Introduction

Romantic relationships serve important developmental functions in adolescence, and they are a prominent aspect of peer relations for many adolescents (Collins 2003). Adolescent interactions with romantic partners outnumber interactions with friends, parents, and siblings by 10th grade (Laursen and Williams 1997), and romantic partners tie with mothers for second place in the support hierarchy of 10th-graders (Furman and Buhrmester 1992). In a study of 12- to 18-year-olds, Carver et al. (2003) found that more than half of teens reported having a romantic relationship in the previous 18 months; furthermore, many of those relationships were long-lasting. Romantic relationships have been shown to make unique contributions to adolescent development (Roisman et al. 2009). Like adults, adolescents view positive characteristics such as dependability, honesty, humor, and kindness as important in romantic partners (Regan and Joshi 2003; Weber and Ruch 2012). However, a growing body of research indicates that antisocial behaviors might also facilitate success in dating relationships. For example, studies have shown that engaging in aggressive behavior predicts dating status and dating popularity both concurrently and longitudinally (Arnocky and Vaillancourt 2012; Connolly et al. 2000b; Pelligrini and Long 2003). The factors that play a role in adolescent dating and perceptions of dating partners are complex.

Peer status has been shown to play a role in adolescents’ choice of dating partner. One of the key functions of romantic relationships in early-to-mid-adolescence concerns youths’ sense of belonging, acceptance, and status in the peer group (Brown 1999). As Brown notes, “[d]ating itself is a source of peer status for adolescents” (Brown 1999, p. 309), and given the high priority many adolescents place on status (LaFontana and Cillessen 2010), it makes sense that young teens might be preoccupied with dating and romantic pursuits that help maintain or improve their social position among peers. This premise also assumes that some adolescents will view high peer status as a desirable quality in a potential romantic partner.

Brown (1999) proposes a four-stage developmental–contextual approach to understanding the role of the peer group in adolescent romantic relationships. He argues that the nature of romantic relationships changes across adolescence, and that these changes are strongly influenced by features of the peer context. The youngest adolescents are characterized as entering the initiation phase of romantic development, in which puberty as well as peer group pressures influence youths to reflect upon their own awakening romantic desires and their own worth as potential romantic partners. Next comes the status phase, in which adolescents focus on the peer group’s perceptions of the romantic partners they choose, and the possible consequences of those perceptions. It is during this phase that concerns about status and belonging are most prominent; adolescents evaluate potential partners in terms of their status-enhancing potential and are more preoccupied with the status implications of their relationships than with the relationships themselves. The third phase of romantic development is the affection phase, in which the focus shifts away from concerns about peers’ perceptions and onto developing meaningful, affectionate ties with romantic partners. Finally, older adolescents enter the bonding phase, in which they contemplate both emotional and practical issues related to forming committed, long-term relationships.

In this study, we focus on mid-adolescents’ concerns about dating from the perspective of the status phase. Drawing from Brown’s developmental–contextual approach, we propose that adolescents’ peer status is a key component of their desirability as a potential romantic partner, and that peers of higher status will be seen as more valuable dating partners. Further, we propose that aggression functions to increase social visibility and access to social resources, such as access to potential partners (e.g., Hawley 1999; Pelligrini and Long 2003). Indeed, peer status, specifically peer popularity, is strongly related to aggressive behavior in adolescence (e.g., Cillessen and Rose 2005). Thus, we predict that aggression will also be positively associated with dating popularity. Below, we describe previous research linking peer status and aggressive behavior to dating and other indices of romantic development.

Peer Status: Being Liked Versus Being Popular

Peer relations researchers distinguish between two types of peer status among adolescents: social preference, or being accepted and well-liked by peers in general, and popularity, which is an index of social visibility and power (Cillessen and Rose 2005; Mayeux et al. 2011; Parkhurst and Hopmeyer 1998). Social preference is typically positively correlated with behavioral characteristics like prosocial behavior and sociability, and negatively correlated with aggression and other forms of antisocial behavior (Rubin et al. 2006). Popularity, on the other hand, is positively correlated with both prosocial and aggressive behavior (Cillessen and Mayeux 2004; Hawley 2003; Mayeux et al. 2011; Puckett et al. 2008). Social preference and popularity are thus distinct forms of peer status.

Romantic development is strongly situated in the peer context. Because romantic relationships among teens often stem from their integration into the broader peer group, teens with more friends in general are more likely to have a romantic partner (Connolly et al. 2000a). Indeed, social preference and popularity—both of which are related to having relatively high social impact in the peer group—have been shown to be related to dating behavior in adolescence. Well-liked adolescents, who generally have the types of skills in social behavior and emotion regulation that are helpful in facilitating healthy relationships, likely benefit from both their social competencies and their integration into the peer network. In a study of seventh-graders, being well-liked by peers was related to having a romantic partner (Miller et al. 2009). Similar findings were found in a study of third-, fifth-, seventh-, and ninth-graders (Carlson and Rose 2007) and in a study of high school students (Franzoi et al. 1994). The source of peer liking may also play a role. Perhaps not surprisingly, dating popularity was more closely related to being liked by the opposite sex than to being liked by same-sex peers in a study of adolescents (Dijkstra et al. 2010).

A small number of studies suggests that popular youths are also more likely to be engaged in romantic relationships than their average-status or unpopular peers. Controversial peer status (which is closely related to being popular; see Parkhurst and Hopmeyer 1998) was associated with having a romantic partner among the seventh-graders in Miller and colleagues’ study (Miller et al. 2009). Carlson and Rose (2007) also found popularity to be associated with being involved in a reciprocated romantic relationship. In a study of seventh- and ninth-graders, a hypothetical popular peer elicited more romantic jealousy than a hypothetical unpopular peer (Mayeux 2011). In an experimental study, Ha et al. (2010) examined the role of attractiveness and peer status on dating desire among adolescent boys and girls. For boys, peer status was only important when evaluating attractive potential partners. For girls, peer status was important for both attractive and unattractive potential partners. Popular peers are probably seen as valuable potential dates for many reasons, including their peer-valued characteristics such as wealth, attractiveness, and sense of humor (Vaillancourt and Hymel 2006). They may also be seen as a means of status improvement; adolescents who wish they could be a part of the popular crowd might view dating a popular peer as a good way of ensuring entry into that crowd (Brown 1999). This strategy may well work: The popularity of an individual’s best friends in one school year predicts changes in popularity for that individual in the next school year. Adolescents who associate with popular peers experience an increase in popularity; those who are friends with unpopular peers experience a decrease in popularity (Marks et al. 2012). The same process may occur in romantic relationships, with a higher-status partner raising the status of a lower-status adolescent simply by association.

Overt and Relational Aggression and Dating

Researchers also typically distinguish between two general types of aggression: physical harm and verbal teasing or threatening behaviors (often referred to simply as overt or physical aggression; Dodge et al. 2006), and aggression that is geared toward harming the target’s social standing or peer relationships (called relational, social, or indirect aggression; Crick 1995; Underwood 2003; Vaillancourt 2005). Both overt and relational aggression have been found to predict dating popularity or dating status. Adolescent girls who were higher in overt physical and verbal forms of aggression were also more likely to report that they had a romantic partner than less aggressive girls in the study by Miller et al. (2009). However, another study found overt aggression to negatively predict dating status 1 year later for boys (Arnocky and Vaillancourt 2012). Gallup et al.’s (2011) retrospective study of undergraduates’ middle and high school dating experiences found an association between same-sex peer aggression and dating outcomes. Girls who engaged in nonphysical modes of aggression, such as teasing and excluding others, tended to begin dating at earlier ages. For males, the use of nonphysical forms of aggression was associated with a greater number of dating partners in adolescence. However, other studies have shown increases in dating popularity over time for boys high in physical assertiveness (Pellegrini and Bartini 2001). Indirect aggression was positively correlated with dating status, and predicted dating status 1 year later, controlling for attractiveness and popularity, in a study of sixth- through ninth-grade boys and girls (Arnocky and Vaillancourt 2012). Relational aggression also predicted increases in dating popularity over time for girls in a study of middle school students (Pelligrini and Long 2003). Thus, while the evidence of indirect or relational forms of aggression enhancing dating is quite consistent, findings regarding overt aggression have been mixed.

Aggression and romantic development are correlated even at more extreme levels of aggression. Bullies have been shown to start dating at a younger age, to engage in a wider variety of dating activities with their partners, and to spend more non-school time with their dating partners than non-bullies (Connolly et al. 2000a, b). This was true despite the fact that bullies were likely to use physical or relational aggression against their own dating partners. Indeed, boys who engage in peer-directed violence are more likely to engage in dating violence (McNaughton Reyes and Foshee 2013; Ozer et al. 2004). Recent research indicates that approximately one-third of high school students report being the victim of relationship violence at some point in high school (O’Leary et al. 2008). Understanding why aggressive behavior and dating behaviors are associated with one another is thus an important research endeavor with potential implications for adolescent physical and emotional health.

Accordingly, various hypotheses have been presented to explain why aggression and dating are correlated. Sexual selection theory (Darwin 1871), for example, suggests that girls will select partners who exhibit signs of dominance among peers, because dominance indicates potential resource-holding power as well as high status among peers. Thus, boys who use overt aggression or assertiveness may be particularly good at securing the attention and interest of potential dates (Pelligrini and Long 2003). On the other hand, girls compete for dates by maximizing their value to boys while at the same time minimizing other girls’ value to those same boys—often via strategies that include gossip, rumors, and other forms of relational aggression that allow them to harm others without the threat of direct physical harm to the self (Arnocky and Vaillancourt 2012; Pelligrini and Long 2003). Thus, girls who are most successful at winning potential dates are likely to be high in relational aggression as well. Furthermore, adolescent girls and young women who are attractive report more relational victimization than less attractive peers do (Arnocky et al. 2012; Leenaars et al. 2008), suggesting that relational aggression is often used strategically to derogate potentially strong romantic rivals (Fisher and Cox 2009). Resource Control Theory offers a similarly useful perspective from which to understand the link between aggression and dating popularity (Hawley 1999). Resource Control Theory focuses on the functional value of behaviors that facilitate the acquisition and control of resources; resources can include material things like wealth or toys, but also intangible social goods like attention from peers or dating opportunities. From a Resource Control Theory perspective, then, aggression can be viewed as a strategy that functions to increase dating popularity via the control of social resources relevant to dating.

Current Goals and Hypotheses

Both peer status and aggression have been shown to predict dating status or dating popularity among adolescents. A growing literature documents these associations and highlights the need for greater understanding of their influence on one another over time. However, research in this area remains limited and the findings inconsistent, particularly in investigating the association between romantic development and popularity, and between romantic development and overt aggression. Further, among the small number of studies addressing the links between peer status, aggression, and dating, a large proportion focuses on children in middle school. Romantic involvement, and dyadic dating relationships in particular, intensify across adolescence, particularly after the transition to high school (Carver et al. 2003; Connolly et al. 1999). Finally, only one study has investigated dating in relation to both peer status and aggression in the context of one study, so our understanding of the interaction of those constructs in predicting romantic development is limited (see Arnocky and Vaillancourt 2012). Given the strong relationship between aggression and popularity in adolescence (Mayeux et al. 2011), as well as the link between peer-directed aggression and dating partner aggression (Connolly et al. 2000a, b), understanding the interaction of popularity and aggression in predicting dating has important health-risk implications.

Thus, the current study builds on this existing literature in three ways. First, we examine the associations of dating popularity with two forms of aggression (overt and relational) and two forms of peer status (social preference and popularity) in the context of one study of ninth-graders, which allows for a better understanding of the unique contributions of each to adolescents’ romantic development at a time when their romantic involvement is intensifying. Second, we investigate the interactions of peer status and aggression in predicting dating popularity, illuminating specific patterns of status and aggression that are most beneficial in terms of controlling romantic resources in the peer group. Third, we address the question of whether these patterns differ significantly by gender, which allows us to test gender-based predictions related to romantic strategies in adolescence. Prior research in this area has found that patterns of association between dating and aggression in particular vary by gender (Pellegrini and Bartini 2001; Pelligrini and Long 2003). However, investigation into whether the interaction of peer status and aggression in the prediction of romantic development is moderated by gender has not been undertaken. This project was designed to address these limitations.

The current study had five specific goals and related hypotheses. We conceptualized dating popularity as being nominated frequently by peers as someone that they would like to go on a date with. First, we investigated the associations of social preference, popularity, and overt and relational aggression with dating popularity in a large sample of ninth-graders. We hypothesized that all four constructs would be positively correlated with dating popularity. Second, we investigated gender as a moderator of these associations. We expected that overt aggression would be more strongly related to dating popularity for boys than for girls, and that relational aggression would be more strongly related to dating popularity for girls than for boys (Pellegrini and Bartini 2001; Pelligrini and Long 2003). Evidence from recent studies suggests that girls consider peer status more consistently than boys do when evaluating a potential romantic partner (Ha et al. 2010), so we anticipated a stronger association between peer popularity and dating popularity for boys than for girls.

Third, we investigated overt and relational aggression as moderators of the associations between peer status and dating popularity. We anticipated that, at relatively high levels of both forms of aggression, there would be a positive association between peer popularity and dating popularity. The combination of social power and aggressive behavior was expected to be particularly effective in attracting potential romantic partners. However, we hypothesized that dating popularity would be relatively low for low-status youths regardless of their level of aggression.

Fourth, we tested the interaction of social preference and peer popularity as a predictor of dating popularity. Previous research has shown evidence of considerable heterogeneity among popular youths: some are characterized by higher levels of antisocial behaviors and other negative qualities, while others are primarily prosocial (de Bruyn and Cillessen 2006; Rodkin et al. 2000). Not surprisingly, these groups differ in their levels of liking by peers (Rodkin et al. 2000). While the focus of our research was not to identify subgroups, we investigated whether popular adolescents who are also relatively high in social preference were higher in dating popularity than popular adolescents who were less socially preferred. We anticipated positive associations between dating popularity and peer popularity at both high and low levels of social preference, but we expected the magnitude of the association to be stronger at higher levels of social preference.

Fifth and finally, we tested whether gender further moderated the two-way interactions described above. Previous studies of aggression in relation to dating popularity have found differing patterns for boys and girls, with overt aggression and assertiveness being more relevant for boys, and relational aggression being more relevant for girls (Pellegrini and Bartini 2001; Pelligrini and Long 2003). Thus, we anticipated similar patterns in our own findings. We hypothesized that the moderating effect of peer status on the association between relational aggression and dating popularity would be especially strong for girls, and that the moderating effect of peer status on the association between overt aggression and dating popularity would be especially strong for boys.

Method

Participants

Participants were 215 ninth-graders (50 % girls; 46 % of the students in the grade) who attended a large high school located in the Midwestern United States. Participants were predominately white and were recruited from a school that serves both suburban and rural, working- to middle-class families. While we did not request participants’ birthdates, the ninth grade year in this school system corresponds to approximately ages 14–15. Students in the school attended class primarily with students in their same grade, changing classes frequently throughout the day. Because all data for this study were derived from peer nominations, data from the entire grade (478 students) were used.

Procedure

Signed parental consent and child assent were obtained before data were collected. Letters were distributed via students’ English classes in April of the school year. One English teacher neglected to send consent forms home with her students. Thus, approximately 80 % of the students in the ninth grade took home a consent letter describing the research and data collection procedures. Parents who wished for their child to participate returned the signed consent form to the school. Two hundred fifteen students (50 % girls) returned a signed consent form and provided peer nomination data for the study. This sample comprised 56 % of the total number of students who took home a consent form, and 46 % of the 478 students in the school grade. While this number is lower than the consent rates for most published studies using sociometric methods, the use of unlimited nomination procedures and, in most cases, multiple items to measure each construct maximize data reliability (Marks et al. 2013; Terry 2000). Students who provided data were higher in social preference than students who did not provide data [t(476) = 3.27, p = .001], but there were no differences in popularity, aggression, or dating popularity between the two groups.

All constructs were measured via peer nomination. To facilitate the process, a roster of the names of all the students in the ninth grade was given to each participant. Each name on the roster was associated with a unique code number that participants were asked to use when nominating that student. Unlimited same- and cross-sex nominations were allowed. Self-nominations were discouraged and were omitted from data analysis. All students in the grade, rather than only the students with permission to participate, were listed on the roster in accordance with university Institutional Review Board guidelines. Listing all students protects the privacy of participants by preventing the identification of which students participated in the study and which did not. All students in participating classrooms received a booklet on the day of testing. Participating students received a survey booklet, while nonparticipating students received a booklet of games and puzzles to solve.

Measures

Social Preference

Social preference was measured using two items: “Who in your grade do you like the most?” and “Who in your grade do you like the least?” Nominations received for each item were counted for each student and then standardized within grade to a z-score with a mean of 0 and a standard deviation of 1, yielding a score that reflected students’ relative level of liking, disliking, etc., compared to their peers. To create the composite social preference score, the standardized liked least score was subtracted from the standardized liked most score. The resulting number was then restandardized.

Popularity

Popularity was also measured using two items: “Who in your grade is the most popular?” and “Who in your grade is the least popular?” Similar to the calculation of social preference, the composite popularity score was created by subtracting the standardized score for least popular from the standardized score for most popular, and restandardizing the resulting score.

Relational Aggression

Three items were used to measure relational aggression (“Who excludes others from the peer group?”, “Who spreads rumors or gossip about some peers?” and “Who ignores others in order to get their way?”; intercorrelations were r = .58–.69; Cronbach’s α = .86). The number of nominations for each item was first standardized, and then the mean of the three scores was computed and used in all analyses.

Overt Aggression

Three items were also used to measure overt aggression (“Who initiates or gets into physical fights with peers?”, “Who hits, shoves or pushes peers?” and “Who tries to dominate or bully people?”; intercorrelations were r = .84–.91; Cronbach’s α = .96). As with relational aggression, the number of nominations for each item was first standardized, and then the mean of the three scores was computed and used in all analyses.

Dating Popularity

One item was used to measure dating popularity: “Who are the students in your grade you would most like to go on a date with?” This score was standardized within grade as well.

Results

Associations of Peer Status and Aggression with Dating Popularity

Bivariate correlations were first obtained between all of the measured constructs (social preference, peer popularity, relational aggression, overt aggression, and dating popularity), separately by gender. These results are presented in Table 1. Dating popularity was positively correlated with social preference and with peer popularity. For both genders, dating popularity was positively correlated with relational aggression, but dating popularity was only correlated with overt aggression for boys. Fisher’s r-to-z transformations indicated that several correlations differed by gender. The positive association between overt and relational aggression was stronger for boys than for girls (rs = .67 and .28 for boys and girls, respectively; p < .001), as was the positive association between dating popularity and social preference (rs = .31 and .15 for boys and girls, respectively; p = .03). The negative association between relational aggression and social preference was stronger for girls than for boys (r = −.39 for girls vs −.24 for boys; p = .04). Finally, popularity and overt aggression were significantly related for boys only (rs = .34 and .02 for boys and girls, respectively; p < .001).

Table 1 Intercorrelations among dating popularity, peer status, and aggression
Full size table

Moderating Effects of Peer Status and Gender

To address the remaining research questions, a hierarchical regression analysis was conducted using dating popularity as the dependent variable. In the first step gender (dummy coded; boys = 0, girls = 1), social preference, peer popularity, relational aggression, and overt aggression were entered. In the next step, all of the possible two-way interactions of the study variables, except the interaction between overt aggression and relational aggression, were entered. The third step included key three-way interactions of gender, peer status, and aggression (see Table 2). All variables in the regression were standard scores, so no centering was needed prior to analysis. Significant interaction terms were plotted in the manner described by Aiken and West (1991) by creating a “high” score one standard deviation above the mean, and a “low” score one standard deviation below the mean.

Table 2 Summary of hierarchical regression analysis predicting dating popularity
Full size table

A total of 43 % of the variance in peer-nominated dating popularity was explained by the model tested. Significant main effects were found for social preference, β = .19, t(472) = 4.03, p < .001, peer popularity, β = .36, t(472) = 7.18, p < .001, and relational aggression, β = .24, t(472) = 4.77, p < .001. Adolescents who were more well-liked, more popular, or higher in relational aggression received higher scores for dating popularity.

Gender moderated the association between peer popularity and dating popularity, β = .20, t(463) = 3.08, p = .002 (see Fig. 1). Highly popular girls were found to be higher in dating popularity than highly popular boys (simple slopes test, t = 3.96, p < .001), but there was no gender difference found for those low in popularity (simple slopes test, t = −.75, p = .46).

Fig. 1
figure 1

Gender moderates the association between peer popularity and dating popularity

Full size image

Both overt aggression, β = −.15, t(463) = −2.71, p = .007, and relational aggression, β = .15, t(463) = 2.46, p = .01, moderated the association between social preference and dating popularity. For adolescents who were high in relational aggression, being well-liked was associated with being high in dating popularity (simple slopes test, t = 4.29, p < .001; see Fig. 2). However, for those low in relational aggression, there was no association between social preference and dating popularity (simple slopes test, t = 1.63, p = .10).

Fig. 2
figure 2

Relational aggression moderates the association between social preference and dating popularity

Full size image

The opposite was true for the case of overt aggression (see Fig. 3). At low levels of overt aggression, adolescents were rated higher in dating popularity when they were also well-liked (simple slopes test, t = 3.87, p < .001), but at high levels of overt aggression, social preference was not linked to their level of dating popularity (simple slopes test, t = 1.77, p = .08).

Fig. 3
figure 3

Overt aggression moderates the association between social preference and dating popularity

Full size image

Popularity moderated the association between overt aggression and dating popularity, β = .15, t(463) = 2.05, p = .04. This interaction was further moderated by gender, β = .12, t(456) = 2.28, p = .02 (see Fig. 4). For popular girls, overt aggression was associated with increased dating popularity. For unpopular girls, the association between overt aggression and dating popularity was negative. Popular, overtly aggressive girls were particularly high in dating popularity, while unpopular overtly aggressive girls were very low in dating popularity. The association between overt aggression and dating popularity was nonsignificant for both popular and unpopular boys.

Fig. 4
figure 4

Three-way interaction of overt aggression, popularity, and gender predicting dating popularity

Full size image

Popularity moderated the association between social preference and dating popularity, β = .31, t(463) = 7.02, p < .001, and this was further moderated by gender, β = .19, t(456) = 2.60, p = .01 (see Fig. 5). For both genders, the positive association between popularity and dating popularity was stronger for those were also well-accepted. This relationship was especially strong for well-liked girls. Unpopular, well-liked girls were particularly low on dating popularity. Finally, the results indicated that being popular, for girls, was a protective factor against the negative effects of being poorly accepted on dating popularity.

Fig. 5
figure 5

Three-way interaction of popularity, social preference, and gender predicting dating popularity

Full size image

Discussion

The current study investigated associations of two forms of peer status and two forms of aggression with dating popularity in a sample of ninth-graders. Using Brown’s developmental–contextual model of adolescent romantic relationships (Brown 1999) and Resource Control Theory (Hawley 1999) as guiding frameworks, we anticipated positive associations between both forms of peer status and dating popularity, and between both forms of aggression and dating popularity. This hypothesis was supported except in the case of overt aggression, which showed no main effect in the prediction of dating popularity. However, our hypothesis that gender would moderate these associations was largely unsupported by the data. Gender moderated only the relation of dating popularity with peer popularity, with the relation being particularly strong for girls (but still positive for boys). Popularity is indeed seen as a desirable attribute in potential dating partners among ninth-graders. Popular girls were seen as especially desirable dates by boys. Popularity is more closely associated with other desirable attributes, such as facial attractiveness or physical maturity, in girls than in boys (Borch et al. 2011; Krantz 1987; Krantz et al. 1985). It may also be true that the peer-valued characteristics that are associated with popularity—things like material wealth and being fashionable (Vaillancourt and Hymel 2006)—are more important for boys’ romantic interests than for girls’, leading to popular girls’ higher levels of attractiveness as potential dates.

Peer Status, Aggression, and Dating Popularity

At the bivariate level, overt aggression was positively related to dating popularity for boys, and the association was modest. However, when important moderating factors such as social preference and peer popularity were also considered, interesting patterns of association emerged. For example, for adolescents who were high in social preference, overt aggression predicted lower dating popularity. Socially preferred youths who used overt aggression were not nominated by others as desirable potential dates. Their likeability may stem from a variety of other positive traits, but their high level of overt aggression made them undesirable as dating partners. This finding is consistent with Arnocky and Vaillancourt’s (2012) longitudinal study, which also found overt aggression to be problematic for boys’ dating opportunities over time. Socially preferred youths who were low in overt aggression were seen as relatively desirable dates. These adolescents’ low levels of overt aggression, in combination with other factors that are associated with their likeability, may have made them stand out as desirable romantic partners.

Peer popularity also moderated the association between overt aggression and dating popularity, and this relationship was further moderated by gender. For boys, overt aggression and dating popularity were unrelated regardless of their level of popularity. However, overt aggression was associated with dating popularity for girls at both high and low levels of peer popularity. Girls who used low levels of overt aggression had similar (low-to-average) dating popularity regardless of their level of popularity. For popular girls, increases in overt aggression were associated with increases in dating popularity. For unpopular girls, overt aggression was linked to lower dating popularity. Boys viewed popular, aggressive female peers are desirable dates. Aggressive, unpopular girls were very rarely nominated as potential dates. Previous investigations have typically found relational aggression to predict dating popularity for girls (Pellegrini and Bartini 2001; Pelligrini and Long 2003), but a small number of studies has also found overt aggression to predict dating popularity for girls (e.g., Miller et al. 2009). However, these studies did not account for the role of peer popularity or other measures of social prominence. From the perspective of Resource Control Theory, for popular girls who are already seen as desirable dates, the use of overt aggression further enables them to control social resources related to romantic relationships, such as attention from the opposite sex. How exactly this plays out in the peer group remains to be explored. It is possible that popular girls’ acts of dominance and bullying make them stand out to peers in ways that also increases the visibility of other more attractive qualities, like assertiveness and confidence, that are attractive to potential partners. Aggressive behaviors enacted by popular girls may also differ qualitatively from the same kinds of behaviors enacted by lower-status peers. For example, their aggression may be more proactive in nature, and appear to others as self-possessed and in control. Lower-status girls may aggress in a more reactive way, potentially in situations in which they feel marginalized and highly aware of their low status in the peer group.

Relational aggression was positively associated with dating popularity, consistent with our hypotheses and with previous findings (e.g., Arnocky and Vaillancourt 2012; Pellegrini and Bartini 2001; Pelligrini and Long 2003). However, there were no gender differences in this association, indicating that both boys and girls who used relational aggression were seen as desirable romantic partners. From the perspective of Resource Control Theory, the use of relational aggression enables adolescents to control social resources related to romantic relationships, such as attention from the opposite sex and the reputations of romantic rivals (Arnocky et al. 2012; Leenaars et al. 2008). However, unlike during early adolescence, in which gendered patterns of aggression in dating relationships have been documented (e.g., Pellegrini and Bartini 2001), slightly older adolescents in high school may control dating resources by using a wider variety of aggressive and manipulative behaviors that cross stereotypical gender lines.

Contrary to our predictions, social preference moderated the link between relational aggression and dating popularity, while peer popularity did not. Low social preference was associated with low dating popularity regardless of the level of relational aggression. Disliked peers are likely not seen as eligible romantic partners under most circumstances, particularly if they are aggressive. Surprisingly, social preference and dating popularity were more closely related at high levels of relational aggression than low. Perhaps socially preferred adolescents use relational aggression in a way that is qualitatively different than their peers—less maliciously, or coupled with other more prosocial or pleasant qualities that buffer the potential ill effects of the aggression (e.g., Hawley 2003).

The Attraction of Being Both Well-Liked and Popular

Our final research goal involved investigating the interaction of social preference and peer popularity in the prediction of dating popularity. We found a positive association between peer popularity and dating popularity at all levels of social preference for both boys and girls, providing further evidence of the importance of peer popularity for adolescents’ romantic attractiveness to peers (Brown 1999). Popular, well-liked girls were seen by boys as particularly desirable dates. The combination of peer popularity and high levels of social preference is likely an indicator of many positive qualities, including sociability, confidence, and prosocial behavior—the “best of both worlds,” socially speaking. Well-accepted, popular youths have been shown to have better relationships with peers than either popular-only or preferred-only youths, and they have high social self-concepts and other attractive personal attributes such as kindness and trustworthiness (Košir and Pecjak 2005; Parkhurst and Hopmeyer 1998). Furthermore, popular adolescents who are also well-accepted are described by their peers as dominant, but not as aggressive (Parkhurst and Hopmeyer 1998). Thus, popular, well-liked adolescents exhibit the status-enhancing features associated with popularity, but do not use aggressive behaviors to the same extent as popular peers who are less liked. Further, their dominance is expressed in the context of other more prosocial and socially competent behaviors and attributes (see also Hawley 2003). This combination of popularity and status, low aggression, and positive interpersonal qualities is likely to be very attractive to peers.

Somewhat surprisingly, girls who were high in social preference but low in peer popularity were especially low in dating popularity. Perhaps, as Brown proposed as part of his developmental–contextual theory of romantic relationships (1999), adolescents view potential partners through the lens of peer status management. Well-liked peers may be valued for their positive, relationship-enhancing qualities in the context of friendships, but without the boost of peer popularity, they are not likely to be seen as status-enhancing for their potential romantic partners. They may be viewed as excellent “friendship material,” but when it comes to dating, their peers might gravitate toward likeable dates who bring enhanced status to the table as well. Still, it is unclear why these dynamics lead to such extreme outcomes for girls in particular. Well-liked, non-popular boys were also relatively low in dating popularity, but their dating popularity score was approximately one standard deviation higher than that of well-liked, non-popular girls. It is possible that many of the same characteristics that make an adolescent boy likeable—perhaps qualities such as being funny or confident—also make them attractive as dating partners, regardless of their level of popularity.

The Meaning of Dating Popularity

The extant literature on adolescent dating is comprised of studies that use a variety of conceptualizations of dating and romantic development. Some have measured actual dating status (Carlson and Rose 2007), while others have measured dating behaviors such as hanging out alone with opposite-sex peers (Pellegrini and Bartini 2001). As a measure of dating popularity, this study relied on nominations of peers with whom participants would most like to go on a date. We provided no further qualifications for the nomination item, and intended it to tap into adolescents’ perceptions of peers who were seen as desirable potential partners in either a short-term or long-term relationship. Given that most White adolescents in the United States have not initiated sexual relationships by the end of ninth grade (Zimmer-Gembeck and Helfand 2008), it is unlikely that participants interpreted the nomination as referring to strictly sexual relationships. We did not collect data on actual dating behavior or dating history, which would have served as interesting comparisons to our dating popularity construct. However, we felt it was important to measure adolescents’ perceptions of desired dating partners, rather than the peers they had actually dated, in order to best understand that roles of aggression and peer status in adolescents’ views on what makes a valuable partner. Further, some ninth-graders have not yet begun to date (Zimmer-Gembeck et al. 2004).

Limitations

Our findings should be interpreted in light of some limitations. For example, our data were collected at one high school, rather than multiple school sites, leading to interpretations that may not be generalizable to all populations. The sample was ethnically homogeneous, and there was relatively little socioeconomic diversity. In addition, the participation rate was less than ideal for studies using peer-nomination data. Obtaining the perspective of a greater percentage of the reference group would have improved the reliability of the measured constructs. However, with the exception of dating popularity, our constructs were measured with multiple items (Marks et al. 2013), which increases their reliability. Dating popularity was indeed measured with only one peer nomination item (“students in your grade that you would most like to go on a date with”), which may have been interpreted in different ways by different participants. Finally, we were not able to include important control variables, such as physical attractiveness and physical maturity or pubertal development, which are strong predictors of attractiveness to peers (Ha et al. 2010; Zimmer-Gembeck and Helfand 2008). Regarding this last point, we expect that the patterns we found in our data would remain significant after controlling for physical appearance. We also suspect that physical appearance variables would have acted as important moderators of our findings, such that adolescents who were popular, aggressive, and physically attractive would have been seen as exceptionally valuable potential dating partners (Borch et al. 2011).

Conclusions

Recent research on adolescent romantic relationships has dispelled a number of myths and misconceptions about them, as well as highlighting concerning patterns in teen dating. For example, adolescent romances are not always short-lived and unimportant (Collins 2003), and highly aggressive youths engage in even more frequent dating activities than their non-aggressive peers (Connolly et al. 2000a, b). Understanding the development, course, and processes of romantic engagement in adolescence is a crucial task, as it relates directly to mental and physical health outcomes (Ozer et al. 2004). For example, physical aggression against peers has been shown to be a strong predictor of sexual violence against a dating partner (McNaughton Reyes and Foshee 2013). Although adolescent sexual dating violence is less prevalent than overt or relational dating aggression (estimates range from 3 to 10 % for sexual violence, compared to 30 % for physical aggression; Foshee and Matthew 2007; O’Leary et al. 2008), it has serious implications for negative physical and mental health outcomes. Furthermore, nonviolent precursors to sexual aggression (such as spreading sexual rumors or making sexual comments to others) are observed even among youths in middle school (Espelage et al. 2012). Clearly, there is a need for researchers to continue to investigate the development of both healthy and unhealthy romantic development among children and teens.

Our findings build on this growing literature by further documenting the complex dynamics of adolescent romantic development. We tested the hypothesis that dating popularity would be positively linked to both peer status and aggressive behavior. Further, we focused on the interaction of aggression with peer status in predicting dating popularity, with attention to differing patterns by gender. Consistent with previous research (Carlson and Rose 2007; Pellegrini and Bartini 2001), our results suggest that adolescents’ most highly-desired dating partners are often characterized by high peer popularity and relational aggression. Importantly, we found relational aggression to predict dating popularity for boys as well as girls. These findings provide support for Resource Control Theory (Hawley 1999), in that adolescent aggression functions to increase access to social resources such as dating partners. We also found support for the developmental–contextual theory of romantic development (Brown 1999), in that popularity emerged as a characteristic of desired partners.

We also found important moderating effects of aggression in the association between both social preference and peer popularity and dating popularity, which have not been documented previously. Some of these findings were further moderated by gender. For example, social preference predicted dating popularity for teens who were high in relational aggression, but also for teens who were low in overt aggression. High levels of overt aggression were associated with high dating popularity for girls who were popular, but overt aggression predicted low dating popularity for girls who were unpopular with peers. In evaluating the desirability of peers as dating partners, adolescents seem to view aggression through the lens of peer status. Aggressive behavior enacted by unpopular peers may be seen as particularly problematic, while aggressive behavior by popular peers is overlooked, or even seen as desirable (Mayeux et al. 2011). According to our findings, boys in particular attach considerable weight to the popularity of potential female partners. As we learn more about the complexity of adolescent romantic development, questions about the roles of status, aggression, and gender provide exciting avenues for future research.