Patterns of relational aggression,narcissism,and self-esteem: Adolescents’ social goals unraveled

Abstract

This study applies latent profile analysis to identify profiles of adolescents differentiated on levels of relational aggression (RAgg), self-esteem, and narcissism. To understand the social aspects of these profiles within the adolescent peer context, we compare them to their reported social goals of dominance, popularity, and intimacy. Greek junior high school students (N = 2,207), selected via random stratified cluster sampling, aged 13–16 (M = 14.04, SD =.81), completed a self-report survey. Based on self-esteem, narcissism, and RAgg scores, four groups emerged: the group with the lowest scores in RAgg, narcissism, and self-esteem (“low risk”), the group with the highest scores in RAgg, narcissism, and self-esteem (“relational aggressors with narcissistic self-esteem”), the group with moderate scores in RAgg, narcissism, and self-esteem (“moderate risk”), and the group with high RAgg and narcissism scores but low self-esteem (“relational aggressors with vulnerable self-esteem”). Relational aggressors with high narcissism and self-esteem scored higher on social goals than those with vulnerable self-esteem (high narcissism but low self-esteem). Preventive policies should replace excessive and unconditional praise with more realistic self-esteem sources and teach adolescents how to respond to negative feedback and to cope with ego threats or social placement concerns.

Keywords

relational aggression narcissism self-esteem social goals adolescence latent profile analysis

Relational aggression (RAgg) refers to covert and volitional acts intending to harm or manipulate an individual's social relationship and status, through rumor spreading, exclusion, or peer rejection (Voulgaridou & Kokkinos, 2015, 2019). In this study, relationally aggressive behavior is conceptualized as a set of strategies available to adolescents to maintain their social goals, which are defined as reflecting underlying interpersonal needs that determine individuals’ interpersonal relationships (Buhrmester, 1998). Adolescents are more prone to participate in RAgg if they perceive it as an efficient technique for achieving social goals. This perspective is consistent with the well-documented view of the literature that it is oversimplifying to explain adolescent aggression only in terms of social competence deficits (Yeager et al., 2018). This is because aggression that successfully serves adolescents’ social goals—planned, strategic attacks on others’ reputations for the sake of enhancing or maintaining one's social prominence, while preserving social status and avoiding punishment—requires a certain degree of social awareness (Faris & Felmlee, 2011). Therefore, emerging evidence suggests that RAgg is used to pursue adolescents’ social goals (e.g., Mayeux, 2014).

Recent evidence reveals that relational forms of aggression and their association with adolescents’ social goals are of substantial concern among students (McEachern & Snyder, 2012), making the present research especially relevant for school practitioners. According to Rigby (2004), there are various intervention strategies that can be implied depending on how RAgg and its correlates are perceived. To provide a specific example, if RAgg is theorized as a way to achieve social goals a socio-cognitive intervention pattern could be employed by both educators and school psychologists (Kerig, 2018). Such programs would aid students in learning to recognize and manage their emotions, recognize others’ physiological arousal, develop coping mechanisms, understand the motivations of others, and create and evaluate viable alternatives to aggressive behaviors (Leff et al., 2010). Thus, the investigation of the social aspects of distinct aggressive profiles can help understand the adolescent peer context.

Ιmportantly, according to Morf and Rhodewalt's (2001) self-regulatory model, self-regulation on one's social goals is tightly associated with individual differences, such as particular personality traits (i.e., narcissism) and other dispositional characteristics including self-esteem. The majority of the existing research on the correlates of narcissism and self-esteem has been conducted using variable-centered methodological approaches. In comparison, person-centered techniques, such as Latent Profile Analysis (LPA), identify groups of individuals who are more similar to one another on certain attributes or relations among attributes than individuals from distinct groups (Laursen & Hoff, 2006). A literature review showed that prior work has employed LPA to evaluate various combinations of self-esteem and narcissism among adults (Bushman et al., 2009), concluding with mixed results (various profiles, different combinations of narcissism, and self-esteem). However, it is unclear whether these findings generalize to adolescence, a period during which narcissism reaches its developmental peak, while self-esteem is reduced. Taken together, this study's major contribution is the adoption of a person-centered approach to identify and disentangle the potential alternative patterns of the aforementioned three constructs (i.e., RAgg, narcissism, and self-esteem) on adolescent social goals.

Three categories of social goals, which have been extensively investigated by peer relations theorists, are explored in this study. The first includes goals oriented to power (social dominance goals), defined as the degree to which an individual strives for dominance acquisition among peers (Mayeux, 2014). The second type of social goal has been referred to as popularity, namely the extent to which a person is popular according to peers’ evaluation (van den Broek et al., 2016). Finally, the third is oriented to interpersonal relations and is mirrored in goals associated with closeness, approval, and intimacy (Kiefer & Ryan, 2008). This study aims at identifying individual differences in terms of narcissism, self-esteem, and relationally aggressive behavior in adolescents’ social goals with the aim of advancing existing practical implications regarding the achievement of social goals in a non-aggressive way. For example, not all adolescents prioritize social dominance or popularity to the same extent, even though adolescence is a developmental period characterized by an increased emphasis on peer social evaluation and acceptance (Yeager et al., 2018). More precisely, we expect that adolescents with certain personality traits (i.e., narcissism) and specific scores on self-esteem and RAgg may score higher on the social goals of dominance and popularity.

While self-esteem and narcissism share certain similarities in that both entail positive self-evaluation (Brummelman et al., 2016), they also have significant distinctions. For instance, narcissists are defined as having not only extremely favorable self-views but also a sense of entitlement—they exploit others and lack empathy for them (Horvath & Morf, 2010). Further, they view themselves as superior to others on agentic (dominance and popularity in the peer group), but not on more communal traits (intimacy). In contrast, high self-esteem individuals perceive themselves as better in both domains (Campbell et al., 2010). Bushman et al. (2009) claim that aggressive behaviors in response to ego-threat are specific to narcissism, and are not attributable to its overlap with self-esteem among youth. Research indicates that one of the narcissists’ social goals in adulthood is to assert their social dominance over others, to demonstrate their grandiosity and superiority by self-promoting in front of important people, and to enhance their popularity among peers (Horvath & Morf, 2010). During adolescence, narcissists seek to develop their grandiose self-view by acquiring social dominance in their peer network and obtaining admiration through aggressive behavior (Fanti & Henrich, 2015), while they may be motivated to behave aggressively when they perceive that their grandiose self-image and popularity are threatened within the peer group (Onishi et al., 2012). Narcissism has been related to all forms of aggression (direct, indirect, physical, and relational). The exploration of aggressive conduct has relied on several theoretical frameworks highlighting the role of various factors (i.e., personality traits and social context). Practically speaking, intervention efforts emphasizing solely adolescents’ self-esteem may not be adequate as it is not clear how self-esteem affects involvement in RAgg as well as adolescents’ social goals. The potential interplay between self-esteem and narcissism highlights an additional factor that determines the role of self-esteem. Therefore, adolescents’ narcissistic traits should not be ignored by school psychologists when they apply intervention strategies aiming to increase their self-esteem (Campbell et al., 2010).

A sound theoretical background that has been systematically applied to the interpretation of aggressive behavior is the General Aggression Model (GAM; Anderson & Bushman, 2002). GAM has been used in RAgg research among adolescents (Kokkinos et al., 2016, Kokkinos et al., 2017; Kokkinos & Voulgaridou, 2017; Voulgaridou & Kokkinos, 2020), positing that individual traits, such as personality characteristics (e.g., narcissism) and self-esteem are associated with relationally aggressive behavior. The sole and considerable association between narcissism and RAgg in both childhood community samples and adolescent clinical samples is not surprising (Bell et al., 2018). Particularly, it has been argued that narcissistic preadolescents typically display behaviors that may resemble RAgg attributes (Onishi et al., 2012). Adolescents high on narcissism tend to present a favorable image through the use of RAgg, in the sense that it provides a means of acting against others and achieving or maintaining social status. As a result, this form of antisocial conduct may help narcissistic individuals to be powerful and dominant in their peer group (Bell et al., 2018; Vaillancourt & Krems, 2018). Thus, adolescents scoring high on narcissism may use RAgg as an efficient way to achieve a privileged position in the peer group and to garner appraisal responses from others (Coyne et al., 2006).

In contrast, self-esteem reflects more communal concerns, as it is determined by an individual's lifetime experiences with social acceptances and rejections (Leary, 2004). For example, in Park and Maner's (2009) study, undergraduate students with high on self-esteem expressed an increased desire to seek contact and intimacy with others in order to restore their self-esteem following unfavorable criticism. On the other hand, individuals low on self-esteem chose to avoid social contact, thereby avoiding the risk of further rejection. Regarding the association between self-esteem and RAgg, the evidence seems controversial. Low self-esteem has been associated with higher RAgg (e.g., Fanti & Henrich, 2015), while in other studies (e.g., Puckett et al., 2008) high self-esteem was found to be related to RAgg, implying that high self-esteem adolescents may use these tactics as more socially acceptable in order to avoid the risk of social rejection and preserve their social goals for dominance and popularity (Horvath & Morf, 2010). Combined with narcissism, high self-esteem buttresses narcissistic adolescents’ grandiose but fragile ego and thereby reduces their experience of ego-threat (Fanti & Henrich, 2015), which arises when an inflated self is perceived as under attack leading to aggressive responding. Thus, narcissism combined with high self-esteem should relate weakly to an array of psychological responses symptomatic of ego-threat, such as RAgg among early adolescents (Ojanen et al., 2012). Similarly, the combination of a low or a fragile self-view, indicative of low self-esteem, with a grandiose self-concept or a desire to be perceived grandiosely by others, indicative of high narcissism, may contribute to the continuation or escalation of relationally aggressive behavior (Fanti & Henrich, 2015). Therefore, the exploration of the association between narcissism and self-esteem will inform preventive policies that promote positive self-image and teach adolescents strategies for coping with ego threats.

Regarding the social goal of intimacy, the findings are inconsistent. Specifically, while according to some evidence adolescents who engage in RAgg are unlikely to value social closeness and intimacy (Kokkinos et al., 2020), other studies have shown a positive association between RAgg and friendship intimacy and exclusivity during adolescence (Banny et al., 2011). Youth who manipulate others and score high on narcissism tend to take advantage of their friends to gain benefits without remorse or they may pick friends that can easily exploit. Thus, individuals scoring high in narcissism and RAgg are likely to have low intimacy goals.

The current study

The aim of the current study is to identify distinct individual profiles based on levels of narcissism, self-esteem, and RAgg. To understand the social goals associated with these profiles within the adolescent peer context, we aim to compare the identified profiles on levels of reported social goals of dominance, popularity, and intimacy. Due to the lack of previous research investigating constellations of such factors, no a priori hypotheses about the number and patterns of groups that would emerge were formulated. As this study targets a community sample of adolescents, it was expected that the majority of participants would be classified as low-risk, with at least moderate scores on self-esteem, as well as low scores on narcissism and RAgg. Additionally, given the comorbidity of psychosocial difficulties, an at-risk group characterized by high RAgg, high narcissism, and low self-esteem was expected to be identified. As the findings regarding the association between self-esteem and narcissism are conflicting, a latent-profile analysis could unveil the distinct combinations of high narcissism with high (narcissistic) or low (vulnerable) self-esteem. This study also explores the differential predictivity of the identified profiles for the emergence of distinct social goals. Although the combined effect of high RAgg, high narcissism, and low self-esteem has not been explored yet, a profile with high scores on these constructs is expected to be related to the social goal of dominance. In contrast, a low-risk profile defined by low scores on RAgg, narcissism, and moderate self-esteem may be associated with the social goal of intimacy.

In terms of gender, while males are more socially dominating, females prefer peer group popularity and higher levels of intimacy (Mayeux, 2014). Moreover, males show higher rates of narcissism (Grijalva et al., 2015) due to established parenting practices and socialization processes aimed to teach males more agenting roles and females more caring and communal roles. Further, males are more likely to score high on self-esteem during adolescence, possibly because girls’ attitudes about themselves (e.g., their attractiveness) become more negative during this period (Bleidorn et al., 2016). Regarding RAgg, adolescent females seem to be more relationally aggressive (Murray-Close et al., 2016), although some findings suggest that adolescent males engage in RAgg as frequently as females or even more frequently than females (Casper et al., 2017; Kokkinos et al. 2016). Despite the mixed evidence on gender differences in RAgg, meta-analytic findings reported negligible and no gender differences regarding experiences of RAgg (Card et al., 2008; Casper et al., 2020). We thus aimed to explore the role of gender on the profiles of narcissism, self-esteem, and RAgg in adolescents. Compared to girls, we expected that boys would report an increased likelihood of belonging to a high narcissism, high self-esteem, and high RAgg profile.

Method

Participants and procedure

The Institute of Education Policy, a scientific and advisory body regulated by the Greek Ministry of Education, Research and Religious Affairs, approved the study. A random stratified cluster sampling with prefecture and locality (urban/rural) as sampling criteria was applied to select the respondents from 10 junior high schools in the broader area of Eastern Macedonia and Thrace including five regions. The sample was selected to reflect the adolescent student population (junior high school students) in this area based on the most recent Greek Bureau of Statistics (GBS) data. The Greek educational system at the secondary level includes 3 years of junior high school (grades 7–9). The students stay in self-contained classrooms and have distinct courses taught by different teachers. The average class size is approximately 20 students. However, the school size differs in terms of the school's locality (urban/rural schools). In each school all three grade classes (grades 7–9) took part, making 30 classes in total, while all students in these classrooms were eligible for inclusion.

School principals were face-to-face informed about the purpose of the study and students were instructed to hand in written consent to their parents and/or legal guardians emphasizing that the participation was voluntary and anonymous and that the collected information was confidential. Surveys were group administered during a class period for approximately 45 min. Approximately 97% (n = 2,439) of the given consent forms were returned by the parents and/or legal guardians and most of them gave active consent for their children to participate (n = 2,229, 89% of the original sample). Twenty-two students and their questionnaires were excluded because of errors or incomplete responses. In total 2,207 individuals (1,165, 52.8% females), aged between 13 and 16 (M = 14.04, SD = .81) participated in the study. A post hoc power analysis was conducted utilizing G * Power. With an alpha level of 0.05, a sample size of 2207, and a medium effect size of 0.46 (Cohen 1992), the achieved power for the study was 1.00.

Measures

Relational aggression. The 30-item Relational Aggression Scale (RAS; Voulgaridou & Kokkinos, 2018) (“I blame some kids behind their backs, so the others don’t like them”) rated on a 5-point scale (0=never to 4=all of the time) was used. The RAS was initially developed in English and was then translated into Greek according to cross-cultural research guidelines (Voulgaridou & Kokkinos, 2018). The scale was used with Greek youth samples and showed adequate psychometric properties (Voulgaridou & Kokkinos, 2018). Previous findings confirmed the validity and reliability of a unidimensional, two correlated two-factor models (i.e., direct and indirect RAgg, proactive and reactive RAgg) and a correlated four-factor model (i.e., proactive direct, proactive indirect, reactive direct, reactive indirect) (Voulgaridou & Kokkinos, 2018). In this study, the RAS uni-dimensional structure was used. Past research has shown adolescents’ responses to this scale to be highly reliable (Voulgaridou & Kokkinos, 2020), and in the present study, a high α = .94 was obtained.

Social goals. The 12-item Kiefer and Ryan's (2008) adapted version of the Social Goal Questionnaire was used to measure dominance (“when I am with people my own age I like it when I make them do what I want”), popularity (“when I am with people my own age I like it when they like me better than anyone else”), and intimacy goals (“when I am with people my own age I like it when someone understands how I feel”) on a 5-point scale (1=not at all true of me to 5=very true of me). The scale was translated and used in prior research with Greek-speaking samples and its psychometric properties were supported (Kokkinos et al., 2020). Internal consistency of dominance, popularity, and intimacy social goals in the current sample was adequate (Cronbach's alphas .92, .87, and .91, respectively).

Narcissism. The 7-item Narcissism subscale (“I act charming or nice to get things I want”) from the Antisocial Process Screening Device-Youth Version (APSD; Frick & Hare, 2001) was used, rated on a four-point Likert scale (0 = not at all true to 3 = definitely true). The scale was translated into Greek and its validity has been supported by Greek-speaking adolescents (Fanti & Henrich, 2015). Cronbach's alpha of .68 for the narcissism subscale was acceptable for this study.

Self-esteem. The brief 4-item unidimensional Lifespan Self-Esteem Scale (Harris et al., 2017) (“How do you feel about yourself?”) was used. The response options were illustrated with cartoon faces depicting the appropriate feeling (from “really sad” to “really happy”). The scale demonstrated adequate psychometric properties among adolescents (Harris et al., 2017), but it has not been previously used with Greek-speaking samples. The Greek translation of the scale was developed using the front and back translation method with the assistance of two bilingual psychologists. Then, the scale was pilot tested with a small sample bearing the same characteristics as the main study's sample for item comprehensibility. The reliability analysis showed a high Cronbach's alpha of .89.

Data analysis plan

Data analyses were performed using SPSS 27.0 (IBM) and MPLUS, Version 7.4. Preliminary analyses (means, standard deviations, Cronbach's alphas, and bivariate correlations using Pearson's r coefficient) were performed. Missing data were handled using full information maximum likelihood estimation. LPA, a person-centered approach that can identify different latent classes within a heterogeneous sample, was used to explore the hypothesized groups based on adolescents’ scores on RAgg, self-esteem, and narcissism. It is a probabilistic, model-based method that estimates posterior probabilities of class membership and classifies individuals into latent classes based on their probability of belonging to a class (Muthén & Muthén, 2012). The final class count is determined using a variety of statistical criteria, including the Bayesian information criterion (BIC) and the Lo-Mendel-Rubin (LMR) statistic (Nylund et al., 2007). Specifically, the model with the lower BIC value is chosen (Schwarz, 1978). In terms of the LMR statistic, a non-significant chi-square value (p >.05) indicates that the model with one less class should be chosen (Lo et al., 2001). Finally, average posterior probabilities and entropy values greater than .80 are preferred since they suggest good model classification and class membership (Clark & Muthén, 2009). After LPA, analysis of variance (ANOVA) was applied to explore the potential mean-differences among distinct groups, and post hoc multiple comparisons using the Tukey HSD test were run. Missing data were handled using full information maximum likelihood.

Results

Descriptive statistics

Table 1 summarizes the descriptive statistics, Cronbach's alphas, and the correlations among the variables. RAgg was significantly positively associated with narcissism and negatively with self-esteem. Positive links were observed between RAgg, dominance, and popularity, while RAgg was negatively related to intimacy. Narcissism was positively linked to all the social goals, while similar associations were found between self-esteem and social goals except for intimacy.

Table 1.
Bivariate correlations between relational aggression, narcissism, self-esteem, and social goals.

RAgg NARC SE D I P

Ragg

NARC .29**

SE −.05* .01

D .29 .29 .18

I –.08 .13 .03 .23*

P .22 .33 .12 .50* .48***

Descriptive statistics

Cronbach's alpha .94 .68 .89 .92 .87 .91

Means .91 1.02 4.13 2.36 2.79 2.54

SD .1.06 .57 .72 .79 .92 .78

Note. A factor correlation matrix was estimated using maximum likelihood with robust standard errors and full information maximum likelihood (FIML) (N = 2,207). RAgg = Relational Aggression; NARC = narcissism; SE = self-esteem; D = Dominance; I = Intimacy; P = Popularity.

p < .001; p < .01; p < .05

Latent profile analysis

To identify the optimal number of groups to retain, four separate LPA models were estimated, ranging from two to five groups. As shown in Table 2, the BIC statistic increased from Class 4 to Class 5 and decreased from Class 3 to Class 4. In addition, the LMR statistic fell out of significance for the five-class model (p = .06). Thus, the 4-class model better represented the data based on the BIC and LMR statistics. The mean posterior probability scores ranged from .82 to .97 and the entropy value was .89, suggesting that the identified classes were well separated. Figure 1 presents the four identified groups: the group with the lowest scores in RAgg, narcissism, and self-esteem (“low risk”), the group with the highest scores in RAgg, narcissism, and self-esteem (“relational aggressors with narcissistic self-esteem”), the group with moderate scores in RAgg, narcissism, and self-esteem (“moderate risk”), and the group with high RAgg and narcissism scores but low self-esteem (“relational aggressors with vulnerable self-esteem”). The number of adolescents identified in each group and the group distribution across adolescents’ gender is shown in Table 3.

Figure 1.
Latent profile analysis of relational aggression, narcissism, and self-esteem.

Table 2.
Model fit statistics based on LPA results.

Classes BIC Entropy LMR

2 11384.27 .95 p < .001

3 11084.46 .91 p < .01

4 10808.55 .89 p < .01

5 10928.95 .88 p = .06

Note. Numbers in bold indicate the selected LPA classes.

LPA = Latent Profile Analysis; BIC = Bayesian information criterion; and LMR = Lo-Mendel-Rubin.

Table 3.
Cross-tabulation between gender and groups (%).^a

Groups Total

Relational aggressors with narcissistic self-esteem Relational aggressors with vulnerable self-esteem Moderate risk Low risk

Males 47 (55.3) 96 (63.6) 168 (46.8) 731 (45.3) 1,042

Females 38 (44.7) 55 (36.4) 191 (53.2) 881 (54.7) 1,165

Total 85 (100) 151 (100) 359 (100) 1,612 (100) 2,207

^a
(χ² (3, Ν = 2,207) = 20.73, p =.000).

Group differences in adolescents’ social goals

One-way ANOVAs evaluated possible significant effects of participant roles on students’ social goals, indicating that adolescents’ scores significantly differed in terms of dominance and popularity, while non-significant findings emerged for intimacy (Table 4). Post-hoc comparisons indicated that low-risk students scored lower than the other groups in dominance and popularity. In addition, the relational aggressors with narcissistic self-esteem group had higher scores than relational aggressors with vulnerable self-esteem in dominance and higher scores than the moderate risk group in popularity. Furthermore, adolescents in the relational aggressors with vulnerable self-esteem group scored higher than the moderate risk group in both dominance and popularity.

Table 4.
Participant role differences in social goals.

Participant role

Relational aggressors with narcissistic self-esteem (n = 85) Relational aggressors with vulnerable self-esteem (n = 151) Moderate risk (n = 359) Low risk (n = 1,612) ANOVA (df = 3, 2,207)

M SD M SD M SD M SD F p

D 2.01_a .86 1.60 .75 1.13_a .69 1.02 .76 68.32 .000

P 2.02_b .88 1.76_b .80 1.39 .75 1.29 .76 38.04 .000

I 1.89 .83 1.76 .85 1.78 .94 1.81 .93 .43 .73

D = dominance; P = popularity; and I = intimacy.

Means within a row sharing the same subscripts indicate pairwise comparisons with non-significant differences at the p <.05 level (Tukey HSD test).

Discussion

This study provides unique contributions to the literature on the associations among narcissism, self-esteem, RAgg, and social goals in adolescents. The results showed that RAgg is positively associated with narcissism and negatively with self-esteem. These findings corroborate past research and emphasize the increased significance of narcissism in understanding relationally aggressive behavior (Hayes et al., 2021). Additionally, the results revealed a non-significant relationship between self-esteem and narcissism, corroborating earlier research that distinguishes these constructs (Fanti & Henrich, 2015). In terms of social goals, dominance was found to have moderately positive correlations with RAgg and narcissism, but only minor associations with self-esteem. Popularity was associated with reduced levels of RAgg, narcissism, self-esteem, and high levels of intimacy, but intimacy was associated with minor levels of narcissism.

The person-centered analysis revealed four distinct groups of participants, with two scoring high on RAgg and narcissism and the other two scoring low or moderate on these variables. Overall, 70% of the adolescents reported low levels of RAgg and narcissism. The most prevalent profile, termed “low risk,” was characterized by low RAgg and narcissism ratings, as well as average self-esteem. A single moderate risk profile with moderate levels of RAgg and narcissism, but below-average self-esteem was also detected. Two high RAgg groups with scores above one standard deviation on narcissism also emerged. The first group, labeled relational aggressors with narcissistic self-esteem, had high narcissism and self-esteem scores. Only a small number of adolescents were assigned to this group. This came as no surprise, given the positively skewed distributions of RAgg and narcissism scores. The other profile had nearly identical levels of RAgg and narcissism, but lower levels of self-esteem, thus it was labeled relational aggressor with vulnerable self-esteem. Members of this group appear to have a fragile self-image, believing that they deserve more but demonstrating a low sense of self-worth. These findings are consistent with research indicating two distinct subgroups of narcissists that have a common sense of entitlement but have significantly differing levels of self-esteem (Stronge et al., 2016).

In terms of gender, males were overrepresented in both relational aggressor groups with high narcissism. Indeed, Greek male adolescents were found to score higher on RAgg than females (Voulgaridou & Kokkinos, 2018). Females, on the other hand, were more likely to be assigned to the “moderate” and “low” risk groups, indicating gender is a factor frequently linked to student’ engagement in RAgg (Voulgaridou & Kokkinos, 2019). Indeed, research undertaken in different cultural contexts (America, Japan, China, Greece, and Italy) revealed inconsistent findings regarding gender variations in RAgg, which have been attributed to how normative RAgg is within the cultural context and how likely this form of aggressive behavior will be punished. Males were more likely to report higher rates of narcissism, confirming prior research (Grijalva et al., 2015). This evidence could be attributed to established parenting practices and socialization processes aimed to make males more agenting.

According to previous studies linking adolescent dominance or popularity to RAgg and narcissism, the “relational aggressors with narcissistic self-esteem” group scored higher on these social goals than the other groups. It appears that adolescents with a strong (high self-esteem) and pleasant (high narcissism) self-view may engage in RAgg to win others’ support and aid (Ojanen et al., 2012). Indeed, positive associations have been found amongst relationally aggressive behavior, social efficacy, and cooperative conduct (Puckett et al., 2008), indicating that RAgg might be associated with adequate social skills. Although the “relational aggressors with low self-esteem” group scored much lower on these social goals than the “moderate” and “low” risk groups, adolescents with both high self-esteem and narcissism are motivated to seek social dominance and popularity. The findings suggest that self-esteem and affiliative behaviors may account for at least some of the attempts made by youth to acquire social support from their peer group.

Notably, the differences in intimacy between the groups did not reach statistical significance. The combination of RAgg, narcissism, and self-esteem seems to have no effect on adolescents’ orientation toward intimate goals in peer interaction. While aggression among children is negatively linked to goals related to peer interaction and closeness, it may be that adolescent aggression, and especially its relational form mirrors mainly goals associated with social dominance rather than low intimacy with peers (Ojanen et al., 2012). This could be indicative of a somewhat normative function of social dominance and aggressive behavior in adolescents during times of RAgg peaking and being used to acquire social approval (Murray-Close et al., 2016).

The current study makes a significant contribution to the existing literature by providing more insight into adolescents’ personality traits (i.e., narcissism), self-esteem, and RAgg as well as their associations with their social goals. The present results shed light on the combination of narcissism and self-esteem that put adolescents at increased risk for RAgg. The finding that higher narcissism is associated with higher RAgg points to the need for interventions focusing on decreasing grandiosity. The implications of this study will help counselors to plan treatment interventions to reduce relationally aggressive behaviors of narcissistic adolescents.

Results indicated that adolescents who score high on RAgg, narcissism, and self-esteem were more likely to report higher levels of social dominance and popularity, while narcissistic but low self-esteemed adolescents were less likely to report these social goals. These findings have significant implications for both developmental and school psychologists, as they highlight the need for the implementation of differentiated intervention strategies that should not solely rely on decreasing adolescents’ narcissism. For instance, preventive policies should aim to replace excessive and unconditional praise with more realistic sources of self-esteem, as well as guide youngsters on how to deal with negative feedback adaptively (Barry et al., 2003). These findings underscore the importance of developing effective interventions that promote positive self-image and teach adolescents various strategies for coping with ego threats or worry about their social placement within their peer group (Derry et al., 2020). For example, classroom-based prevention and intervention programs should include supervised activity in school sessions that allow teenagers to channel their aggressive behavior productively and develop their self-esteem in a friendly competitive atmosphere (Campbell et al., 2010). Additionally, interventions should not overlook adolescent’ social goals, in that other means of achieving them should be made available to students by their caregivers, such as participation in extracurricular activities that can divert aggressors away from relationally aggressive behavior. A challenge for intervention strategies in terms of narcissism and self-esteem would be for parents, educators, and stakeholders to increase global self-worth without simultaneously inflating these perceptions by providing young people with encouragement rooted in worthiness as an individual rather than in comparison to others (Brummelman et al., 2016). Although the social comparison is unavoidable, it should only be used to gain a true awareness of one's own strengths and limitations. Intrinsic reflection should be promoted as a means of personal development of one's abilities, talents, and values, which can potentially contribute to good psychosocial adjustment (Farrell & Vaillancourt, 2021).

Several limitations of this study should be noted. First, the data were collected only by self-report, which may have resulted in underreporting of unwanted relationally aggressive conduct, while the shared method variance may have inflated the obtained correlations (Fanti & Henrich, 2015). Self-report measures, on the other hand, have the advantage of obscuring the individual's opinions and emotions from other observers. Additionally, self-report measures of psychopathology and personality appear to be more valid during adolescence, but parent and teacher report measures appear to lose validity during adolescence (Fanti & Henrich, 2015). Nonetheless, future research might employ a multi-informant approach and focus on adolescents of multiple age groups. Second, as this study used measures that did not differentiate RAgg's functions or narcissism dimensions, future research could investigate the reactive and proactive function of RAgg and the adaptive and maladaptive dimensions of narcissism and their potentially different associations.

Overall, this evidence supports that distinct profiles of RAgg, narcissism, and self-esteem profiles have been identified in a sample of adolescents. Also, this study explored whether these profiles differentiate adolescents’ scores on social goals of dominance, popularity, and intimacy, indicating that relational aggressors with narcissistic self-esteem (high narcissism and self-esteem) scored higher on social goals than relational aggressors with vulnerable self-esteem (high narcissism but low self-esteem). These findings pave the way for future research using person-centered analysis to investigate such associations across development.

	RAgg	NARC	SE	D	I	P
Ragg
NARC	.29**
SE	−.05*	.01
D	.29**	.29**	.18**
I	–.08**	.13**	.03	.23***
P	.22**	.33**	.12**	.50***	.48***
Descriptive statistics
Cronbach's alpha	.94	.68	.89	.92	.87	.91
Means	.91	1.02	4.13	2.36	2.79	2.54
SD	.1.06	.57	.72	.79	.92	.78

Classes	BIC	Entropy	LMR
2	11384.27	.95	p < .001
3	11084.46	.91	p < .01
4	10808.55	.89	p < .01
5	10928.95	.88	p = .06

	Groups	Total
Males	47 (55.3)	96 (63.6)	168 (46.8)	731 (45.3)	1,042
Females	38 (44.7)	55 (36.4)	191 (53.2)	881 (54.7)	1,165
Total	85 (100)	151 (100)	359 (100)	1,612 (100)	2,207

	Participant role
D	2.01_a	.86	1.60	.75	1.13_a	.69	1.02	.76	68.32	.000
P	2.02_b	.88	1.76_b	.80	1.39	.75	1.29	.76	38.04	.000
I	1.89	.83	1.76	.85	1.78	.94	1.81	.93	.43	.73

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Greek State Scholarships Foundation (grant no. 10451).

ORCID iD

Constantinos M. Kokkinos

Author biographies

Ioanna Voulgaridou, BEd, MEd, MA, PhD is a contract assistant professor at the Department of Education Sciences in Early Childhood, Democritus University of Thrace, Greece (Educational Psychology). Her research focuses on adolescents’ relationally aggressive behavior, and in antisocial behavior and its intrapersonal and interpersonal correlates.

Constantinos M. Kokkinos, PhD, is a professor of Educational Psychology at the Department of Primary Education, Democritus University of Thrace, Greece. He studies the role of individual differences and contextual correlates of children’s, youth and emerging adults’ aggression.

Kostas Fanti, PhD, is an associate professor of Developmental Psychopathology at the Department of Psychology, University of Cyprus. His research focus on how individual, environmental, and biological factors influence maladaptive behavior and especially antisocial behavior from preschool to adolescence. His current research focuses on novel cognitive neuroscience informed treatments for children and adolescents at risk for conduct disorder.

References

Anderson

C. A.

Bushman

B. J.

(2002). Human aggression. Annual Review of Psychology, 53(1), 27–51. https://doi.org/10.1146/annurev.psych.53.100901.135231

Banny

A. M.

Heilbron

Ames

Prinstein

M. J

. (2011). Relational benefits of relational aggression: Adaptive and maladaptive associations with adolescent friendship quality. Developmental Psychology, 47(4), 1153–1166. https://doi.org/10.1037/a0022546

Barry

C. T.

Frick

P. J.

Killian

A. L.

(2003). The relation of narcissism and self-esteem to conduct problems in children: A preliminary investigation. Journal of Clinical Child and Adolescent Psychology, 32(1), 139–152. https://doi.org/10.1207/S15374424JCCP3201_13

Bell

G. R.

Crothers

L. M.

Hughes

T. L.

Kanyongo

G. Y.

Kolbert

J. B.

Parys

(2018). Callous-unemotional traits, relational and social aggression, and interpersonal maturity in a sample of behaviorally disordered adolescents. Journal of Applied School Psychology, 34(6), 65–85. https://doi.org/10.1080/15377903.2017.1345814

Bleidorn

Arslan

R. C.

Denissen

J. J.

Rentfrow

P. J.

Gebauer

J. E.

Potter

Gosling

S. D.

(2016). Age and gender differences in self-esteem—A cross-cultural window. Journal of Personality and Social Psychology, 111(3), 396–410. https://doi.org/10.1037/pspp0000078

Brummelman

Thomaes

Sedikides

(2016). Separating narcissism from self-esteem. Current Directions in Psychological Science, 25(1), 8–13. https://doi.org/10.1177/0963721415619737

Buhrmester

. (1998). Need fulfillment, interpersonal competence, and the developmental contexts of early adolescent friendship. In Bukowski

W. M.

Newcomb

A. F.

Hartup

W. W.

(Eds.), The company they keep: Friendship in childhood and adolescence (pp. 158–185). Cambridge University Press.

Bushman

B. J.

Baumeister

R. F.

Thomaes

Begeer

West

S. G.

(2009). Looking again, and harder, for a link between low self-esteem and aggression. Journal of Personality, 77(2), 427–446. https://doi.org/10.1111/j.1467-6494.2008.00553.x

Campbell

R. L.

Eisner

Riggs

. (2010). Sources of self-esteem: From theory to measurement and back again. New Ideas in Psychology, 28(3), 338–349. https://doi.org/10.1016/j.newideapsych.2009.09.008

10.

Card

N. A.

Stucky

B. D.

Sawalani

G. M.

Little

T. D.

(2008). Direct and indirect aggression during childhood and adolescence: A meta-analytic review of gender differences, intercorrelations, and relations to maladjustment. Child Development, 79(5), 1185–1229. https://doi.org/10.1111/j.1467-8624.2008.01184.x

11.

Casper

D. M.

Card

N. A.

Barlow

(2020). Relational aggression and victimization during adolescence: A meta-analytic review of unique associations with popularity, peer acceptance, rejection, and friendship characteristics. Journal of Adolescence, 80, 41–52. https://doi.org/10.1016/j.adolescence.2019.12.012

12.

Casper

D. M.

Card

N. A.

Bauman

Toomey

R. B.

(2017). Overt and relational aggression participant role behavior: Measurement and relations with psychometric status and depression. Journal of Research on Adolescence, 27(3), 661–673. https://doi.org/10.1111/jora.12306

13.

Clark

S. L.

Muthén

(2009). Relating latent class analysis results to variables not included in the analysis. Retrieved from http://www.statmodel.com/download/relatinglca.pdf.

14.

Cohen

. (1992). Statistical power analysis. Current Directions in Psychological Science, 1(3), 98–101. 10.1111/1467-8721.ep10768783

15.

Coyne

S. M.

Archer

Eslea

(2006). “We're not friends anymore! unless …”: The frequency and harmfulness of indirect, relational, and social aggression. Aggressive Behavior: Official Journal of the International Society for Research on Aggression, 32(4), 294–307. https://doi.org/10.1002/ab.20126

16.

Derry

K. L.

Ohan

J. L.

Bayliss

D. M

. (2020). Fearing failure: Grandiose narcissism, vulnerable narcissism, and emotional reactivity in children. Child Development, 91(3), e581–e596. https://doi.org/10.1111/cdev.13264

17.

Fanti

K. A.

Henrich

C. C.

(2015). Effects of self-esteem and narcissism on bullying and victimization during early adolescence. The Journal of Early Adolescence, 35(1), 5–29. https://doi.org/10.1177/0272431613519498

18.

Faris

Felmlee

(2011). Status struggles: Network centrality and gender segregation in same-and cross-gender aggression. American Sociological Review, 76(1), 48–73. https://doi.org/10.1177/0003122410396196

19.

Farrell

A. H.

Vaillancourt

(2021). Joint trajectories of adolescent narcissism and self-esteem predict interpersonal features in young adulthood. Journal of Research in Personality, 93, 104127. https://doi.org/10.1016/j.jrp.2021.104127

20.

Frick

P. J.

Hare

R. D.

(2001). The Antisocial Process Screening Device. Multi-Health System.

21.

Grijalva

Newman

D. A.

Tay

Donnellan

M. B.

Harms

P. D.

Robins

R. W.

Yan

(2015). Gender differences in narcissism: A meta-analytic review. Psychological Bulletin, 141(2), 261–310. https://doi.org/10.1037/a0038231

22.

Harris

M. A.

Donnellan

M. B.

Guo

McAdams

D. P.

Garnier-Villarreal

Trzesniewski

K. H.

(2017). Parental co-construction of 5-to 13-year-olds’ global self-esteem through reminiscing about past events. Child Development, 88(6), 1810–1822. https://doi.org/10.1111/cdev.12944

23.

Hayes

N. L.

Marsee

M. A.

Russell

D. W.

(2021). Latent profile analysis of traditional and cyber-aggression and victimization: Associations with dark triad traits and psychopathology symptoms. Journal of Psychopathology and Behavioral Assessment, 43, 399–412. https://doi.org/10.1007/s10862-020-09835-2

24.

Horvath

Morf

C. C.

(2010). To be grandiose or not to be worthless: Different routes to self-enhancement for narcissism and self-esteem. Journal of Research in Personality, 44(5), 585–592. https://doi.org/10.1016/j.jrp.2010.07.002

25.

Kerig

P. K

. (2018). Polyvictimization and girls’ involvement in the juvenile justice system: Investigating gender-differentiated patterns of risk, recidivism, and resilience. Journal of Interpersonal Violence, 33(5), 789–809. https://doi.org/10.1177/0886260517744843

26.

Kiefer

S. M.

Ryan

A. M.

(2008). Striving for social dominance over peers: The implications for academic adjustment during early adolescence. Journal of Educational Psychology, 100(2), 417–428. https://doi.org/10.1037/0022-0663.100.2.417

27.

Kokkinos

C. M.

Karagianni

Voulgaridou

. (2017). Relational aggression, big five and hostile attribution bias in adolescents. Journal of Applied Developmental Psychology, 52, 101–113. https://doi.org/10.1016/j.appdev.2017.07.007

28.

Kokkinos

C. M.

Kountouraki

Voulgaridou

Markos

(2020). Understanding the association between big five and relational aggression: The mediating role of social goals and friendship jealousy. Personality and Individual Differences, 160, Article e109946. https://doi.org/10.1016/j.paid.2020.109946

29.

Kokkinos

C. M.

Voulgaridou

. (2017). Links between relational aggression, parenting and personality among adolescents. European Journal of Developmental Psychology, 14(3), 249–264. https://doi.org/10.1080/17405629.2016.1194265

30.

Kokkinos

C. M.

Voulgaridou

Markos

(2016). Personality and relational aggression: Moral disengagement and friendship quality as mediators. Personality and Individual Differences, 95, 74–79. https://doi.org/10.1016/j.paid.2016.02.028

31.

Laursen

Hoff

. (2006). Person-centered and variable-centered approaches to longitudinal data. Merrill-Palmer Quarterly, 52(3) 377–389.

32.

Leary

M. R

. (2004). The self we know and the self we show: Self-esteem, self-presentation, and the maintenance of interpersonal relationships. In Brewer

M. B.

Hewstone

(Eds.), Emotion and motivation (pp. 204–224). Blackwell Publishing.

33.

Leff

S. S.

Waasdorp

T. E.

Crick

N. R

. (2010). A review of existing relational aggression programs: Strengths, limitations, and future directions. School Psychology Review, 39(4), 508–535. 10.1080/02796015.2010.12087739

34.

Mendell

N. R.

Rubin

D. B.

(2001). Testing the number of components in a normal mixture. Biometrika, 88(3), 767–778. https://doi.org/10.1093/biomet/88.3.767

35.

Mayeux

(2014). Understanding popularity and relational aggression in adolescence: The role of social dominance orientation. Social Development, 23(3), 502–517. https://doi.org/10.1111/sode.12054

36.

McEachern

A. D.

Snyder

(2012). Gender differences in predicting antisocial behaviors: Developmental consequences of physical and relational aggression. Journal of Abnormal Child Psychology, 40, 501–512. https://doi.org/10.1007/s10802-011-9589-0

37.

Morf

C. C.

Rhodewalt

(2001). Unraveling the paradoxes of narcissism: A dynamic self-regulatory processing model. Psychological Inquiry, 12(4), 177–196. https://doi.org/10.1207/S15327965PLI1204_1

38.

Murray-Close

Nelson

D. A.

Ostrov

J. M.

Casas

J. F.

Crick

N. R.

(2016). Relational aggression: A developmental psychopathology perspective. In Cicchetti

(Ed.), Developmental psychopathology (pp. 660–722). Wiley.

39.

Muthén

L. K.

Muthén

B. O.

(2012). Mplus user’s guide. Muthén & Muthén.

40.

Nylund

K. L.

Asparouhov

Muthén

B. O.

(2007). Deciding on the number of classes in latent class analysis and growth mixture modeling: A Monte Carlo simulation study. Structural Equation Modeling: A Multidisciplinary Journal, 14(4), 535–569. https://doi.org/10.1080/10705510701575396

41.

Ojanen

Findley

Fuller

(2012). Physical and relational aggression in early adolescence: Associations with narcissism, temperament, and social goals. Aggressive Behavior, 38(2), 99–107. https://doi.org/10.1002/ab.21413

42.

Onishi

Kawabata

Kurokawa

Yoshida

(2012). A mediating model of relational aggression, narcissistic orientations, guilt feelings, and perceived classroom norms. School Psychology International, 33(4), 367–390. https://doi.org/10.1177/0143034311421433

43.

Park

L. E.

Maner

J. K

. (2009). Does self-threat promote social connection? The role of self-esteem and contingencies of self-worth. Journal of Personality and Social Psychology, 96(1), 203–217. https://doi.org/10.1037/a0013933

44.

Puckett

M. B.

Aikins

J. W.

Cillessen

A. H. N.

(2008). Moderators of the association between relational aggression and perceived popularity. Aggressive Behavior, 34(6), 563–576. https://doi.org/10.1002/ab.20280

45.

Rigby

(2004). Addressing bullying in schools theoretical perspectives and their implications. School Psychology International, 25(3), 287–300. https://doi.org/10.1177/0143034304046902

46.

Schwarz

(1978). Estimating the dimension of a model. The Annals of Statistics, 6(2), 461–464.

47.

Stronge

Cichocka

Sibley

C. G.

(2016). Narcissistic self-esteem or optimal self-esteem? A latent profile analysis of self-esteem and psychological entitlement. Journal of Research in Personality, 63, 102–110. https://doi.org/10.1016/j.jrp.2016.06.016

48.

Vaillancourt

Krems

J. A.

(2018). An evolutionary psychological perspective of indirect aggression in girls and women. In Coyne

Ostrov

(Eds.), The development of relational aggression (pp. 111–126). Oxford University Press.

49.

van den Broek

Deutz

M. H.

Schoneveld

E. A.

Burk

W. J.

Cillessen

A. H.

(2016). Behavioral correlates of prioritizing popularity in adolescence. Journal of Youth and Adolescence, 45, 2444–2454. https://doi.org/10.1007/s10964-015-0352-7

50.

Voulgaridou

Kokkinos

C. M.

(2015). Relational aggression in adolescents: A review of theoretical and empirical research. Aggression and Violent Behavior, 23, 87–97. https://doi.org/10.1016/j.avb.2015.05.006

51.

Voulgaridou

Kokkinos

C. M.

(2018). The relational aggression scale (RAS): Psychometric properties of a newly developed measure of relational aggression. European Review of Applied Psychology, 68(1), 11–22. https://doi.org/10.1016/j.erap.2017.12.001

52.

Voulgaridou

Kokkinos

C. M.

(2019). Measuring relational aggression in children and adolescents: A systematic review of the available instruments. Aggression and Violent Behavior, 46, 82–97. https://doi.org/10.1016/j.avb.2019.02.002

53.

Voulgaridou

Kokkinos

C. M

. (2020). The mediating role of friendship jealousy and anxiety in the association between parental attachment and adolescents’ relational aggression: A short-term longitudinal cross-lagged analysis. Child Abuse & Neglect, 109, 104717. https://doi.org/10.1016/j.chiabu.2020.104717

54.

Yeager

D. S.

Dahl

R. E.

Dweck

C. S.

(2018). Why interventions to influence adolescent behavior often fail but could succeed. Perspectives on Psychological Science, 13(1), 101–122. https://doi.org/10.1177/1745691617722620

	Groups				Total
	Relational aggressors with narcissistic self-esteem	Relational aggressors with vulnerable self-esteem	Moderate risk	Low risk
Males	47 (55.3)	96 (63.6)	168 (46.8)	731 (45.3)	1,042
Females	38 (44.7)	55 (36.4)	191 (53.2)	881 (54.7)	1,165
Total	85 (100)	151 (100)	359 (100)	1,612 (100)	2,207

	Participant role
	Relational aggressors with narcissistic self-esteem (n = 85)		Relational aggressors with vulnerable self-esteem (n = 151)		Moderate risk (n = 359)		Low risk (n = 1,612)		ANOVA (df = 3, 2,207)
	M	SD	M	SD	M	SD	M	SD	F	p
D	2.01_a	.86	1.60	.75	1.13_a	.69	1.02	.76	68.32	.000
P	2.02_b	.88	1.76_b	.80	1.39	.75	1.29	.76	38.04	.000
I	1.89	.83	1.76	.85	1.78	.94	1.81	.93	.43	.73