure of Psychopathic Personality Traits in Non-Criminal Population.”Journal of Personality Assessment 66(3):488–524.

31.

Lilienfeld

S. O.

Hess

Tanya

. 2001. “Psychopathic Personality Traits and Somatization: Sex Differences and the Mediating the Role of Negative Emotionality.”Journal of Psychopathology and Behavioural Assessment 23(1):11–24.

32.

McCord

William. M.

1982. The Psychopath and Milieu Therapy. New York, NY: Academic Press.

33.

McCuish

Evan

Bouchard

Martin

Beauregard

Eric

Corrado

Raymond

. 2019. “A Network Approach to Understanding the Structure of Core Symptoms of Psychopathic Personality Disturbance in Adolescent Offenders.”Journal of Abnormal Child Psychology 47:1467–82.

34.

McPherson

Miller

Smith-Lovin

Lynn

Cook

James

. 2001. “Birds of a Feather: Homophily in Social Networks.”Annual Review of Sociology 27(1):415–44.

35.

Mehra

Ajay

Kilduff

Martin

Brass

Daniel J.

2001. “The Social Networks of High and Low Self-Monitors: Implications for Workplace Performance.”Administrative Science Quarterly 46(1):121–46.

36.

Merluzzi

Jennifer

Burt

Ronald S.

2013. “How Many Names Are Enough? Identifying Network Effects with the Least Set of Listed Contacts.”Social Networks 35(3):331–37.

37.

Muñoz

Luna C.

Kerr

Margaret

Besic

Nejra

. 2008. “The Peer Relationships of Youth with Psychopathic Personality Traits: A Matter of Perspective.”Criminal Justice and Behavior 35(2):212–27.

38.

Patrick

Christopher J.

Fowles

Don C.

Krueger

Robert F.

2009. “Triarchic Conceptualization of Psychopathy: Developmental Origins of Disinhibition, Boldness, and Meanness.”Developmental Psychopathology 21(3):913–38.

39.

Rauthmann

John F.

2011. “The Dark Triad and Interpersonal Perception: Similarities and Differences in the Social Consequences of Narcissism, Machiavellianism, and Psychopathy.”Social Psychology and Personality Science 3(4):487–96.

40.

Roberts

Sam G.

Dunbar

Robin I. M.

2011. “Communication in Social Networks: Effects of Kinship, Network Size, and Emotional Closeness.”Personal Relationships 18(3):439–52.

41.

Roberts

Sam G. B.

Wilson

Ruth

Fedurek

Pawel

Dunbar

Robin I. M.

2008. “Individual Differences and Personal Social Network Size and Structure.”Personality & Individual Differences 44(4):954–64.

42.

Selden

Mary

Goodie

Adam S.

2018. “Review of the Effects of Five Factor Model Personality Traits on Network Structures and Perception of Structures.”Social Networks 52:81–99.

43.

Skeem

Jennifer L.

Poythress

Norman

Edens

John F.

Lilienfeld

Scott O.

Cale

Ellison M.

2003. “Psychopathic Personality or Personalities? Exploring Potential Variants of Psychopathy and Their Implications for Risk Assessment.”Aggression and Violent Behavior 8(5):513–46.

44.

Tsang

Siny

Salekin

Randall T.

Coffey

Adam C.

Cox

Jennifer

. 2018. “A Comparison of Self-Report Measures of Psychopathy among Nonforensic Samples Using Item Response Theory Analyses.”Psychological Assessment 30(3):311–27.

45.

Uzieblo

Katarzyna

Verschuere

Bruno

Van den Bussche

Eva

Crombez

Geert

. 2010. “The Validity of the Psychopathic Personality Inventory–Revised in a Community Sample.”Assessment 17(3):334–46.

46.

World Health Organization. 1992. The ICD-10 Classification of Mental and Behavioral Disorders: Clinical Descriptions and Diagnostic Guidelines. Geneva: Work Health Organization.

47.

Zhu

Xiumei

Woo

Sang

Porter

Caitlin

Brzezinski

Michael

. 2013. “Pathways to Happiness: From Personality to Social Networks and Perceived Support.”Social Networks 35(3):382–93.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB