Is Aggression Associated With Biased Perceptions of One’s Acceptance and Rejection in Adolescence?

Abstract

In adolescence, being rejected by one’s peers is positively associated with aggression. However, whether self-perceptions of being rejected or accepted by peers, and biases in these perceptions, are linked to aggression remains unclear, as the literature points to 2 perspectives: Youth are more likely to be aggressive when they (a) know or believe that they are rejected or (b) overestimate their acceptance. By addressing some of the limitations of prior studies, this study aimed to clarify how self-perceptions of status are related to concurrent and future aggression, and whether high levels of aggression are predictive of biased self-perceptions of acceptance and rejection. Data were collected in 2 consecutive school years from 572 high school students (M_age = 15.06, SD = .75; 55.4% girls). The ethnic/racial composition of the sample was 47.5% Asian, 43.1% Latino/Hispanic, 4.0% White, and 5.3% other. For well-liked youth, awareness of one’s acceptance was positively associated with concurrent relational aggression, whereas for highly rejected youth, awareness of one’s rejection was positively associated with concurrent overt aggression. Awareness of being rejected (i.e., high levels of both self-perceived and actual rejection) was associated with elevated levels of aggression over time. There was no evidence that youth with high levels of aggression had more biased perceptions of their status (concurrently or longitudinally) than youth with low levels of aggression. These findings help clarify how youth’s status-related perceptions relate to the development of aggression.

Research on predictors of aggression in childhood and adolescence has revealed that peer status plays a significant role in the initiation or maintenance of such behavior (see Leary et al., 2006). Acceptance amongst peers (i.e., being liked; also referred to as social preference when assessed by subtracting rejection nominations from acceptance nominations) tends to be negatively associated with aggression (e.g., Tomada & Schneider, 1997; see Casper et al., 2020). Being rejected—or disliked—leads to increases in aggression over time (e.g., Cillessen & Mayeux, 2004; Dodge et al., 2003; Lansford et al., 2010; Miller-Johnson et al., 2002; Ojanen & Findley-Van Nostrand, 2014). Given the deleterious effects of peer aggression on victims’ mental (Christina et al., 2021; Schwartz et al., 2015) and physical health (Bogartet al., 2014; Gini et al., 2014), it is essential to effectively reduce perpetration. This requires a comprehensive understanding of youth’s social cognitions that may motivate their aggression.

Several theories suggest that the extent to which individuals think that they are liked or disliked by others (i.e., self-perceived acceptance and rejection), as well the degree to which their perceptions are correct (i.e., congruent with their actual status; Bellmore & Cillessen, 2003; Kenny, 1994), may be key determinants of aggressive behavior in children and youth (e.g., Leary et al., 2006). However, the extant literature points to two perspectives for the link between self-perceived status and aggression—youth are more likely to be aggressive when: 1. they know or believe that they are rejected, or 2. they overestimate their peer acceptance. These perspectives are not mutually exclusive: Some youth may know they are not well-liked but still overestimate the degree to which they are liked, and some might overestimate both their acceptance and rejection (e.g., if they think they are liked by some and disliked by others; Coie et al., 1982). Due to inconsistent findings as well as methodological limitations (e.g., not differentiating between acceptance and rejection) in the literature, it is not yet clear whether or not there is more support for one of these perspectives. The unique effects of self-perceived status on aggression beyond actual status are also not well-established. Moreover, the dearth of longitudinal studies prohibits us from understanding the direction of effects. Are self-perceptions or biases in self-perceptions of status leading to more aggression, or does aggression predict changes in self-perceptions of status over time?

In addition to these limitations, most studies have focused on childhood rather than adolescence. As the ability to evaluate one’s status is related to cognitive skills (e.g., Malloy et al., 2007), biases in adolescence are more likely to be due to personal characteristics other than individual differences in cognitive development. At the same time, there is some evidence that youth are more likely to demonstrate self-protective positive biases in the social domain in adolescence than in childhood, as social interactions and status become more salient in adolescence (Hoza et al., 2010).

Therefore, although the links between aggression and self-perceptions of status have been a focus of peer relations’ research for decades, the question of how self-perceived status, biases in perceptions of one’s status, and aggression are related remains unclear, particularly in adolescence. Taking into account possible gender differences and different forms of aggression (Mayeux & Cillessen, 2008), the present study examined the associations of two types of self-perceived status (acceptance and rejection) with concurrent and future aggression in adolescence. Furthermore, we investigated whether biases in self-perceptions of status are concurrently and longitudinally associated with aggression in two ways. First, we examined whether concordance between self-perceived and actual status is associated with, and predictive of, aggression (i.e., whether biases or accuracy potentially contribute to the development of aggression). Second, we examined whether aggression moderates the association between actual status and self-perceived status (i.e., whether high levels of aggression potentially contribute to the development of more biased or more accurate self-perceptions).

Is Self-Perceived Status Associated With Aggression?

Interpersonal rejection - being disliked - being disliked - often plays a role in aggression (e.g., Lansford et al., 2010), and most theoretical explanations for this effect imply that it is self-perceived rejection that triggers aggressive responses, via several mechanisms (Leary et al., 2006). Youth who feel rejected by peers may experience negative emotions (e.g., loneliness, anger), and their pain or frustration may result in aggression (e.g., “frustrationaggression hypothesis”; Dollard et al., 1939). Furthermore, negative emotions elicited by (perceived) rejection can lead to biased social information processing (SIP) which in turn is associated with subsequent aggression (e.g., Dodge et al., 2003; Lansford et al., 2010). Whereas some proposed explanations for a link between self-perceived rejection and aggression focus on automatic reactions to the perceived rejection (e.g., frustration), the link may also be explained by interpersonal dynamics. Self-perceptions of rejection may lead to aggression as a way for people to reestablish self-efficacy, enact revenge for the perceived slight, or demonstrate that they should be respected (Leary et al., 2006). In these circumstances, youth who perceive themselves to be disliked may be aggressive in order to “save face” or “even the score.” In addition, some of them may be especially sensitive to rejection (e. g., “confirmed inadequacy”; Downey et al., 1998; Sandstrom & Herlan, 2007). That is, some youth may be primed or expect to receive negative feedback from peers, and if they perceive cues confirming their negative self-perceptions, they respond aggressively. Consistent with this perspective, hostile attribution bias (i. e., interpreting ambiguous actions of others as intentionally hostile; Dodge et al., 2003) is also associated with elevated aggression (Verhoef et al., 2019). Surprisingly, despite strong theoretical arguments for why self-perceived rejection may lead to aggression, few studies have found an association between the two variables. Self-perceived rejection was not found to be positively correlated with physical (Guerra et al., 2004) or general aggression (García-Bacete et al., 2019) in young children, nor with peerreported relational aggression (which consists in causing harm by damaging someone’s relationships or status) or with self-reported overt aggression in early adolescents (Zimmer-Gembeck et al., 2013).

The question of how self-perceived acceptance may relate to aggression is also open. It is of course conceivable that not feeling well-liked could promote aggression for the same reasons that feeling rejected could make one aggressive. However, other processes are also possible: Whereas perceiving oneself as highly rejected may reflect a perception of hostility from others, seeing oneself as low in acceptance could simply reflect feeling neglected (i.e., neither liked nor disliked) or isolated, which is not typically associated with aggression (e.g., Coie & Kupersmidt, 1983). Alternately, perceiving oneself as well-liked could facilitate youth’s aggressive behavior by instilling a sense of entitlement; they may assume that they can afford to be aggressive because their social position is secure. The literature on the links between self-perceived acceptance and aggression appears to support these two possibilities. Some studies found no association between the two variables in children samples (Stephens et al., 2015; White & Kistner, 2011). In adolescence, the degree to which participants thought they were liked was found to be unrelated to bullying (Garandeau & Lansu, 2019) and self-perceived social preference was unrelated to relational or overt aggression in 9th- and 10th-graders (Mayeux & Cillessen, 2008). Consistent with the “sense of entitlement” perspective, self-perceived social preference was found to be positively associated with bullying among early adolescents (Pozzoli & Gini, 2021) and with overt aggression in 11th -and 12th-graders (Cillessen & Mayeux, 2004).

The current study examined the extent to which perceptions of one’s rejection and acceptance are concurrently and longitudinally associated with aggression. Importantly, the absence of consistent findings in past studies could indicate that the effects of self-perceived status on aggression depend on actual status. For example, feeling rejected may be predictive of aggression only for youth who are actually disliked and therefore experience the rejecting contexts that can elicit aggression. Thus, it is important to consider actual status and to examine the potential role of biases in self-perceptions of status in the development of aggression.

Self-Perceived Status, Actual Status, and Aggression

The current study examined whether biases in self-perceived status are associated with aggression in two ways. First, we examined how the concordance between self-perceived and actual status relate to aggression, which indicates whether biases are associated with aggression. Second, we examined whether the degree of accuracy in the perception of one’s status differs for youth high in aggression and for their less aggressive peers.

Are Biases in Self-Perceived Status Associated With Aggression?

Past research has posited that aggressive children and adolescents are characterized by a propensity for self-enhancement, that is by tendencies to overestimate their level of acceptance and underestimate their level of rejection. Such biases in one’s perceptions of status have been theorized to facilitate aggression through various processes. As affection is a fundamental human need (e.g., “the need to belong”; Baumeister & Leary, 1995) and interpersonal rejection is known to be particularly distressing (see Leary, 2015), it is possible that a lack of awareness of being rejected (i.e., underestimating one’s rejection) prevents aggressive youth from trying to win back others’ affection by ceasing their negative behavior. Knowing they are rejected could encourage youth to act in ways that will promote liking from others- that is, to refrain from aggression and engage in respectful behavior (see Coie, 1990). If these youths do not recognize that they are rejected, they may not feel a need to modify their behavior, as they do not perceive any status costs to their aggression.

Moreover, consistent with the threatened egotism hypothesis, overestimating one’s acceptance may be linked to higher aggression via narcissism, which refers to a tendency for inflated self-views in general (Bushman & Baumeister, 1998). Youth who (incorrectly) perceive themselves very positively (e.g., well-liked) may lash out when these positive self-views are threatened. In addition, aggressive youth do tend to be perceived as popular (de Bruyn et al., 2010), an indicator of visibility and prestige, which could lead them to confuse this positive attention for peer acceptance. In this case, positive biases in perceived acceptance would be positively associated with aggression.

In support of this theory, several studies with children (aged 9–11, approximately) have found a positive, concurrent association between overestimation of peer acceptance and aggression, using both standardized residual scores between self- and peer-ratings of social acceptance (e.g., David & Kistner, 2000; Orobio de Castro et al., 2007; Stephens et al., 2015; Zimmer-Gembeck et al., 2013) and difference scores (e.g., Sandstrom & Herlan, 2007; Stephens et al., 2016; Zakriski & Coie, 1996). With the same age group, Morrow and colleagues (2016) found similar results using dyadic nominations (e.g., “who do you really like?” and “who really likes you?”) to generate accuracy scores, such that acceptance bias (calculated as the number of “false positives” minus “misses”) was positively associated with aggression. There is also some evidence that overestimation of peer acceptance is prospectively associated with aggression. Using difference scores, a study by Lynch et al. (2016) found that positively biased self-perceptions of acceptance were associated with increases in relational aggression. Using standardized residual scores, Brendgen and colleagues (2004) found that overestimation of acceptance was associated with an increase in aggression, particularly for children who were initially high or average in aggression.

However, findings on the effects of biased self-perceptions of status on aggression are not completely consistent. Among adolescents, some studies found no evidence that proactive aggression was associated with biased self-perceptions of being liked (Garandeau & Lansu, 2019; White & Kistner, 2011) and a study by Mayeux and Cillessen (2008) found that those who knew they were being rejected showed increases in overt aggression, while those who overestimated their acceptance showed decreases in overt aggression. In the study by Brendgen et al. (2004) conducted with 9- to 13-year-olds, evidence was found for a curvilinear association, such that extreme underestimation of acceptance was also associated with an increase in aggression over time, though only for children initially high in aggression. In young children, underestimating one’s rejection has been linked with being less aggressive (García-Bacete et al., 2019).

Does the Association Between Actual and Perceived Status Depend on Aggression?

Whereas biases in self-perceived status may predict aggression, aggression may also influence the extent to which actual status and self-perceived status are related concurrently and over time. Social information processing theory suggests that aggressive youth differ from less aggressive youth in how they interpret social cues, such as the intentions of others in ambiguous social situations (e.g., Dodge et al., 2003). Thus, youth who are high in aggression may differ from those low in aggression in how they use cues from the social context to derive their perceptions of their own status. However, this has, to our knowledge, never been tested. Therefore, to clarify whether biases in self-perceived status depend on adolescents’ level of aggression, the current study also examined whether aggression moderates the concurrent and longitudinal association between actual and self-perceived status.

Methodological Considerations

The literature on biases in self-perceived status has raised several methodological issues. First, most studies did not distinguish between biases in peer acceptance (i.e., being liked) and biases in rejection (i.e., being disliked), even though there is evidence that they differ (Bellmore & Cillessen, 2003). Having low levels of peer acceptance is not synonymous with being rejected, and acceptance and rejection are not opposite ends of one continuum (Bukowski et al., 2000; Gorman et al., 2011). Moreover, being rejected/disliked by peers is more strongly related to aggression than not being well-liked (e.g., Gorman et al., 2011). Therefore, youth’s rejection, rather than their acceptance, may be more pertinent for aggression.

Second, the operationalization of bias is problematic in some studies, which have used standardized residual scores to account for the portion of variance in self-perceived status that is not accounted for by actual status in a regression model (e.g., Brendgen et al., 2004; David & Kistner, 2000; Lynch et al., 2016). As pointed out by De Los Reyes and Kazdin (2004), when the correlation between two variables is low, the variance accounted for in the dependent variable (e.g., self-perceived status) by the independent variable (e.g., actual status) is also low, and the residual score (i.e., difference between actual and predicted values on the dependent variable) will be almost completely predicted by the participants’ scores on the dependent variable; that is, the standardized residual score will be highly correlated with the dependent variable but not the independent variable. Consequently, any unique characteristic or contribution of actual status (or self-perceived status if it is the independent variable) would not be detected. Although difference scores may have some advantages for addressing certain questions (e.g., Owens et al., 2007), they can also be problematic when different measures are used for self- and peer-reports of status (White & Kistner, 2011). Moreover, the resulting difference score is completely predictable from the two variables (i.e., the difference score does not represent a new construct; see Zuckerman et al., 2002).

As both difference scores and standardized residual scores have their limitations, a simple alternative to examine the correlates of a discrepancy between self-perceived and actual status is by testing interactions (Zuckerman et al., 2002). If the main question of interest is how biases in perceived status affect a third variable (e.g., aggression), testing the interaction between self-perceived status and actual status in predicting this third variable will be appropriate (Zuckerman et al., 2002). Alternately, if the goal is to see whether awareness of how (dis)liked one is depends on aggression, this can be achieved by examining the moderating effect of aggression on the association between actual and self-perceived status.

The Current Study

Despite years of research on adolescents’ self-perceived status and aggression, a lingering question remains as to how self-perceived rejection and acceptance, as well as biases in self-perceived status, relate to aggression, both concurrently and over time. The current study addresses three main research questions: 1) Is self-perceived status concurrently and longitudinally associated with aggression? 2) Are biases in self-perceptions of status concurrently and longitudinally associated with aggression? 3) Does aggression moderate the extent to which actual status is concurrently or longitudinally associated with perceived status?

In an extension of past research, and to provide further understanding of the unique contributions of actual and self-perceived status to aggression, we examined whether the pattern of findings differed when considering actual and self-perceived acceptance versus actual and self-perceived rejection. Given past research demonstrating different patterns for relational and overt aggression (e.g., Mayeux & Cillessen, 2008), we conducted our analyses separately for relational and overt aggression. In addition, we explored gender as a possible moderator as there is some evidence of gender differences in the association between aggression and self-perceived status; however, we did not have any a priori hypotheses because prior findings have not been consistent (e.g., Mayeux & Cillessen, 2008; Portillo & Fernández-Baena, 2020).

Method

Participants and Procedure

At the initial assessment (T1), 9th and 10th graders from the participating Los Angeles high school were invited to participate in the study (N = 1151). Participants included 735 students (56% girls; M_age = 15.10, SD = .75: 50.2% 9th graders) who received parental consent and assented to be part of the project (63.9% participation rate). Students were followed one year later, when they were in 10th and 11th grade. Of the 735 participants, 572 participated at both time points. Retained students did not differ from attrited students on any of the T1 study variables. The self-reported ethnic/racial composition of the participants was 47.5% Asian, 43.1% Latino/Hispanic, 4.0% White, and 5.3% other or mixed ethnicity. The composition of this sample is highly representative of the larger school district in Southern California. The majority of students in this school qualified for free or reduced-price lunch (EdData, 2019). Among the participating students, 508 students had full data at T1 and T2. Little’s (1988) MCAR test indicated that the missing values were completely missing at random, χ² = 9.14, df = 20, p = .98. To address missingness, multiple imputation was conducted (5 imputations; Little & Rhemtulla, 2013) using the mice package in R (van Buuren & Groothuis-Oudshoorn, 2011).

Similar data collection procedures were used at T1 and T2. At each year of the project, participants completed self-report measures and peer nominations during 70-minute classroom-based data collection sessions toward the end of the school year. Trained research staff administered the questionnaires and read standardized instructions to the participants. The research team was available to answer any questions during data collection. The project was conducted in compliance with the ethical standards of the American Psychological Association and was approved by the University of Southern California’s Internal Review Board (study title: “Academic Success Project”, Protocol number: UP-17-00663). This study was not preregistered and the data and study materials for this study are not available. The study analysis code is available upon request.

Measures

Actual (Peer-Perceived) Social Status and Self-Perceived Status

Two dimensions of peer status (acceptance and rejection) were assessed with peer nominations. Participants were given a list of 50 randomly selected peers from their grade (e.g., Bellmore et al., 2010) and were asked to nominate their grademates on a variety of descriptors. Participants were asked to nominate students “that you really like” and “that you do not like that much” to assess acceptance and rejection, respectively. A proportion score for each status item was created, which represented the number of received nominations divided by the total number of possible nominations for each item. To evaluate self-perceived status, additional questions were asked right after the peer nomination items on status, where participants reported whether they thought other students picked them for that question (e.g., “Did a lot of other students say that they like you?”), with responses ranging from 1 (“definitely not”) to 7(“definitely yes”).

Aggression

Peer nominations were also used to measure two types of aggression. To assess overt aggression, participants were asked to identify peers who “hit or push other students” and “start fights with other students by pushing or punching them” (r = .65 and .67 at T1 and T2). To assess relational aggression, participants were asked to identify peers who “try to be mean to other students by ignoring them or excluding them” and “gossip about other students” (r = .60 and .63 at T1 and T2). For each item, the number of received nominations was divided by the total number of possible nominations.

Analytic Plan

To examine how actual and self-perceived rejection and acceptance were uniquely associated with concurrent levels of aggression, regression analyses were conducted separately for overt and relational aggression. To examine whether biases in perceptions of one’s status was associated with aggression, the interactions between self-perceived and actual rejection, as well as self-perceived and actual acceptance were also included in the regressions. For a significant interaction to indicate that bias is associated with aggression, high levels of aggression should be associated with high levels of self-perceived status and low levels of actual status (or vice versa). To investigate whether self-perceived status and biases in self-perceived status predict the development of aggression over time, this analysis was repeated to predict aggression at T2, adding aggression at T1 as a control variable.

To examine whether those high in aggression differ from others in how accurately they perceive their own acceptance and rejection, we tested the moderating effect of aggression on the association between self-perceived and actual status by regressing T1 self-perceived status on T1 actual status, aggression, and the interaction between actual status and aggression. A significant interaction between actual status and aggression would indicate that the extent to which youth overestimated or underestimated their status was dependent on how aggressive they were.

Next, to test whether actual status is differentially related to self-perceived status over time for aggressive youth, we repeated these analyses with self-perceived status at T2 as the dependent variable, controlling for initial levels of self-perceived status. In all analyses, separate models were conducted for overt and relational aggression.

Gender and race were dummy-coded and controlled for in all models. Gender was coded as 0 = girl and 1 = boy. Three dummycodes were created for race (with “Asian” as the reference group): 1. Latino/Hispanic, 2. Non-Hispanic White, and 3. Other. All continuous variables were centered at their mean before multiple imputation. Tests of regions of significance were conducted for significant interactions. For interpretation of the regions of significance, we translated the centered scores used in the interactions back to the original proportion score by adding the mean. For all analyses, gender was explored as a possible moderator and we examined whether including interactions with gender resulted in a significant change in R². We also conducted additional sensitivity analyses to examine whether our analyses differed by grade level.

Results

Means, standard deviations and correlations for the main study variables are presented in Table 1. Girls had higher levels of relational aggression and lower levels of overt aggression than boys at both time points. In addition, girls reported higher levels of self-perceived rejection at T2. There were no other significant gender differences. Overt and relational aggression were positively correlated at T1 and T2 (rs = .52 and .49, respectively). Both forms of aggression were positively associated with most other variables at T1 (rs ranging from .10 to .60), ps < .02. Relational aggression at T1 was positively correlated with self-perceived rejection at T1 and T2 (rs = .17 and .12, respectively) and with self-perceived acceptance at T1 (r = .10). In contrast, overt aggression at T1 was not significantly associated with self-perceived acceptance at T1 (r = .06) or T2 (r = .04). It was positively correlated with T1 self-perceived rejection (r = .12) but not with T2 self-perceived rejection (r = .04). Actual rejection was positively associated with self-perceived rejection (r = .12, p = .007), and actual acceptance was positively associated with self-perceived acceptance (r = .19, p < .001) at T1. There were significant, but modest, negative associations between self-perceived rejection and acceptance at T1 and T2 (rs = −.19 and −.16, respectively), which further emphasizes the importance of examining them separately.

Table 1.

Correlations, Means, and Independent Sample T-Tests of Gender Differences

Variable	1	2	3	4	5	6	7	8	9	M (SD)	M (SD)boys	M (SD)girls	t
1. Overt Aggression T1	—									.03 (.05)	.04 (.06)	.02 (.04)	−5.13***
2. Relational Aggression T1	.52***	—								.08 (.07)	.06 (.05)	.09 (.08)	5.52***
3. Actual Rejection T1	.47***	.60***	—							.03 (.03)	.03 (.03)	.03 (.03)	−0.89
4. Self-perceived Rejection T1	.12**	.17***	.12**	—						3.08 (1.44)	2.96 (1.49)	3.17 (1.39)	1.72
5. Actual Acceptance T1	.16**	.29***	.04	.06	—					.11 (.07)	.10 (.07)	.11 (.07)	1.53
6. Self-perceived Acceptance T1	.06	.10*	.04	−.19***	.19**	—				4.51 (1.31)	4.52 (1.47)	4.50 (1.17)	−.13
7. Overt Aggression T2	.44***	.30***	.35***	.16***	.04	.03	—			.02 (.04)	.03 (.04)	.02 (.04)	−.3.44***
8. Relational Aggression T2	.22***	.52***	.36***	.15***	.20***	.10*	.49***	—		.06 (.07)	.05 (.04)	.08 (.08)	5.95***
9. Self-perceived Rejection T2	.04	.12**	.07	.41***	.01	−.17***	.08	.17***	—	3.06 (1.39)	2.89 (1.45)	3.19 (1.32)	2.47*
10. Self-perceived Acceptance T2	.04	−.00	−.02	−.20***	.11**	.44***	−.03	.02	−.16***	4.47 (1.30)	4.52 (1.44)	4.42 (1.18)	−0.85

Note.

^*p < .05.

^**p < .01.

^***p < .001.

Effects of Self-Perceived Status and Biases in Status Perceptions on Aggression

First, we tested the concurrent main and interactive effects of self-perceived and actual status on aggression (see Table 2). The models predicting T1 overt aggression, Fs(10, 561) > 27.09, p < .001, and T1 relational aggression, Fs(10, 561) > 56.54, p < .001 were both significant. Neither self-perceived rejection nor self-perceived acceptance were significantly associated with T1 overt aggression (ps > .10), whereas both were positively associated with T1 relational aggression (bs = .004, SE = .002, ps < .04). Actual acceptance and actual rejection were positively associated with both T1 overt (bs = .12 and .58, SE = .03 and .05, ps < .001, respectively) and relational aggression (bs = .25 and 1.18, SE = .03 and .06, ps < .001, respectively).

Table 2.

Predicting T1 Aggression From Self-Perceived Social Status and Actual Social Status

	Overt aggression T1			Relational aggression T1
Variable	b	SE	95% CI	b	SE	95% CI
Male	0.02***	0.00	[0.02, 0.03]	−0.03***	0.00	[−0.04, −0.02]
Latino/Hispanic	0.01***	0.00	[0.01, 0.02]	0.01	0.00	[−0.00, 0.02]
White	−0.01	0.01	[−0.03, 0.00]	−0.01	0.01	[−0.03, 0.01]
Other	0.01	0.01	[−0.00, 0.03]	−0.00	0.01	[−0.02, 0.02]
Actual Rejection T1	0.58***	0.05	[0.48, 0.68]	1.16***	0.06	[1.05, 1.30]
Self-perceived Rejection T1	0.00	0.00	[−0.00, 0.00]	0.00*	0.00	[0.00, 0.01]
Actual Acceptance T1	0.12***	0.03	[0.07, 0.17]	0.25***	0.03	[0.19, 0.32]
Self-perceived Acceptance T1	0.00	0.00	[−0.00, 0.00]	0.00*	0.00	[0.00, 0.01]
Actual Rejection × Self-perceived Rejection T1	0.12**	0.03	[0.05, 0.19]	0.09	0.05	[−0.01, 0.18]
Actual Acceptance × Self-perceived Acceptance T1	0.01	0.02	[−0.03, 0.05]	0.07**	0.03	[0.02, 0.12]
Total Adjusted R2	.318			.497

^*p < .05.

^**p < .01.

^***p < .001.

The interaction between self-perceived and actual rejection was a significant predictor of T1 overt aggression (b = .12, SE = .03, p < .001; see Figure 1A), whereas the interaction between self-perceived and actual acceptance was not significant (b = .01, SE = .02, p = .76). At low levels of actual rejection, self-perceived rejection at T1 was not associated with overt aggression at T1 (simple slopes test, b = −.001, SE = .002, t = −.45, p = .65). However, at high levels of actual rejection, self-perceived rejection was associated with higher T1 overt aggression (simple slopes test, b = .005, SE = .002, t = 3.26, p = .001). Youth who were highly disliked and thought they were highly disliked had high concurrent levels of overt aggression. There was a marginal association between self-perceived rejection and T1 overt aggression at average levels of actual rejection (simple slopes test, b = .002, SE = .001, t = 1.87, p = .06). Tests of regions of significance showed that when actual rejection was greater than .03, the slope of self-perceived rejection was significant and positive. Gender did not further qualify this association.

Figure 1.

Interaction Between T1 Self-Perceived Rejection and T1 Actual Rejection Predicts T1 Overt Aggression (A) and Interaction between T1 Self-Perceived Acceptance and T1 Actual Acceptance Predicts T1 Relational Aggression (B)

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

The effect of the interaction between self-perceived and actual rejection on relational aggression at T1 was trending but did not reach significance (b = .09, SE = .05, p = .066); an exploratory probe of the interaction found the same pattern as for overt aggression. The interaction between self-perceived and actual acceptance, however, was a significant predictor of relational aggression at T1 (p = .008; see Figure 1B). At low levels of actual acceptance, self-perceived acceptance was not associated with relational aggression at T1 (simple slopes test, b = .000, SE = .002, t = .12, p = .90). At high and average levels of actual acceptance, self-perceived acceptance was associated with higher T1 relational aggression (simple slopes test, b = .009, SE = .003, t = 3.29, p = .001 and b = .004, SE = .002, t = 2.58, p = .01). Thus, youth who thought they were well-liked and who were at least moderately well-liked also had elevated concurrent levels of relational aggression. When actual acceptance was greater than .09, the slope of self-perceived acceptance was significant and positive. Gender did not further qualify this association.

Next, we examined the prospective association between self-perceived status and T2 aggression. The model testing the main and interactive effects of T1 self-perceived and actual status on T2 overt and relational aggression were both significant, Fs(11, 560) > 17.67, p < .001 (see Table 3). Overt and relational aggression were both relatively stable from T1 to T2 (bs = .23 and .33, SEs = .04 and 05, p < .001, respectively). Self-perceived and actual acceptance were not significantly associated with T2 overt or relational aggression. Actual rejection was positively associated with both T2 overt and relational aggression (bs = .19 and .20, SE = .05 and .09, ps < .02). Self-perceived rejection was positively associated with T2 overt aggression (b = .003, SE = .001, p = .002) and marginally associated with T2 relational aggression (b = .003, SE = .002, p = .06).

Table 3.

Predicting T2 Aggression From Self-Perceived Social Status and Actual Social Status

	Overt aggression T2			Relational aggression T2
Variable	b	SE	95% CI	b	SE	95% CI
Male	0.01*	0.00	[0.00, 0.01]	−0.02***	0.00	[−0.03, −0.01]
Latino/Hispanic	0.01*	0.00	[0.00, 0.01]	−0.00	0.00	[−0.01, 0.01]
White	0.01	0.01	[−0.01, 0.02]	0.00	0.01	[−0.02, 0.03]
Other	0.03***	0.01	[0.01, 0.04]	0.03*	0.01	[0.01, 0.05]
Aggression T1	0.23***	0.04	[0.16, 0.31]	0.33***	0.05	[0.24, 0.43]
Actual Rejection T1	0.19***	0.05	[0.09, 0.28]	0.20*	0.09	[0.03, 0.37]
Self-perceived Rejection T1	0.00**	0.00	[0.00, 0.01]	0.00	0.00	[−0.00, 0.01]
Actual Acceptance T1	−0.01	0.02	[−0.05, 0.04]	0.06	0.04	[−0.01, 0.13]
Self-perceived Acceptance T1	0.00	0.00	[−0.00, 0.00]	0.00	0.00	[−0.00, 0.01]
Actual Rejection × Self-perceived Rejection T1	0.06*	0.03	[0.00, 0.12]	0.13**	0.05	[0.03, 0.22]
Actual Acceptance × Self-perceived Acceptance T1	0.01	0.02	[−0.02, 0.05]	0.02	0.03	[−0.03, 0.08]
Total Adjusted R2	.248			.304

^*p < .05.

^**p < .01.

^***p < .001.

Furthermore, the interaction between self-perceived rejection and actual rejection significantly predicted both T2 overt and relational aggression (Figure 2A and 2B). Self-perceived rejection at T1 was positively associated with T2 overt aggression at high and average levels of T1 actual rejection (simple slopes test, bs = .006 and .003, SEs = .001, ps < .004, respectively), but not at low levels of T1 actual rejection (b = .000, SE = .002, t = .27, p = .78). Thus, self-perceived rejection positively predicted overt aggression over time, for youth who were at least moderately disliked. Tests of regions of significance showed that when actual rejection was greater than .02, the slope of self-perceived rejection was significant and positive.

Figure 2.

Interaction Between T1 Self-Perceived Rejection and T1 Actual Rejection Predicts T2 Overt Aggression (A) and T2 Relational Aggression (B)

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

A similar pattern was found for relational aggression at T2. At high levels of T1 actual rejection, self-perceived rejection was associated with higher T2 relational aggression (simple slopes test, b = .008, SE = .002, t = 3.56, p < .001). However, at low and average levels of T1 actual rejection, self-perceived rejection at T1 was not associated with relational aggression at T2 (simple slopes test, b = −.002, SE = .002, t = −.79, p =.43 and b = .003, SE = .002, t = 1.83, p = .07). This suggests that adolescents thinking they were disliked subsequently had elevated levels of relational aggression, but only when they were actually disliked. Tests of regions of significance indicated that when actual rejection was greater than .03, the slope of self-perceived rejection was significant and positive. Exploratory analyses testing for gender differences did not result in a significant change in R², and there was no significant 3-way interaction between self-perceived status, actual status, and gender when predicting either form of T2 aggression.

Moderating Effect of Aggression on the Concurrent and Prospective Associations Between Actual and Self-Perceived Status

First, we examined whether the interaction between actual status and aggression was associated with self-perceived status at T1. The models with self-perceived rejection at T1 as the dependent variable were significant, Fs(7, 564) > 2.07, ps < .045 (see Table 4). Relational aggression (b = 3.06, SE = 1.21, p = .01), but not overt aggression, was positively associated with self-perceived rejection at T1. Neither the interaction between actual rejection and overt aggression (b = 12.06, SE = 26.46, p = .65) nor the interaction between actual rejection and relational aggression (b = −9.51, SE = 14.27, p = .51) were significant, suggesting that the association between actual rejection and self-perceived rejection - or accuracy of self-perceived rejection - did not depend on level of aggression. The 3-way interactions between gender, actual rejection, and either form of aggression were not statistically significant (ps > .29).

Table 4.

Predicting T1 Self-Perceived Social Status From the Interaction Between Aggression and Actual Social Status

	Dependent variable: Self-Perceived social status T1
	Rejection			Acceptance
Variable	b	SE	95% CI	b	SE	95% CI
Male	−0.29*	0.13	[−.54, −.05]	−0.00	0.11	[−0.22, 0.22]
Latino/Hispanic	0.02	0.13	[−.24, .28]	−0.25*	0.12	[−0.47, −0.02]
White	0.04	0.31	[−.57, .65]	0.33	0.28	[−0.22, 0.88]
Other	−0.32	0.28	[−.87, .24]	−0.27	0.25	[−0.76, 0.23]
Overt Aggression T1	2.58	1.63	[−.62, 5.79]	1.25	1.21	[−1.13, 3.62]
Actual Social Status T1	2.73	2.13	[−1.45, 6.91]	3.41***	0.85	[1.75, 5.08]
Overt Aggression × Actual Social Status T1	12.06	26.46	[−39.91, 64.03]	−3.95	17.70	[−38.72, 30.82]
Total Adjusted R2	.016			.037
	b	SE	95% CI	b	SE	95% CI
Male	−0.15	0.13	[−0.40, 0.10]	0.07	0.11	[−0.15, 0.29]
Latino/Hispanic	0.03	0.13	[−0.22, 0.28]	−0.24*	0.11	[−0.46, −0.01]
White	0.06	0.31	[−0.56, 0.67]	0.34	0.28	[−0.22, 0.89]
Other	−0.27	0.28	[−0.82, 0.29]	−0.26	0.25	[−0.75, 0.23]
Relational Aggression T1	3.06*	1.21	[0.68, 5.43]	1.54	0.92	[−0.27, 3.36]
Actual Social Status T1	1.90	2.34	[−2.70, 6.49]	3.21***	0.86	[1.52, 4.91]
Relational Aggression × Actual Social Status T1	−9.51	14.27	[−37.55, 18.52]	−7.76	10.01	[−27.44, 11.92]
Total Adjusted R2	.021			.040

^*p < .05.

^**p < .01.

^***p < .001.

The models were also significant when testing whether the interaction between actual acceptance and overt and relational aggression was associated with T1 self-perceived acceptance, Fs(7, 564) > 3.81, ps < .001 (see Table 4). In both models, actual acceptance was positively associated with self-perceived acceptance (bs > 3.21, SEs < .86, ps < .001). None of the interactions between actual acceptance and either overt or relational aggression were significant (bs = −3.95 and −7.76, SEs = 17.70 and 10.01, ps > .44), again indicating that the association between actual and self-perceived status did not depend on level of aggression. There was no statistically significant 3-way interaction between gender, actual acceptance, and either form of aggression (ps > .64).

Next, we examined whether the interaction between actual status and aggression was associated with T2 self-perceived status, controlling for T1 self-perceived status (see Table 5). There were no significant interactions between actual status (rejection or acceptance) and aggression (overt or relational), ps > .11. Exploratory analyses testing for gender differences did not result in a significant change in R² when examining the interaction between actual status, aggression, and gender.

Table 5.

Predicting T2 Self-Perceived Social Status From the Interaction Between Aggression and Actual Social Status

	Dependent variable: Self-Perceived social status T2
	Rejection			Acceptance
Variable	b	SE	95% CI	b	SE	95% CI
Male	−0.21	0.11	[−0.43, 0.01]	0.08	0.10	[−0.12, 0.29]
Latino/Hispanic	0.00	0.11	[−0.22, 0.23]	0.02	0.11	[−0.19, 0.23]
White	−0.25	0.28	[−0.80, 0.30]	−0.33	0.26	[−0.84, 0.18]
Other	−0.12	0.25	[−0.61, 0.37]	−0.37	0.23	[−0.83, 0.08]
Self-Perceived Social Status T1	0.38***	0.04	[0.30, 0.45]	0.42***	0.04	[0.35, 0.50]
Overt Aggression T1	0.86	1.53	[−2.17, 3.90]	0.27	1.12	[−1.92, 2.46]
Actual Social Status T1	1.26	1.90	[−2.47, 4.98]	0.85	0.82	[−0.77, 2.48]
Overt Aggression × Actual Social Status T1	−22.61	25.13	[−72.24, 27.02]	−9.01	16.60	[−41.66, 23.63]
Total Adjusted R2	.160			.186
	b	SE	95% CI	b	SE	95% CI
Male	−0.18	0.11	[−0.40, 0.04]	0.06	0.10	[−0.14, 0.27]
Latino/Hispanic	0.01	0.11	[−0.21, 0.23]	0.04	0.10	[−0.17, 0.24]
White	−0.26	0.28	[−0.81, 0.29]	−0.34	0.26	[−0.84, 0.17]
Other	−0.12	0.25	[−0.61, 0.38]	−0.36	0.23	[−0.80, 0.09]
Self-Perceived Social Status T1	0.38***	0.04	[0.30, 0.45]	0.43***	0.04	[0.35, 0.50]
Relational Aggression T1	0.86	1.13	[−1.36, 3.09]	−0.58	0.83	[−2.20, 1.05]
Actual Social Status T1	0.20	2.07	[−3.87, 4.28]	1.21	0.84	[−0.46, 2.88]
Relational Aggression × Actual Social Status T1	−5.28	13.78	[−32.61, 22.05]	−14.96	9.22	[−33.11, 3.18]
Total Adjusted R2	.160			.192

^*p < .05.

^**p < .01.

^***p < .001.

Additional Sensitivity Analyses

All our models were also conducted separately for 9th and 10th graders as 9th grade was a transition year to a new school and this could influence associations between social status and aggressive behavior (e.g., Farmer et al., 2011). Thus, we explored whether these processes differed for 9th and 10th graders. Adding grade as a moderator only resulted in a significant change in R² for one model. The 2-way interaction between self-perceived rejection and actual rejection in predicting T2 relational aggression was further qualified by grade (p < .001). For 9th graders, self-perceived rejection was positively associated with T2 relational aggression at high (b = .01, SE = .003, t = 4.71, p < .001) and average levels (b = .004, SE = .002, t = 2.00, p = .046) of actual rejection, but not low levels of actual rejection (b = −.005, SE = .003, t = −1.49, p = .137); for 10th graders, the association between these two variables was not significant, regardless of students’ actual rejection (bs = .001, .001, −.001, SEs = .004, .003, .004, ps > .66 for low, average, and high levels of actual rejection, respectively).

Discussion

There is ample evidence that youth’s actual social status—rejection and acceptance—matter for their aggressive behavior (Casper et al., 2020). However, the literature on the links between aggression and youth’s perceptions of their own status, as well as biases in self-perceived status, are less clear. Whereas theoretical perspectives on why actual rejection leads to aggression tend to imply that it is related to one’s self-perceptions of being rejected, findings on the association between self-perceived status and aggression have been inconsistent. Previous research is also inconsistent with regard to the associations between biases in self-perceived status and aggression. As the answer to these questions have important theoretical and practical implications, the current study aimed to further clarify the concurrent and longitudinal associations of adolescents’ self-perceptions of status and biases in self-perceived status with their aggressive behavior. Adolescence was a particularly relevant period to examine these associations as it is an age when cognitive skills (e.g., Kuhn, 2000; Malloy et al., 2007) should be sufficiently developed, yet when the increased importance of status among peers may promote positive biases in the social domain (Hoza et al., 2010). To address some of the methodological limitations of past studies, we considered both acceptance and rejection separately and relied on moderation analyses to investigate bias in aggressive youth’s assessment of their own status.

Are Self-Perceptions of Status Related to Aggression?

We first investigated how self-perceptions of acceptance and rejection were related to concurrent and future levels of aggression. Self-perceived acceptance and rejection were both positively associated with concurrent relational aggression, but not concurrent overt aggression. Self-perceived rejection, but not self-perceived acceptance, was positively associated with future overt and relational aggression (although the association was only marginally significant for relational aggression). However, some of these positive associations varied as a function of actual status (see below). Both actual acceptance and actual rejection were positively associated with concurrent overt and relational aggression, whereas only actual rejection predicted elevated aggression over time. The findings that rejection and acceptance were only moderately correlated with one another and were both positively associated with concurrent aggression supports the argument that the two types of status are not opposite ends of one continuum (e.g., Gorman et al., 2011); their differential effects on future aggression further emphasize the importance of examining them as distinct constructs (see Véronneau & Dishion, 2010). Although it may seem surprising that acceptance was positively associated with aggression, this may reflect the fact that some aggressive youth are both liked and disliked (i.e., “controversial”; Coie et al., 1982). Moreover, these findings could indicate that several different pathways (i.e., self-perceived rejection, self-perceived acceptance) could underlie aggression (i.e., equifinality). For example, self-perceived rejection may be associated with aggression due to the negative emotions elicited by feeling rejected (e.g., frustrationaggression hypothesis), whereas self-perceived acceptance may be associated with aggression out of a sense of entitlement (i.e., belief in an ability to be aggressive without suffering consequences).

Are Biases in Self-Perceived Status Related to Aggression?

As we expected, whether or not self-perceived status was related to concurrent aggression depended on actual status. At moderate and high levels of actual acceptance, high self-perceived acceptance was related to elevated levels of concurrent relational aggression, compared to when actual acceptance and self-perceived acceptance were incongruent. Our findings demonstrate that awareness of being highly well-liked is associated with higher relational aggression. At moderate levels of acceptance, there is some indication that overestimating acceptance is linked to elevated relational aggression. In general, acceptance is considered to be a “lower-risk” type of status (compared to perceived popularity), and aggression is generally thought of as a behavior that is met with disapproval by peers in terms of acceptance; these findings suggest that it may not be that simple. Adolescent aggressors are typically viewed as popular but disliked (e.g., Sandstrom & Cillessen, 2006). However, as also indicated by the positive correlation between aggression and acceptance in this study, it is important to remember that youth can be both aggressive and well-liked. It is possible that adolescents who are aware of being well-liked are able to strategically use relational aggression in a way that does not risk their acceptance (e.g., by using aggression against low-risk targets; Veenstra et al., 2010). Consistent with this idea, we found that being at least moderately liked and perceiving oneself as accepted was significantly associated only with relational aggression, which may involve more social savvy than overt aggression (Björkqvist et al., 2000). This finding is also consistent with past research demonstrating that aggressive youth are not universally disliked by peers (de Vries et al., 2021; Hafen et al., 2013). Furthermore, youth need to utilize social relationships to effectively use relational aggression; thus, it makes sense that youth who use relational aggression can also be well-liked and integrated in the peer group.

In addition, we did not find support for the notion that underestimating one’s rejection was associated with higher levels of concurrent aggression. Rather, perceiving oneself as highly rejected was associated with high levels of concurrent overt aggression, even at moderate levels of rejection. This could suggest that for youth who experience even some level of rejection from peers, it is overestimation (not underestimation) of one’s rejection that is associated with overt aggression (a similar pattern was found for relational aggression but did not reach statistical significance). One should note that actual rejection remained positively associated with concurrent overt aggression, regardless of the adolescent’s self-perceived rejection. This could indicate that adolescents with high levels of overt aggression are met with disapproval from peers whether they realize it or not; however, the highest levels of overt aggression were found for those who realized they were rejected.

When predicting aggression one year later, initial levels of self-perceived rejection positively predicted both forms of aggression, controlling for prior levels of aggression. For overt aggression only, at least moderate rejection from peers was associated with elevated levels of aggression over time when self-perceived rejection was also high. Tests of regions of significance showed that relational aggression was higher over time when actual rejection was at least slightly above average and self-perceived rejection was high. At low levels of self-perceived rejection, actual rejection was not significantly associated with subsequent overt or relational aggression. This suggests that it is only when youth feel rejected that actual rejection is related to future aggression. In contrast, some youth may be blissfully unaware of being rejected by their peers and thus do not respond aggressively. Our findings suggest that awareness of being rejected had more of an impact on future overt and relational aggression than awareness of being accepted. For relational aggression, however, this finding should be interpreted with caution, as our sensitivity analyses demonstrated that this pattern was present in 9th graders, but not 10th graders. As 9th grade was a transition year, youth who are and feel rejected may try to use relational aggression (a more strategic form of aggression) to advance their position in the peer group before reputations become more stable.

Does the Association Between Actual and Perceived Status Depend on Aggression?

Although these findings generally suggest that adolescents who tend to be aware of being highly rejected or highly accepted are more aggressive, they do not answer the question of whether those high in aggression differ from others in the extent to which they use social cues regarding their actual status to inform their self-perceptions of status. Adolescents in general were not very accurate in assessing their own status, as reflected by the modest correlations between actual rejection and self-perceived rejection (r = .13) and between actual acceptance and self-perceived acceptance (r = .18). Furthermore, our findings did not provide any evidence that this pattern differed at high versus low levels of aggression. We conducted concurrent analyses, in which the association between actual status and self-perceived status was not significantly moderated by either form of aggression. Therefore, the hypothesis that biases in self-perceived status would be stronger for adolescents with high levels of aggression was not supported. We then tested whether aggression moderated the extent to which actual status predicted self-perceived status over time. As with the concurrent analyses, aggression did not moderate the association between initial actual status and future self-perceived status. Thus, our findings do not indicate that adolescents with high levels of aggression show biases when inferring their status based on cues in their environment about their actual status.

This is consistent with the body of studies showing that adolescents and children with high levels of aggression do not exhibit more biases when evaluating their acceptance or rejection than their peers (e.g., Garandeau & Lansu, 2019; García-Bacete et al., 2019; Mayeux & Cillessen, 2008; White & Kistner, 2011). We have known for a long time that not all perpetrators of aggression are rejected and maladjusted youth, and the current results support this perspective that aggressors do not necessarily show deficiencies in their processing of social information and can be socially competent (Garandeau & Cillessen, 2006; Guy et al., 2017; Sutton et al., 2001). Importantly, the current study highlights that youth with high levels of aggression are not oblivious of their peer status, which suggests that attempts at raising students’ awareness that their aggressive behavior may be costly in terms of acceptance may not be the optimal intervention strategy.

Although our tests of aggression as the moderator of the links between self-perceived and actual status are not consistent with past findings of self-enhancing tendencies in aggressive youth (e.g., Stephens et al., 2016), these results do not necessarily imply that aggression is not associated with a tendency for self-enhancement when it comes to qualities other than status (e.g., intelligence, athleticism or physical appearance). The present findings could instead be an indication that aggressive youth do not perceive acceptance as an enviable quality and value other qualities more (e.g., popularity; Garandeau & Lansu, 2019). In other words, it is conceivable that aggression is associated with a narcissistic propensity for self-enhancement for a number of qualities that aggressive youth consider desirable, and acceptance is not one of them.

Alternatively, it is possible that these differences are related to the age of our sample. Many of the studies that have found support of a positive link between aggression and biased perceptions of status have been conducted with children. Although it is beyond the scope of the current study, a possible explanation is that the positive association between aggression and bias found in childhood may be due to the fact that both can be explained by lower cognitive skills, at least for some children.

Taken together, our findings suggest that, for those high in rejection or acceptance, awareness of one’s status is associated with concurrent aggression, and, for those high in rejection, awareness of one’s rejection is also associated with future aggression. Moreover, youth with high levels of aggression were not more likely than others to show biases in self-perceptions of status. Our findings also shed light on how self-perceived rejection is associated with aggression, which has received much less empirical attention in the literature. The reasons for why being aware of being disliked may lead to aggression have already been discussed (see Leary et al., 2006). Nevertheless, given that individuals have a “need to belong” and should be highly motivated to avoid social rejection (Baumeister & Leary, 1995), this finding raises another important question—if youth are aware of being disliked, then why does this not deter them from being aggressive? Some aggressive youth may be disliked by many peers but well-liked by a few (e.g., friends), which may be enough for them. Indeed, the positive correlation between (overt and relational) aggression and acceptance found in our study—though much weaker than the positive correlation between aggression and rejection—does suggest that aggressive youth are not universally disliked and do have some peers who like them, which may complicate efforts to reduce aggression.

Strengths, Limitations, and Future Directions

The extant literature on how self-perceived status and actual status relate to aggression has been limited in several ways. The majority of studies have focused on measures of social preference (the difference score between nominations of liked most and liked least) rather than examining acceptance and rejection separately. Moreover, most studies have operationalized biases in perceived status using difference scores or standardized residual scores, both of which have significant limitations (see De Los Reyes & Kazdin, 2004). The current study is one of the few studies to examine the prospective association between self-perceptions of status and aggression, and highlighted that actual and self-perceived rejection play more of a role in aggression over time than actual and self-perceived acceptance, while both were relevant for concurrent levels of aggression. Furthermore, the current study built on past research by testing whether the interaction between self-perceived status and actual status was moderated by aggression, with independent examination of acceptance and rejection; this design allowed us to avoid the limitations of other methodologies (Zuckerman et al., 2002).

Nevertheless, the current study has several limitations. First, the lack of information regarding dyadic nominations implies that we could not examine by whom participants were liked or disliked. Some youth may have a reputation in the classroom as disliked compared to other peers, but still be liked by certain classmates. For example, the association between bullying and being disliked was found to be primarily driven by peers who were highly victimized (Hafen et al., 2013). As general status is distinct from liking and disliking within specific dyads, a valid assessment of biases in self-perceived status would require a differentiation between accuracy of one’s general reputation in the peer group and accuracy about whether one is liked or disliked by specific others. The current study shows that youth with high levels of aggression generally do not significantly differ in the accuracy of their perception of their general social status compared to less aggressive youth, but it is possible that there are differences regarding their level of dyadic accuracy. Future research should use dyadic nominations, not only of actual but also self-perceived acceptance and rejection, to examine whether perpetrators differ from their less aggressive classmates. In addition, future research could examine whether feeling liked by the peers they care most about (e.g., their friends, popular peers) despite feeling generally disliked contributes to the aggressive behavior of some youth.

Second, there are variables related to one’s actual and self-perceived status that may impact their association with aggression but were not accounted for in the current study. For example, whether or not youth who are rejected and know they are rejected become more aggressive over time may be related to their social goals or how much value they place on being well-liked. Given that not all youth value being well-liked to the same extent (e.g., Garandeau & Lansu, 2019), this is another important factor for future research to consider. Aggressive youth may be less likely to overestimate their acceptance when they do not perceive acceptance as important. Relatedly, youth’s actual popularity is another relevant factor to consider, as aggressive adolescents who enjoy high levels of popularity may be more likely to overestimate their acceptance if they mistake peers’ attention with affection.

Third, the variable-centered analyses used in the current study do not account for the possible heterogeneity of aggressive adolescents. The inconsistencies in the literature, as well as those found in the current study, in regard to whether aggressive youth overestimate their acceptance or know that they are disliked may simply reflect the fact that aggressive adolescents form a heterogenous group. Future research should utilize person-centered analyses to determine whether there are meaningful differences in accuracy of self-perceived status within aggressive youth, and, if so, what are the key factors related to these differences. Finally, self-perceived and actual status were each measured with only one item; however, single items have been used by past research (e.g., Mayeux & Cillessen, 2008) and allow for precise measurement of the construct. Furthermore, each individual score is computed on the basis of multiple informants.

Conclusion

Understanding the causes of aggressive behavior requires good knowledge of the cognitions of youth. The current study investigated youth’s social cognitions related to their acceptance and rejection among peers and focused on adolescence, when most individuals have acquired the cognitive abilities needed for assessing one’s own peer status and place value on such status. Avoiding the limited methodologies often used by past studies, it adds to this discussion by demonstrating that for youth who were highly rejected or accepted, awareness of their status was positively associated with concurrent aggression. Moreover, our results suggest that actual and self-perceived rejection played a larger role in the development of aggression than actual and self-perceived acceptance. Our longitudinal analyses further showed that awareness of being highly rejected was positively associated with future aggression. In addition, the current study did not detect any evidence that youth with high levels of aggression had more biased perceptions of their acceptance or rejection than their peers, concurrently or over time. From an intervention perspective, this suggests that attempts to increase adolescents’ awareness of the potential status costs of their behavior is unlikely to be effective in reducing their propensity for aggression. To properly inform interventions, future research should identify the exact reasons why the awareness of being disliked does not deter youth from being aggressive.