Being in the Know

Abstract

We tested the hypothesis that early adolescents’ (N = 951) knowledge of who bullies whom is differentially distributed across peers. Knowledge of bullying was assessed by asking middle school students to nominate grademates who bullied or were bullied using physical or relational means. We hypothesized that peers who were closer to bullies and victims in the grade-level friendship network, such as bullies, victims, their friends, and socially central youth, would be more knowledgeable than would peers who were socially distant. Findings revealed that approximately half of the adolescents were able to identify bullies and victims, and those who were most able to do so were socially closer, particularly victims and their friends. Bullies and their friends were less likely to name peers as bullies or victims.

Keywords

bullying victimization peer relationships networks knowledge

For many youth, middle school represents a tremendous social transition that is characterized by a reorganization of social roles and a corresponding increase in interpersonal problems, such as bullying. Observational studies conducted in elementary schools have shown that bullying takes place as often as every 5 to 15 minutes when children gather in unstructured settings (Craig & Pepler, 1998; Craig, Pepler, & Atlas, 2000; Pepler & Craig, 1995). Comparable estimates of the observed frequency of middle school bullying have not been established, but studies using self-report and peer-report methodologies suggest that rates of bullying are even higher at this age, particularly just after the middle school transition (Nansel et al., 2001; Pellegrini & Long, 2002). Approximately 30% of adolescents report moderate or frequent involvement in bullying as a bully and/or victim, and approximately 60% to 70% of adolescents report witnessing bullying (Aboud & Miller, 2007; Bradshaw, Sawyer, & O’Brennan, 2007; Cuadrado-Gordillo, 2012; Nansel et al., 2001; Rivers, Poteat, Noret, & Ashurts, 2009). Furthermore, middle school students and school personnel indicate relatively higher levels of concern about bullying compared with students and educational staff at other developmental levels (Bradshaw et al., 2007).

These findings could be interpreted to imply that bullying in middle school is rampant and that students are acutely aware of bullying incidents and of who is bullying whom. Yet, such an interpretation seems to contradict findings that hint that young adolescents may not be as knowledgeable about bullies and their victims as one might think (Craig & Pepler, 1998; Parault, Davis, & Pellegrini, 2007). Such knowledge requires the opportunity to become informed about a bullying event (by witnessing it or hearing about it second hand) as well as the necessary motivation to remember the details of the event (such as who bullied whom and how). Despite the fact that rates of bullying are relatively high in middle school, bullying behaviors still occur less frequently than non-bullying behaviors (e.g., such as solitary activity or social, non-aggressive peer activity; Parault et al., 2007). And, when a bullying event occurs, it is typically brief in duration (Craig & Pepler, 1998). Thus, students are less likely to witness bullying interactions than they are to witness non-bullying peer interactions. Moreover, young adolescents may ignore or misinterpret bullying behaviors as non-bullying behaviors, which may decrease their overall knowledge of the specifics of bullying (Terasahjo & Salmivalli, 2003).

These seemingly contradictory findings lead to important questions: In the complex and dynamic world of middle school peer interactions, how knowledgeable are students of the details of bullying episodes and who is most likely to know? With knowledge comes the ability to retrieve the information and use it, such as in identifying bullies and victims, understanding how and why bullying occurs, or intervening effectively to change the course of harmful social interactions and protect victims (Wyer, 1980). As Cross, Parker, Prusak, and Borgatti (2001) have noted, knowledge, and the sharing of knowledge, is crucial to solving problems. Herein lies the significance of this issue. Peer interventions in bullying interactions and relationships can be effective in reducing or stopping bullying (Kärnä et al., 2011). Yet, for peers to be able to effectively intervene, there must be sufficient knowledge of who is bullying whom and how the bullying is occurring. The purpose of the present study was to assess the degree to which students can identify those who engage in physical and relational bullying and their victims and to consider how peers’ social positions in the grade-level friendship network (e.g., social distance and social centrality, status as bullies and victims, and friendships with bullies and victims) contribute to knowledge about bullies and victims. In other words, in middle school, who is “in the know” about bullying?

Being in the Know

To what extent are youth knowledgeable of the details of bullying? Almost no research has directly addressed this issue. Surveys have shown that 60% to 70% of middle school students report having witnessed bullying, but such surveys typically ask students to respond to general questions about whether one has been exposed to bullying without providing an opportunity for respondents to detail the extent of their knowledge (Aboud & Miller, 2007; Bradshaw et al., 2007; Rivers et al., 2009). Peer nominations, which tap social knowledge by asking peers to identify bullies or victims, are typically aggregated to produce reputational estimates for each classmate and, thus, are used as indicators of prevalence of a behavior among nominated peers rather than as indicators of the knowledge of nominating peers. As such, questions about individual differences in peers’ knowledge of bullies and victims are often left unanswered. Yet, there are reasons to believe that there is variability in knowledge of the details of bullying. For instance, Card, Rodkin, and Garandeau (2010) recently considered the triadic relations among peers and the classmates whom they nominated as bullies and victims. The findings of this study showed that there is variability across peers in their perceptions of who is (or is not) aggressive and who participates in aggressor/victim dyadic relationships. Similarly, studies of non-reciprocated friendship nominations and differences among self-, peer, and teacher reports of social constructs indicate that any two individuals may have different perceptions of the same phenomenon (Adams, Bukowski, & Bagwell, 2005; Gest, 2006; Juvonen, Nishina, & Graham, 2001; Newcomb & Bagwell, 1995). Moreover, sociometric studies have suggested that characteristics of peer nominators predict the nominations given (i.e., who is nominated for what; LaFontana & Cillessen, 1999; Rock, Cole, Houshyar, Lythcott, & Prinstein, 2011; Rodkin, Farmer, Pearl, & Van Acker, 2006). This variability among peers in their perceptions of social situations requires greater consideration.

Within a social group (e.g., all sixth graders in a school), some individuals are likely to have more knowledge of social events than others. The more frequently one is presented with information about bullying, the more recently one has been made aware of bullying, the more relevant information about bullying is to one’s own interests, or the more vivid and prototypical one’s information about bullying, the more knowledge of bullying one should have (Greenhoot, Tsethlikai, & Wagoner, 2006; Tversky & Kahneman, 1973; Wyer, 1980). Knowledge of social phenomena is gleaned from past experiences and can come from participating in or witnessing bullying interactions, as well as by hearing about them from those who were present during the original interactions.

Information that is obtained from direct involvement in bullying is likely to be more detailed than information obtained from secondhand sources, but it may be more difficult to come by. That is, the likelihood that any particular individual will be exposed to any particular bullying incident is quite small (Craig & Pepler, 1998). There are many reasons for this. Bullying interactions occur less often than non-bullying interactions, even though bullying interactions tend to be repeated over time (Parault et al., 2007). Moreover, any individual act of bullying is typically short-lived, often lasting for less than 1 minute (Craig & Pepler, 1998; Craig et al., 2000; Pepler & Craig, 1995). In addition, bullying often occurs in locations that are unstructured and bustling with an activity that detracts attention away from a bullying interaction (e.g., lunchroom, hallway), making it easy for bystanders to overlook a bullying incident (Bradshaw et al., 2007; Parault et al., 2007).

Different forms of aggression may engender more or less knowledge due to the extent to which they are readily noticed by peers. On the one hand, physical bullying is overt, involving behaviors such as poking, hitting, and shoving. Thus, physical bullying may easily garner visual and auditory attention by its distinctiveness. Yet, physical bullying is also considered a serious offense in schools. Thus, when it does occur, students may intentionally try to hide it to avoid sanctions from school authorities. On the other hand, subtle and covert bullying behaviors, such as those that characterize relational bullying, can be even easier to hide. Even so, young adolescents are astute observers of their social world, watching the social interactions going on around them and taking note of the social news that comes through the peer network. Moreover, relationally aggressive behaviors are important to middle schoolers. They are often used in the negotiation of status positions in the social hierarchy, making them important behaviors to attend to (Hawley, 1999). In fact, those who use relationally aggressive behaviors tend to also be socially prominent in the peer group (Cillessen & Mayeux, 2004). Thus, middle schoolers may be motivated to attend to relationally aggressive behaviors and, thus, have greater knowledge of these relative to physically aggressive behaviors.

Who Is in the Know

Knowledge about bullying is unlikely to be equally distributed across all peers. In a typical U.S. middle school, in which there can be hundreds of peers and the relevant peer reference group is the entire grade (rather than one’s class, as is often true during the elementary years), it is likely that some students would have more knowledge of bullying than others. With this in mind, one’s social distance to a victim or perpetrator of bullying should be predictive of knowledge about bullying behavior. Social distance includes both direct and indirect relationship ties, such that students who have direct ties to victims or bullies in the grade-level social network would have low social distance and those who have indirect ties would have higher social distance. Social distance is related to one’s construal of events, in that the closer one is, the more concrete and detailed one’s mental representations are; in contrast, the farther one is, the more abstract and simplistic one’s mental representations are (Trope & Liberman, 2003). In other words, students who are closer to bullies and victims should have more information about who bullies whom than students who are distant from bullying, bullies and victims, and bullying episodes (see also Card et al., 2010).

Bullies and Victims

Those who have direct experience with bullying—as a bully or a victim—should have relatively high knowledge. As Wyer (1980) noted, social knowledge is related to the degree to which an event is personally relevant. Youth who engage in aggressive and bullying behavior and youth who are victimized by their peers are intimately involved in, and can be deeply affected by, bullying. Thus, they should have knowledge of who bullies and who is victimized (Greenhoot et al., 2006). Because knowledge is more important to individuals who have low levels of control over their social environment and who feel that the information is valuable (Swann, Stephenson, & Pittman, 1981), knowledge might be even greater for victims than for bullies. Victims have a need to protect themselves by, for example, avoiding those who are most likely to bully; creating bonds with other non-victimized youth for friendship, protection, and support; or even by distancing oneself from other victims (Pellegrini, Bartini, & Brooks, 1999; Scholte et al., 2009). Thus, knowing who bullies and who is victimized is information that has high utility for victims.

Friends and Socially Central Peers

Students who are not themselves bullies or victims may also have high levels of knowledge of who bullies and who is victimized. As Cohen, Hsueh, Russell, and Ray (2006) have argued, aggression is a multi-level social phenomenon impacting not only the bullies and victims themselves but also others in the peer group. Peers may come to know about bullying by witnessing it or by hearing about it from others. Witnessing bullying requires being in the right place, at the right time, as well as paying attention to the bullying interaction while it is occurring. Similarly, gaining secondhand knowledge of bullying events requires having the right connections with peers who are likely to have sufficient information about bullying events to relay the details. In either case, students who are friends with bullies or victims and students who are central in the peer network have more access to peers and information; thus, they should have an advantage relative to their peers to learn about who bullies whom and about how that bullying occurs (i.e., physical versus relational).

Friends of bullies and victims are likely to be in close proximity, and visually attentive, to bullies and/or victims (e.g., Vaughn, Colvin, Azria, Caya, & Krzysik, 2001). Thus, they should have opportunities to gain knowledge about bullying by directly witnessing bullying episodes. Furthermore, friends of either bullies or victims are likely to be invested in the bullying episodes. They may involve themselves in the moment, perhaps by encouraging and reinforcing the bully (as might bullies’ friends) or defending the victim (as might victims’ friends; Hodges, Boivin, Vitaro, & Bukowski, 1999; Huitsing & Veenstra, 2012). Friends of either bullies or victims may also provide comfort or other forms of social and emotional support following the bullying episode (Holt & Espelage, 2007). Even when bullying episodes occur outside of friends’ purview, they are likely to hear about it. Because middle school friendships are characterized by engagement with one another, intimacy, and loyalty (Newcomb & Bagwell, 1995), bullies and victims likely discuss the episodes with their friends, thereby engendering support and involvement from friends. Thus, friends are likely to be impacted by bullying episodes, and therefore, they should be motivated to attend to the details of it. And, the more bullies (or victims) in one’s friendship network, the more knowledge one should have.

Those who are socially central in the larger peer network should also have knowledge of bullying. Youth who are central members of the peer network are connected to many peers who are themselves connected to many other peers. The more people one knows, the more likely one is to gain social information. For instance, even among young children, who are less cognitively sophisticated than middle schoolers, social knowledge is related to popularity and social participation, such that those who are more embedded in the peer group have greater social knowledge (Krantz, 1982). Moreover, because socially central youth tend to also have relatively well-developed social skills, they are also better than their lower status peers at discriminating others’ behavioral intentions (Dodge, Murphy, & Buchsbaum, 1984). Thus, for both of these reasons, central individuals should have relatively high knowledge of who bullies whom. Indirectly supporting this hypothesis are studies showing that, in early adolescence, the peers who are perceived as popular among their classmates are also those who are likely to defend victims during bullying episodes (Caravita, DiBlasio, & Salmivalli, 2009; Sainio, Veenstra, Huitsing, & Salmivalli, 2010). Before one can defend a victim, one must know who to defend from whom and how to defend him or her.

Present Study

The present study was designed to test the hypothesis that middle schoolers’ knowledge of bullying is differentially distributed, with some students having greater or lesser knowledge of who bullies whom. The first goal was to determine the extent of middle schoolers’ knowledge of bullying. That is, what percentage of students are in the know regarding who is bullying and who is being bullied? We asked this question for physical and relational forms of bullying separately, hypothesizing that students would have greater knowledge of relational bullies and victims as compared with physical bullies and victims.

The next goal was to consider the idea that those youth who are closer to bullies and victims in the grade-level friendship network would be more likely to be “in the know” about bullying than those youth who are more distant. We refer to this as the social distance hypothesis. Building on this idea, we considered whether knowledge of bullies and victims was greater for bullies, victims, their friends, and socially central youth. We refer to this as the social role hypothesis. In all analyses, we considered students’ knowledge of both bullies and victims. Moreover, by testing social distance and social role hypotheses separately for knowledge of physical and relational bullies and victims, we allowed for the possibility that social distance and social roles would differentially predict knowledge regarding each form of aggressive behavior. This was considered to be an exploratory goal. Despite the voluminous literature on physical and relational aggression, little is known about whether and how peers differentially perceive those who engage in these behaviors. Yet, one recent study suggests that middle schoolers who are themselves either relationally or overtly victimized show different patterns in making friendship nominations, speaking to the value of considering the possibility in this study (Sijtsema, Rambaran, & Ojanen, 2013).

These hypotheses were addressed using a middle school sample of sixth-, seventh-, and eighth-grade girls and boys and diverse methodological approaches drawn from the developmental peer research tradition, structural approaches of social network analysis, and social knowledge models. Multiple factors determine “being in the know,” so multiple approaches need to be used to provide an accurate depiction of knowledge of bullying processes.

Method

Participants

Participants were drawn from an ethnically diverse, urban, southwestern middle school that was participating in a 2-year, three wave longitudinal study of intrapersonal and interpersonal processes. The current study utilized data from the second wave (fall of Year 2) because all relevant measures were obtained at this wave. Prior to beginning the data collection, information letters and consent forms were sent home with all students in the school (N = 1,051, sixth-eighth grade students) to inform parents of the intent to collect student survey data. Information letters and consent forms were available in Spanish for Spanish-speaking parents. This study used passive consent procedures, with consent from the school (which acted in loco parentis) and approval from the university Institutional Review Board. Sixty-one of the students’ parents opted out of the study. Four additional students did not provide assent, resulting in 986 students with permission to participate. Of these, 18 students were absent from school during survey administration, and 17 students had withdrawn from the school by the time of survey administration. This resulted in a final sample of 951 participants.

Participants were 339 sixth graders (M age = 11.12 years, SD = .51), 299 seventh graders (M age = 12.12 years, SD = .53), and 313 eighth graders (M age = 13.15 years, SD = .52), with an approximately equal number of boys and girls (49.9% girls). Participants were from ethnically diverse backgrounds, with children self-identifying as Latino (54%), Black/African American (18%), White/European American (14%), Native American (9%), and Other (5%). The majority of adolescents were from low socioeconomic status families (79% were eligible for free lunches, 9% were eligible for reduced-price lunches). Forty-six percent of the participants and their parents were U.S. born, 12% of participants were U.S. born with one parent foreign born, 30% were U.S. born with both parents foreign born, and 12% were foreign born with both parents foreign born. Of those participants who were foreign born, 77% were born in Mexico, 7% were born in Africa, 6% were born in Central or South America, 3% were born in Southeast Asia or Oceania, and the remaining 7% were born in various other countries. Finally, 45% of participants came from two-parent, married families, 33% were from single-parent families, and 16% were from two-parent, unmarried families. The remaining 6% of particpants came from other family configurations.

Procedure

All measures used in the current study were collected in the fall semester of the school year. Participating adolescents completed a questionnaire packet in a classroom setting, in pencil and paper format. Researchers read each item aloud. Individualized assistance was provided as needed to adolescents who had difficulty completing the questionnaires (e.g., students with learning disabilities or language difficulties). The questionnaire packet was administered on 2 consecutive days and took approximately 2 hours to complete. As a token of appreciation for their participation, students were given a pencil and a water bottle with the school logo.

Variables used in the present study were calculated from students’ nominations of peers as bullying perpetrators, victims, and friends. Nominations were obtained by giving students a list of grademates (which contained grademates’ first and last name, as well as a unique identifying number that was created for the study [ID]) and asking them to identify those who fit each description (i.e., bully, victim, friend). Students recorded peers’ first name, last initial, and ID. Participants were told that they could not nominate themselves, but that they could nominate the same person for more than one question. If they could not identify peers for an item, they were instructed to leave the space blank.

Nominations of bullies and victims

Students identified bullies and their victims using a modified version of the “Who Bullies Whom” measure (Ahn, Rodkin, & Gest, 2013). This measure was modified such that separate nominations were made for physical and relational bullies and their victims. Using the list of participating grademates, students were asked to nominate up to three peers who physically bullied others (“hits, kicks, or pushes others”) and up to three peers who relationally bullied others (“gossips about others or excludes others”). For each peer nominated as a bully, participants were asked to identify up to three peers toward whom the nominated person directed their aggressive behavior. For example, the item assessing relational bullying read, “List one person who gossips about others or excludes others,” and the corresponding item assessing relational victimization read, “Who does Person 1 gossip about or exclude the most?” Thus, by linking bully and victim, this measure inherently taps into the relationship between the bully and the victim; this is crucial to the measurement of bullying (Rodkin, Espelage, & Hanish, 2015). Participants answered this question up to 3 times, listing up to three bullies and up to three victims for each bully.

Nominations of friendships

To report on friendships, students again used the list of grademates to identify up to 10 of their closest friends in their grade, starting with their best friend. Students were permitted to nominate any peer within their grade.

Measures

Knowledge of bullies and victims

Often, peer nominations are aggregated across nominators to provide an index of the nominated peers’ reputation as a bully or victim. In this case, however, we used peer nominations to assess the nominators’ own knowledge of bullies and victims. Therefore, four dependent variables were created, which tapped nominators’ knowledge of physical bullies, knowledge of physical victims, knowledge of relational bullies, and knowledge of relational victims. Each of these variables was calculated as a sum of the number of bullies and victims nominated, weighted by the number of times each bully or victim was nominated by other students. The weighting was done to control for variations in reputational prevalence of bullies and victims within the form of bullying (relational or physical). For example, if an individual nominated three peers as physical bullies, one of whom had been nominated as a bully 3 times, one who had been nominated 2 times, and one who had been nominated 5 times, the individual’s weighted score would be calculated as 1(3) + 1(2) + 1(5) = 10. Using this same technique, separate variables were calculated for weighted knowledge of physical victims, relational bullies, and relational victims.

Sum scores, rather than proportion scores (as are commonly used), were calculated because the nominations were obtained at grade level, resulting in very large divisors and, concomitantly, small proportion scores that were difficult to interpret. Because variations in sample sizes across grades were very small, the findings obtained with sum scores were virtually identical to those obtained using proportion scores. Thus, for ease of interpretation, we retained the sum scores. Note, however, that grade was included as a covariate in the models to control for the minor variations in group size across grades.

Social distance

Data based on friendship nominations were arranged in binary matrices for each grade, with each row and column representing an individual within the grade. In the matrix, cell X_ij corresponds to i’s relation to j, as reported by i. That is, if i nominated j as a friend, cell X_ij was coded as 1. If i did not nominate j as a friend, cell X_ij was coded as 0. Social distance was calculated as geodesic distance, which defines distance as the shortest path from one individual to another in the overall grade-level network (Wasserman & Faust, 1994). Geodesic distance was calculated for each pair of actors within the grade-level network, such that a distance value was calculated between each actor j and each actor i (geodesic distance = d[n_j, n_i]).

Bullying and victimization

Adolescents’ own bullying and victimization were calculated as the number of nominations received. Nominations were summed separately for physical bullying, physical victimization, relational bullying, and relational victimization to yield scale scores.

Friends’ average bullying and victimization

As described above, each individual nominated up to 10 friends in their grade. Physical bullying scores for each friend were averaged across all friends to create a measure of friends’ physical bullying. Similar variables were created to represent friends’ physical victimization, friends’ relational bullying, and friends’ relational victimization by averaging friends’ scores on physical victimization, relational bullying, and relational victimization, respectively.

Social centrality

Using the same friendship matrix as described above, social centrality was calculated as Bonacich centrality, which represents an individual’s centrality or power, considering their own position within the peer network as well as their connection to powerful others (Bonacich, 1987; Wasserman & Faust, 1994). This index was calculated by weighting each actor’s centrality by the centrality of those to whom he or she sends ties (Wasserman & Faust, 1994). Where X is the total friendship network, α is a scaling factor (determined mathematically to allow the equation to be solved), β is a power weight reflecting the degree of dependence of an actor’s prestige on the extent of prestige of the peers to whom the actor is connected (which was set equal to .1), Bonacich centrality was computed as X (α, β) = α × (I − [β × X])⁻¹ X₁.

Results

The purpose of this study was to determine the extent to which middle school students are knowledgeable of who bullies whom and which students are in fact knowledgeable. With this in mind, the initial goal was to assess how widespread knowledge is about bullies and victims. We next considered the social distance hypothesis, by comparing dyadic level distances between students who could identify bullies and victims and those who could not. Building on this, the final goal was to explore who is most likely to be in the know about bullying and victimization. To address this goal, regressions were conducted to test whether knowledge of bullies and victims was greater for bullies, victims, their friends, and socially central youth. Across all analyses, we tested knowledge for both physical and relational bullying in separate models, although we controlled for knowledge of the alternate form of bullying in each model.

Descriptive Analyses

As a preliminary step, the primary variables were assessed for skewness and kurtosis; square-root transformations to correct non-normality were needed only for physical and relational bullying measures and for measures of knowledge of physical bullies and knowledge of relational bullies. Means, standard deviations, and ranges for all of these variables prior to the noted transformation are presented in Table 1.

Table 1.

Means, Standard Deviations, and Ranges of Study Variables.

Variable	M	SD	Range
Own physical bullying	0.81	1.88	0.00-26.00
Own physical victimization	1.53	1.76	0.00-12.00
Own relational bullying	1.37	1.98	0.00-20.00
Own relational victimization	2.83	2.95	0.00-23.00
Friends’ physical bullying	0.97	0.86	0.00-6.50
Friends’ physical victimization	1.78	0.98	0.00-7.67
Friends’ relational bullying	1.61	1.03	0.00-7.00
Friends’ relational victimization	3.45	1.70	0.00-13.00
Social centrality	0.94	0.46	0.00-1.89
Knowledge of physical bullies	4.32	7.75	0.00-45.00
Knowledge of physical victims	5.66	10.39	0.00-72.00
Knowledge of relational bullies	5.83	6.75	0.00-38.00
Knowledge of relational victims	17.04	22.15	0.00-111.00

Note. Ns range from 908 to 951. Descriptive statistics for non-transformed variables are shown.

Being in the Know: The Extent to Which Students Can Identify Bullies and Victims

In the present sample, 38.9% of students received at least one nomination for physically bullying others, and 66.5% received at least one nomination for being physically victimized (physical bullying: M = 0.81, SD = 1.88; physical victimization: M = 1.53, SD = 1.76). Similarly, 57.8% of students were nominated by peers as relationally bullying others, and 78.4% were nominated as victims of relational bullying (relational bullying: M = 1.37, SD = 1.98; relational victimization: M = 2.83, SD = 2.95). These rates of bullying are similar to those reported in other studies of middle school samples that report raw ranges of nominations (Nansel et al., 2001). There was overlap across forms of bullying for both perpetration and victimization. Of those nominated as a bully, 39.2% were nominated at least once as both physical and relational bully. Of those nominated as a victim, 66.4% were nominated at least once as both physical and relational victim.

Less than half of the middle schoolers in this study were able to identify peers who were involved in physical bullying. Specifically, 46.9% of students were able to identify at least one peer who perpetrated physical bullying, and 40.2% of the students were able to identify at least one peer who was a victim of physical bullying (knowledge of physical bullies: M = 4.32, SD = 7.75; knowledge of physical victims: M = 5.66, SD = 10.39, using the weighted knowledge variables that account for both number of bullies and victims nominated, as well as how many others nominated those individuals as bullies or victims). In comparison, 65.1% of students were able to identify at least one peer as a perpetrator of relational bullying, and 58.5% of the students were able to identify at least one peer as a victim of relational bullying (knowledge of relational bullies: M = 5.83, SD = 6.75; knowledge of relational victims: M = 17.04, SD = 22.15, again using the weighted knowledge variables). Paired-samples t tests indicated that knowledge of physical victims was significantly greater than knowledge of physical bullies, t(911) = 4.59, p < .001, Cohen’s d = .15, and knowledge of relational victims was significantly greater than knowledge of relational bullies, t(911) = 18.66, p < .001, Cohen’s d = .69 (see Table 1 for means and standard deviations). In addition, knowledge of relational bullies was significantly greater than knowledge of physical bullies, t(911) = 5.25, p < .001, Cohen’s d = .21, and knowledge of relational victims was significantly greater than knowledge of physical victims, t(911) = 17.38, p < .001, Cohen’s d = .66.

Who Is in the Know: The Social Distance Hypothesis

ANOVAs were conducted at the dyadic level to compare the geodesic distance of pairs in which at least one individual nominated the other as a bully or victim with the geodesic distance of pairs in which neither individual nominated the other as a bully or victim. This analysis was conducted 4 times using Welch’s modified t tests, once each for nominations of physical bullies, physical victims, relational bullies, and relational victims. Welch tests were used because they provide a more robust test of the equality of means under violations of assumptions of non-normality and unequal variances than the standard t test (Tomarken & Serlin, 1986; Zimmerman & Zumbo, 1993). Unequal variance is often associated with unequal group size, which was the case in the current analyses (the number of pairs where neither individual nominated the other as a bully or victim was much larger than the number of pairs where one did nominate the other).

Results indicated that social distance in the grade-level friendship network was shorter for dyads in which at least one student nominated the other as a physical bully (M = 2.73, SD = 1.30) compared with dyads in which neither student nominated the other as a physical bully (M = 3.68, SD = 1.14), Welch’s F(1, 763.95) = 407.07, p < .001, Cohen’s d = .83. Similarly, distance was shorter for dyads in which one nominated the other as a physical victim (M = 2.64, SD = 1.36) than dyads where neither nominated the other as a physical victim (M = 3.69, SD = 1.14), Welch’s F(1, 1173.38) = 695.13, p < .001, Cohen’s d = .92; shorter for dyads in which one nominated the other as a relational bully (M = 2.64, SD = 1.28) than dyads where neither nominated the other as a relational bully (M = 3.69, SD = 1.14), Welch’s F(1, 1279.50) = 849.81, p < .001, Cohen’s d = .92; and shorter for dyads in which one nominated the other as a relational victim (M = 2.49, SD = 1.39) than dyads where neither nominated the other as a relational victim (M = 3.69, SD = 1.14), Welch’s F(1, 2165.64) = 1,583.82, p < .001, Cohen’s d = 1.05. Thus, as hypothesized, for both bullying and victimization, students, on average, were closer in terms of social distance to those whom they nominated as bullies and victims than those whom they did not nominate as bullies and victims.

Who Is in the Know: The Social Role Hypothesis

Covariates

Table 2 shows gender, grade, and ethnicity variations in the study variables. For the social role variables, independent-samples t tests revealed gender differences in own and friends’ physical bullying and victimization (favoring boys), own and friends’ relational bullying and victimization (favoring girls), and social centrality (also favoring girls), ts(949) ranged from 3.40 to 13.71, ps < .01. ANOVAs, which were conducted to test for grade and ethnicity differences, indicated that there were grade differences on all social role variables except physical bullying and social centrality, significant Fs (2, 948) ranged from 8.71 to 115.61, ps < .001. Ethnicity effects were considerably less robust in this multi-ethnic sample. In analyses comparing Latino/a, non-Latino White, and Other minority groups, students who identified as Latino/a had friends who physically and relationally bullied and were physically victimized less than students who were identified as being in the Other minority group, Fs(2, 948) = 3.11 and 3.30, ps < .05. In addition, students who identified as Latino/a were higher in social centrality than students who identified as non-Latino White or Other minority, F(2, 948) = 5.98, p < .01.

Table 2.

Means and Standard Deviations of Study Variables by Gender, Grade, and Ethnicity.

	Gender				Grade						Ethnicity
	Girls		Boys		Sixth graders		Seventh graders		Eighth graders		White		Latino/a		Other minority
	M	SD	M	SD	M	SD	M	SD	M	SD	M	SD	M	SD	M	SD
Own physical bullying	0.59^a	1.54	1.04^a	2.15	0.97	2.09	0.64	1.81	0.82	1.70	0.55	0.82	0.78	1.99	0.98	2.01
Own physical victimization	1.27^a	1.45	1.78^a	2.00	1.81^b	1.80	1.16^b,d	1.25	1.57^d	2.07	1.52	1.45	1.42	1.71	1.71	1.96
Own relational bullying	1.59^a	2.14	1.15^a	1.77	1.73^b,c	2.29	1.07^b	1.59	1.26^c	1.88	1.27	1.58	1.33	2.01	1.47	2.07
Own relational victimization	3.32^a	3.27	2.35^a	2.49	3.75^b,c	3.39	2.21^b	2.40	2.43^c	2.65	3.07	3.27	2.75	3.01	2.88	2.69
Friends’ physical bullying	0.87^a	0.72	1.08^a	0.97	0.99^b	0.87	0.81^b,d	0.70	1.11^d	0.95	0.93	0.78	0.92^f	0.83	1.08^f	0.93
Friends’ physical victimization	1.60^a	0.84	1.97^a	1.09	2.10^b,c	0.94	1.36^b,d	0.58	1.84^c,d	1.18	1.73	0.90	1.73^f	0.94	1.90^f	1.09
Friends’ relational bullying	1.77^a	1.02	1.45^a	1.03	2.01^b,c	1.15	1.29^b,d	0.79	1.50^c,d	0.96	1.54	0.95	1.56^f	0.98	1.74^f	1.15
Friends’ relational victimization	3.77^a	1.68	3.12^a	1.66	4.46^b,c	1.82	2.69^b,d	1.28	3.09^c,d	1.38	3.62	1.61	3.36	1.75	3.53	1.66
Social centrality	1.13^a	0.41	0.76^a	0.42	0.92	0.48	0.95	0.47	0.96	0.42	0.88^e	0.47	0.99^e,f	0.43	0.89^f	0.50
Knowledge of physical bullies	5.10^a	9.03	3.54^a	6.10	5.73^b,c	8.41	3.66^b	8.05	3.50^c	6.46	4.67	8.33	4.00	7.67	4.72	7.63
Knowledge of physical victims	6.29	10.61	5.02	10.14	7.00^b	10.60	2.94^b,d	5.99	6.89^d	12.79	4.15	9.24	5.67	10.37	6.27	10.85
Knowledge of relational bullies	6.92^a	7.06	4.73^a	6.25	8.49^b,c	7.44	3.71^b,d	4.70	4.98^c,d	6.71	5.79	5.76	5.28^f	6.43	6.80^f	7.56
Knowledge of relational victims	21.49^a	24.44	12.58^a	18.58	26.28^b,c	26.53	10.79^b	15.98	13.00^c	18.48	17.29	21.34	16.15	22.19	18.45	22.43

Note. Ns range from 908 to 951. Descriptive statistics for non-transformed variables are shown. Comparisons on variables within gender, grade, and ethnicity were made using t tests and ANOVAs. Significant gender, grade, and ethnicity differences on each variable are indicated with matching subscripts.

Indicates significant (p < .01) difference between boys and girls.

Indicates significant (p < .05) difference between sixth and seventh graders.

Indicates significant (p < .05) difference between sixth and eighth graders.

Indicates significant (p < .05) difference between seventh and eighth graders.

Indicates significant (p < .05) difference between White and Latino/a ethnicity.

Indicates significant (p < .05) difference between Latino/a and Other minority ethnicity.

In addition, for the knowledge variables, girls had more knowledge than boys for all types of bullying and victimization, although only marginally for physical victimization, t(949) = 1.83, p = .068, ts(949) ranged from 3.05 to 6.23, p < .05. Sixth graders had more knowledge of physical victims than seventh graders, and had more knowledge of physical bullies, relational bullies, and relational victims than seventh or eighth graders, Fs(2, 948) ranged from 8.09 to 50.05, p < .001. Eighth graders also had more knowledge of physical victims and relational bullies than seventh graders, Fs(2, 948) = 15.17 and 46.32, ps < .001. Finally, those who identified as Other minority had more knowledge of relational bullies than those who identified as Latino/a, F(2, 948) = 4.77, p < .01. Taken together, these analyses indicate the need to control for gender, grade, and ethnicity in subsequent analyses.

Correlations among variables

All predictor and outcome variables (with the exception of the relations between social centrality and indicators of physical bullying, and the relation between knowledge of relational victims and friends’ physical bullying) were significantly positively correlated with one another. Correlations were of low to modest magnitude (see Table 3).

Table 3.

Correlations of Study Variables.

	1	2	3	4	5	6	7	8	9	10	11	12	13
1. Own physical bullying	—
2. Own physical victimization	.35***	—
3. Own relational bullying	.37***	.38***	—
4. Own relational victimization	.25***	.39***	.48***	—
5. Friends’ physical bullying	.16***	.23***	.14***	.10**	—
6. Friends’ physical victimization	.24***	.36***	.19***	.21***	.51***	—
7. Friends’ relational bullying	.12***	.27***	.23***	.28***	.53***	.45***	—
8. Friends’ relational victimization	.11**	.23***	.25***	.36***	.26***	.49***	.58***	—
9. Social centrality	−.04	.04	.15***	.19***	−.01	.06	.27***	.29***	—
10. Knowledge of physical bullies	.08*	.23***	.16***	.18***	.12***	.22***	.19***	.22***	.16***	—
11. Knowledge of physical victims	.08*	.28***	.14***	.17***	.11**	.25***	.22***	.22***	.19***	.73***	—
12. Knowledge of relational bullies	.12***	.18***	.25***	.36***	.09**	.21***	.28***	.35***	.15***	.41***	.38***	—
13. Knowledge of relational victims	.14***	.20***	.31***	.10***	.04	.19***	.29***	.41***	.22***	.43***	.51***	.68***	—

Note. Ns range from 908 to 951.

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

Hypothesis testing

To test the hypothesis that knowledge of bullying was differentially distributed among peers, with some students having greater or lesser knowledge of who bullies and who is victimized, we used the measure of knowledge of bullies or victims as the dependent variable in four separate hierarchical regressions (one each for knowledge of physical bullies, physical victims, relational bullies, and relational victims). For each regression, demographic covariates included ethnicity (with separate indicators of White and Latino/a, dummy coded with Other minority ethnicity as the reference category), grade (dummy coded with sixth grade as the reference category), and gender (coded such that 1 = male, 0 = female), and they were entered in the first step. In addition, in each regression, knowledge of the other form of bullies or victims was also entered as a covariate. For instance, in the regression predicting knowledge of physical bullies, knowledge of relational bullies was included as a covariate. Predictors, entered in a second step, included one’s own bullying and victimization scores, friends’ bullying and victimization scores, and social centrality. Note that one’s own and friends’ bullying and victimization scores were entered such that they matched the form of bullying being predicted. In other words, one’s own physical bully and victim scores and friends’ physical bully and victim scores were entered when predicting knowledge of physical bullies and victims, whereas relational bully and victim scores and friends’ relational bully and victim scores were entered when predicting knowledge of relational bullies and victims.

For knowledge of physical bullies, covariates were entered first and produced an R² of .17 (F = 29.30, p < .001). Of all the covariates, only knowledge of relational bullies significantly predicted knowledge of physical bullies. In the second step, the predictors were entered producing an R² change of .04 (F = 8.32, p < .001), indicating that 4% of the variance in knowledge of physical bullies was accounted for by the set of predictors, controlling for the covariates. Own physical victimization, friends’ physical victimization, and social centrality significantly and positively predicted knowledge of physical bullies (see Table 4). Thus, those who were more physically victimized, had friends who were more physically victimized, and who were more socially central had more knowledge of physical bullies.

Table 4.

Regression Predicting Knowledge of Physical Bullies and Victims.

	Knowledge of physical bullies	Knowledge of physical victims
	β	β
Covariates
Knowledge of relational bullies/victims	.09***	.05***
White	.01	−33
Latino/a	−.08	−.03
Male	.02	.08
Seventh grade	−.16	−.13
Eighth grade	−.17	.37*
R ²	.17	.53
F	29.30***	55.39***
Predictors
Own physical bullying	−.06	−.20*
Own physical victimization	.13***	.18***
Friends’ physical bullying	.04	−.01
Friends’ physical victimization	.14*	.24**
Social centrality	.30*	.24
R ²	.21	.33
ΔR²	.04	.05
ΔF	8.32***	12.16***

Note. Ethnicity variables dummy coded with “other ethnicity” as the reference group. Male is coded such that males = 1, females = 0. Grade variables dummy coded with sixth grade as the reference group.

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

Results for knowledge of physical victims were very similar. As before, when the covariates were entered, knowledge of relational victims predicted knowledge of physical victims. In addition, being in eighth grade (as compared with sixth grade) positively predicted knowledge of physical victims. When predictors were added, the results indicated that those who were more physically victimized and who had friends who were more physically victimized had more knowledge of physical victims. Own physical bullying negatively predicted knowledge of physical victims (see Table 4). The R² change after adding in the predictors was .05 (F = 12.16, p < .001), indicating that 5% of the variance in knowledge of physical victims was accounted for.

For knowledge of relational bullies, knowledge of physical bullies positively predicted knowledge of relational bullies (see Table 5). In addition, those who were White and Latino/a (compared with other ethnicity), male (compared with female), and in seventh and eighth grade (compared with sixth grade) had less knowledge of relational bullies. Next, the predictors were entered, which produced an R² change of .07 (F = 16.58, p < .001), indicating that 7% of the variance in knowledge of relational bullies was accounted for by the set of predictors, controlling for covariates. Specifically, own relational bullying (marginally), own relational victimization, and friends’ relational victimization positively predicted knowledge of relational bullies.

Table 5.

Regression Predicting Knowledge of Relational Bullies and Victims.

	Knowledge of relational bullies	Knowledge of relational victims
	β	β
Covariates
Knowledge of physical bullies/victims	1.46***	5.34***
White	−1.51*	−.31
Latino/a	−1.15*	−1.38
Male	−1.93***	−7.22***
Seventh grade	−3.90***	−11.03***
Eighth grade	−2.64**	−11.57***
R ²	.24	.35
F	47.28***	77.13***
Predictors
Own relational bullying	.51^†	3.00***
Own relational victimization	.41***	1.37***
Friends’ relational bullying	.36	.04
Friends’ relational victimization	.36*	1.55**
Social centrality	.03	1.67
R²	.31	.44
ΔR²	.07	.09
ΔF	16.58***	26.14***

†

p = .06. *p < .05. **p < .01. ***p < .001.

Finally, for knowledge of relational victims, the covariates were entered and produced an R² of .35 (F = 77.13, p < .001). Knowledge of physical victims was positively related to knowledge of relational victims. In addition, male students (compared with females) and seventh and eighth graders (compared with sixth graders) had less knowledge of relational victims. In the second step, social role predictors were entered and produced an R² change of .09 (F = 26.14, p < .001), indicating that social role predictors accounted for 9% of the variance in knowledge of relational victims. As with knowledge of relational bullies, students who were more relationally bullied and victimized and students whose friends were more relationally victimized had greater knowledge of relational victims (see Table 5).

Discussion

Bullying touches some youth’s lives more deeply than others. It has long been known that some youth are more likely to be involved in bullying as either bullies or victims (or both), and the results of this study suggest that a similar conclusion is warranted with regard to middle school grademates’ knowledge of who bullies whom. Approximately half (roughly 40% to 65%) of the middle schoolers in this sample were able to identify grademates who bully and who were bullied using physical or relational means. This estimate is slightly lower than estimates of students’ overall exposure to bullying, wherein between 60% and 70% of students reported that they have witnessed bullying (Aboud & Miller, 2007; Bradshaw et al., 2007; Cuadrado-Gordillo, 2012; Rivers et al., 2009). This apparent difference may indicate that only a subset of those who have witnessed bullying have sufficient information about what they witnessed to allow them to identify bullies and victims. Indeed, even among those who witness bullying, there is likely to be variation in how much specific knowledge individuals have. This would depend on how frequently or recently individuals were exposed to bullying, the nature of the bullying episode(s), and their own tendency to attend to bullying events (Greenhoot et al., 2006; Tversky & Kahneman, 1973; Wyer, 1980). The middle school students in this sample moved from class to class over the course of the school day. This is a typical experience of many middle school students in U.S. schools. As such, they are exposed to a greater breadth of grademates than those students who are contained within a single class throughout the day (as is the case for most U.S. elementary school classrooms), but they have also less persistent and sustained interactions with classmates. It is plausible that estimates of the degree to which students are knowledgeable of who bullies whom would vary by the setting characteristics and the amount of exposure that students have to their peers. Thus, it will be important for future research to replicate this study in samples that vary in school setting characteristics. For instance, one might find that knowledge of bullying and victimized classmates is greater in elementary school samples (as compared with a middle school sample), but that knowledge of bullying and victimized grademates from different classes would be lower.

However, we also found variations in knowledge by form of aggressive behavior. Knowledge of relationally aggressive bullies and victims (for which an average of 62% of students could identify those involved) was considerably greater than knowledge of physically aggressive bullies and victims (for which an average of 44% of students could identify those involved). This finding may reflect the fact that relationally aggressive behaviors serve a unique social function. In early adolescence, relationally aggressive behaviors are more commonly used and more socially sanctioned than physically aggressive behaviors (Cillessen & Mayeux, 2004). Moreover, because relationally aggressive behaviors are more effective than physically aggressive behaviors in manipulating one’s own and others’ status in the peer group, they may be highly relevant to middle schoolers, who tend to be concerned with negotiating the social hierarchy (Hawley, 1999). In fact, the use of relational aggression has been positively linked with social prominence (Cillessen & Mayeux, 2004). As such, this behavior may garner attention and awareness from peers, at least from those peers who are socially connected to bullying episodes.

Effects of Social Distance and Social Role

Knowledge of who bullies whom was not randomly distributed among all members of the peer group. Rather, those middle schoolers who had greater knowledge tended to be those who were most likely to participate in, witness, or hear about bullying because of their social positions. For instance, being victimized oneself and being friends with victimized youth were associated with greater knowledge of bullies and their victims. These were robust findings, evident across all analyses. It is not surprising that victims and their friends had relatively high levels of knowledge. The more relevant an event is to an individual, the more event-related knowledge that individual should have (Greenhoot et al., 2006; Swann et al., 1981; Wyer, 1980). To illustrate, victimized youth are aware of the harm that accompanies bullying behavior (Cuadrado-Gordillo, 2012). They might cope with it by, for example, seeking support, avoidance, or retaliation. A coping strategy that involves using knowledge of bullies, victims, and bullying episodes to manage or minimize risk from threatening situations can be effective in reducing short- or long-term victimization risk (Visconti & Troop-Gordon, 2010). Moreover, because friends can serve as a buffer for direct victimization as well as for victimization-related maladjustment (Holt & Espelage, 2007; Pellegrini et al., 1999; Scholte et al., 2009), ensuring that friends are knowledgeable about bullying can be beneficial as well.

Whereas predictions from victimization (one’s own and that of one’s friends) were quite robust, being nominated by peers as a bully was unrelated to knowledge of physical bullies and negatively related to knowledge of victims of physical aggression. In contrast, being nominated by peers as a bully was positively related to knowledge of relational bullies (albeit only marginally) and victims of relational aggression. Although these findings were far from robust, the relatively positive effects for relational versus physical aggression may reflect a tendency toward greater social sensitivity or social awareness in those who use relationally aggressive behaviors; further research might consider this point in greater depth. However, we caution against over-interpreting this possibility here. A more prudent interpretation is that these mixed findings are indicative of an overall null effect of bullying on knowledge. In further support of this idea was the fact that there was no evidence that friends of bullies were able to identify either bullies or victims. This finding stands in contrast to the very robust findings for victims and their friends. In other words, despite the fact that both bullies and victims (and their friends) are equally close to bullying and, therefore, should have similar opportunities to learn about who bullies whom, they were differentially likely to name peers as bullies or victims. There are two likely explanations for this.

Perhaps the apparent difference in knowledge of bullying for those who bully (and their friends) compared with those who are victimized (and their friends) reflects perceptual differences. In general, young adolescents define bullying largely on the occurrence of harmful behaviors (e.g., physical, relational, or verbal aggression or harassment), and they rarely consider bullying to be characterized by factors such as intentionality, repetition, and power imbalances (Vaillancourt et al., 2008). Bullies and their friends may not perceive their own behavior (or that of other bullies) as negative or harmful, particularly in the case of physical bullying. In fact, the moral disengagement that often accompanies bullying behavior can lead bullies to misinterpret their own behavior, for example, as a joke or justified in some way (Gini, 2006; Hymel, Schonert-Reichl, Bonanno, Vaillancourt, & Henderson, 2010). In further support of this point, attributional research has suggested that physically aggressive youth misattribute responsibility for aggression to the victim. For example, they tend to interpret others’ behaviors as hostile or provocative (e.g., hostile attribution bias) and blame the victim for his plight (Dodge, 2006; Slaby & Guerra, 1988). Thus, physically aggressive bullies (and their friends) may interpret their own behavior as righteous, rather than as aggressive. The fact that students high on relationally aggressive bullying did nominate peers as victims might indicate that they are more socially sensitive than those high on physically aggressive bullying and thus able to identify bullies and victims.

This difference in identification of bullies and victims for involved youth may also reflect motivational differences in knowledge (Greenhoot et al., 2006; Swann et al., 1981). That is, victims and their friends may have something to gain by being willing to identify bullies and victims, such as retaliation and retribution (with regard to identifying bullies) or support and assistance (with regard to identifying victims). In comparison, bullies and their friends may fear punishment and so avoid naming names in an attempt to protect oneself. Further research that could link motivational and perceptual features with such knowledge would help to clarify the reasons that underlie apparent victim-bully differences in knowledge.

Being relatively high in social centrality was also associated with greater knowledge of bullies, but only for physically aggressive bullying. Socially central youth have social ties to many peers. Because they are so highly connected in the friendship network, they should have more opportunities (relative to their peers) to learn about bullying episodes. This might happen directly, as when they witness bullying interactions in the company of peers. It might also happen indirectly, as their connections in the network should enable them to hear about events second hand. The fact that this effect was only evident for knowledge of physically aggressive bullies suggests that information about relatively visible and impactful bullying events is more likely to make its way to socially central students than is information about events that are more often subtle and covert, as is the case for relational behaviors.

Implications

The present findings have both methodological and practical implications. On the methodological side, it is notable that there was variability across grademates in their nominations of peers as bullies and victims. Others have similarly reported variability across peers in their perceptions of bullies and victims, suggesting that perceptions of who bullies whom are related to one’s own experiences with those particular peers (Card et al., 2010). Indeed, in our own study, variables representing victims and friends of victims were consistently related to knowledge. Peer nominations are often used to assess students’ involvement in bullying perpetration and victimization. Typically, nominations are aggregated to produce an estimate of the overall peer consensus of individual children’s social positions; this is done to create a measure that minimizes the unique biases of any particular peer (Kenny et al., 2007). Yet, given the findings of the present study, even an aggregated peer nomination measure likely is weighted by the perceptions of certain peers, namely those who are more closely connected to bullying, relative to others. As such, peer nominations of bullying might be interpreted as representing the impressions of those who are most likely to be in the know more so than the impressions of those who are less knowledgeable.

On the practical side, the findings also speak to interventions to reduce bullying in schools. On the one hand, knowledge can be beneficial in that it enables one to process and use the information (Wyer, 1980). It is for this very reason that early school-based anti-bullying interventions included awareness-raising components (e.g., Olweus, 1991). More recent intervention programs have assumed peer knowledge and have built on that assumption by teaching peers empathic responses to victims, anti-bullying attitudes, and efficacy to intervene in bullying episodes (e.g., KiVa; Kärnä et al., 2011). Indeed, some positive effects are evident for such programs (Polanin, Espelage, & Pigott, 2012). However, as the present findings suggest, it is unlikely that all students in a school would be equally responsive to an intervention approach that teaches peers to step in and stop bullying (see also Caravita et al., 2009; Espelage, Green, & Polanin, 2012; Pöyhönen, Juvonen, & Salmivalli, 2010; Sainio et al., 2010). Those students who are more socially distant to bullying and, therefore, less likely to know who is bullying whom and the circumstances surrounding the bullying are presumably less able to intervene, even if they have the skills.

On the other hand, there is cause to be cautious about knowledge and how it is generated and used. Heightened knowledge of bullying and who bullies whom might result in an environment in which bullying becomes the social norm; thus, attunement to bullying can have negative effects (Cialdini, 2003). Aggressive social norms covary with a peer structure that is dominance-oriented, hierarchical, and rigid, and with the use of aggressive behaviors (contemporaneously and over time) to harm others or to display and achieve power (Ahn, Garandeau, & Rodkin, 2010; Garandeau, Ahn, & Rodkin, 2011; Henry et al., 2000; Huesmann & Guerra, 1997). This raises questions for future research to address about what the optimal level of knowledge is (or should be) and how knowledge should be managed to facilitate positive, rather than negative, outcomes.

Limitations and Conclusions

In the present study, we considered the following questions: In the complex and dynamic world of middle school peer interactions, how knowledgeable are students of the details of bullying episodes and who is most likely to know? The findings paint a picture of heterogeneity with regard to the bullying-related peer context at this developmental level, suggesting that some students are more knowledgeable about bullying than others. Several issues must be considered in interpreting the findings. The first issue has to do with the extent to which the findings of this study should be interpreted as referencing bullying, specifically, or aggression, more generally. Certainly there is overlap, possibly substantial overlap, between the two constructs (Card & Hodges, 2009; Coie et al., 1999). Yet, bullying is also thought to be a specific type of aggression that is characterized by an existing relationship between aggressor and victim (albeit a negatively toned relationship) that is marked by greater social, psychological, or physical power of the aggressor relative to the victim and that enables repeated attacks over time. Differentiating between the two related constructs is tricky, and although the conceptual literature regarding their differentiation is well-developed, the empirical literature relevant to exactly how to distinguish the two, and the extent to which the two are indeed distinctly different, is extremely limited (Rodkin et al., 2015). Regardless, there is reason to believe that the current findings are reflective of bullying per se. That is, we assessed bullying using a measurement procedure that embeds the relational features that are characteristic of bullying within the assessment by asking students to nominate bullies and their victims (Ahn et al., 2013). This method has been used in previous studies that speak to the relationship between bully and victim (e.g., Card et al., 2010; Rodkin & Berger, 2008; Veenstra et al., 2007). Moreover, this measure is sensitive to power imbalances between the bully and his or her victim and to the relationship of the dyad, both of which uniquely distinguish bullying from aggression (Rodkin, Hanish, Wang, & Logis, 2014).

We measured knowledge of bullying by asking students to peer-nominate bullies and their victims. This is an innovative use of peer nomination methods. Peer nominations are most often used to create an index of nominated peers’ reputational status; however, in the present study, we used peer nominations to create an index of nominators’ knowledge. This knowledge may come from direct involvement in bullying episodes, witnessing bullying episodes, or secondhand information about bullying; the methods used in the present study did not allow differentiation of the source of knowledge. We controlled for prevalence of bullies and victims within role (i.e., bullies and victims), grade, and form (i.e., relational and physical bullying). Nonetheless, it is possible that comparisons of knowledge across role, grade, or form reflect, in part, variations in prevalence of bullies versus victims, variations in prevalence of bullying across the middle school grades, or variations in prevalence of relational and physical bullying. Thus, these comparisons should be replicated using methods that allow more fine-grained controls for prevalence. Nevertheless, the tests of the primary social distance and social role hypotheses should not be affected by this possibility because these tests were conducted separately for relational and physical bullies and victims.

Furthermore, in calculating the knowledge variable, students who could identify one or more peers for each question received a non-zero score (with higher scores reflecting greater knowledge), and those who could not identify peers received a score of zero. It is possible that those students who did not respond to the measure did so for other reasons. Systematic alternative explanations for lack of response seem improbable, however. For instance, perhaps those who did not respond were fatigued. Although this measure was placed near the end of the questionnaire packet, students who failed to identify peers as bullies and victims did complete subsequent questions. Alternatively, perhaps those who did not respond were afraid to nominate peers. This explanation might apply to bullies and their friends (as discussed above), but the fact that victims and their friends did nominate peers as bullies and victims suggests that students’ concern for their own or others’ personal safety was unlikely to impact nominations. Finally, many of the students who did not identify bullies and victims verbally informed study personnel that they did not have sufficient information to answer the questions. In other words, alternative explanations are unlikely.

In the present study, we tested the hypotheses that one’s social distance to bullying and one’s social role predicted knowledge. Yet, there are numerous alternative predictors that we did not consider. For instance, knowledge may also be a function of intrapersonal factors, such as motivation to attend to social events, social sensitivity or social anxiety, empathy, morality, or assertiveness. In addition, knowledge may be influenced by contextual factors. For instance, students in classrooms, grades, or schools with higher rates of bullying may have greater knowledge of bullying than students in contexts in which bullying is less common. Exploring contextual-level predictors would be an interesting next step for this line of research and it would support generalizability (notably, however, our findings converge with those of Card et al., 2010). It would require, however, a much larger sampling at the contextual level than was available in this study. The significant, but modest, effect sizes that we obtained in the present study hint that exploring other possible predictors would be fruitful. Thus, this study represents but one step in a larger program of research related to understanding the bullying-related climate of schools and the extent to which bullying functions within the social sphere.

Overall, the findings speak to the need to further expand our models of the relational features of bullying (Hanish et al., 2013). In a field where studies of psychological processes predominate, continuing to explore the social embedding of bullying behavior adds an important dimension to understanding this phenomenon. A challenge for the future of bullying research is to build on this knowledge, feeding it into the next generation of prevention programs.

Footnotes

Acknowledgements

The authors would like to thank the students who contributed to this project as well as the school, teachers, and children for their participation.

Authors’ Note

Portions of this research were presented at the Society for Research on Child Development in 2013.

Declaration of Conflicting Interests

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

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported, in part, by the T. Denny Sanford School of Social and Family Dynamics and by the T. Denny Sanford Foundation.

Author Biographies

Laura D. Hanish, PhD, T. Denny Sanford School of Social and Family Dynamics, Arizona State University. Research interests include aggression and victimization, gender, peer relationships, and school-related contexts and outcomes.

Naomi C. Z. Andrews, MS, T. Denny Sanford School of Social and Family Dynamics, Arizona State University. Research interests include aggression, gender, and social networks.

Olga Kornienko, PhD, Department of Psychology and Institute for Interdisciplinary Bioscience Research, Arizona State University. Research interests include peer networks and network dynamics.

Carlos E. Santos, PhD, Department of Counseling and Counseling Psychology, Arizona State University. Research interests include identity development and social contexts.

Philip C. Rodkin, PhD, died during completion of this article. We acknowledge his contribution to the article and his leadership in advancing understanding of the social dynamics of adolescence.

Kristen L. Granger, MS, T. Denny Sanford School of Social and Family Dynamics, Arizona State University. Research interests include teacher influences on peer relationships.

References

Aboud

Miller

(2007). Promoting peer intervention in name-calling. South African Journal of Psychology, 37, 803-819. doi:10.1177/008124630703700409

Adams

R. E.

Bukowski

W. M.

Bagwell

(2005). Stability of aggression during early adolescence as moderated by reciprocated friendship status and friend’s aggression. International Journal of Behavioral Development, 29, 139-145. doi:10.1080/01650250444000397

Ahn

H.-J.

Garandeau

C. F.

Rodkin

P. C.

(2010). Effects of classroom embeddedness and density on the social status of aggressive and victimized children. Journal of Early Adolescence, 30, 76-101. doi:10.1177/0272431609350922

Ahn

H.-J.

Rodkin

P. C.

Gest

S. D.

(2013). Teacher-student agreement on “bullies and kids they pick on” in elementary school classrooms: Gender and grade differences. Theory Into Practice, 52, 257-263.

Bonacich

(1987). Power and centrality: A family of measures. American Journal of Sociology, 92, 1170-1182.

Bradshaw

C. P.

Sawyer

A. L.

O’Brennan

L. M.

(2007). Bullying and peer victimization at school: Perceptual differences between students and school staff. School Psychology Review, 36, 361-382.

Caravita

S. C. S.

DiBlasio

Salmivalli

(2009). Unique and interactive effects of empathy and social status on involvement in bullying. Social Development, 18, 140-163. doi:10.1111/j.1467-9507.2008.00465.x

Card

N. A.

Hodges

E. V. E.

(2009). It takes two to fight in school, too: A social relations model of the psychometric properties and relative variance of dyadic aggression and victimization in middle school. Social Development, 19, 447-469.

Card

N. A.

Rodkin

P. C.

Garandeau

C. F.

(2010). A description and illustration of the Triadic Relations Model: Who perceives whom as bullying whom? International Journal of Behavioral Development, 34, 374-383.

10.

Cialdini

R. B.

(2003). Crafting normative messages to protect the environment. Current Directions in Psychological Science, 12, 105-109. doi:10.1111/1467-8721.01242

11.

Cillessen

A. H. N.

Mayeux

(2004). From censure to reinforcement: Developmental changes in the association between aggression and social status. Child Development, 75, 147-163. doi:10.1111/j.1467-8624.2004.00660.x

12.

Cohen

Hsueh

Russell

K. M.

Ray

G. E.

(2006). Beyond the individual: A consideration of context for the development of aggression. Aggression and Violent Behavior, 11, 341-351. doi:10.1016/j.avb.2005.10.004

13.

Coie

J. D.

Cillessen

A. H. N.

Dodge

K. A.

Hubbard

J. A.

Schwartz

Lemerise

E. A.

Bateman

(1999). It takes two to fight: A test of relational factors and a method of assessing aggressive dyads. Developmental Psychology, 35, 1179-1188.

14.

Craig

W. M.

Pepler

D. J.

(1998). Observations of bullying and victimization in the school yard. Canadian Journal of School Psychology, 13(2), 41-60. doi:10.1177/082957359801300205

15.

Craig

W. M.

Pepler

D. J.

Atlas

(2000). Observations of bullying in the playground and in the classroom. School Psychology International, 21, 22-36. doi:10.1177/0143034300211002

16.

Cross

Parker

Prusak

Borgatti

S. P.

(2001). Knowing what we know: Supporting knowledge creation and sharing in social networks. Organizational Dynamics, 30, 100-120. doi:10.1016/S0090-2616(01)00046-8

17.

Cuadrado-Gordillo

(2012). Repetition, power imbalance, and intentionality: Do these criteria conform to teenagers’ perception of bullying? A role-based analysis. Journal of Interpersonal Violence, 27, 1889-1910. doi:10.1177/0886260511431436

18.

Dodge

K. A.

(2006). Translational science in action: Hostile attributional style and the development of aggressive behavior problems. Development and Psychopathology, 18, 791-814. doi:10.1017/S0954579406060391

19.

Dodge

K. A.

Murphy

R. R.

Buchsbaum

(1984). The assessment of intention-cue detection skills in children: Implications for developmental psychopathology. Child Development, 55, 163-173. doi:10.2307/1129842

20.

Espelage

D. L.

Green

H. D.

Polanin

(2012). Willingness to intervene in bullying episodes among middle school students: Individual and peer-group influences. Journal of Early Adolescence, 32, 776-801. doi:10.1177/0272431611423017

21.

Garandeau

C. F.

Ahn

H.-J.

Rodkin

P. C.

(2011). The social status of aggressive students across contexts: The role of classroom status hierarchy, academic achievement, and grade. Developmental Psychology, 47, 1699-1710. doi:10.1037/a0025271

22.

Gest

S. D.

(2006). Teacher reports of children’s friendships and social groups: Agreement with peer reports and implications for studying peer similarity. Social Development, 15, 248-259. doi:10.1046/j.1467-9507.2006.00339.x

23.

Gini

(2006). Social cognition and moral cognition in bullying: What’s wrong? Aggressive Behavior, 32, 528-539.

24.

Greenhoot

A. F.

Tsethlikai

Wagoner

B. J.

(2006). The relation between children’s past experiences, social knowledge, and memories for social situations. Journal of Cognition and Development, 7, 313-340. doi:10.1207/s15327647jcd0703_6

25.

Hanish

L. D.

Bradshaw

C. P.

Espelage

D. L.

Rodkin

P. C.

Swearer

S. M.

Horne

(2013). Looking toward the future of bullying research: Recommendations for research and funding priorities. Journal of School Violence, 12, 283-295. doi:10.1080/15388220.2013.788449

26.

Hawley

P. H.

(1999). The ontogenesis of social dominance: A strategy-based evolutionary perspective. Developmental Review, 19, 97-132. doi:10.1006/drev.1998.0470

27.

Henry

Guerra

N. G.

Huesmann

Tolan

P. H.

Van Acker

Eron

(2000). Normative influences on aggression in urban elementary school classrooms. American Journal of Community Psychology, 28, 59-81. doi:10.1023/A:1005142429725

28.

Hodges

E. V. E.

Boivin

Vitaro

Bukowski

W. M.

(1999). The power of friendship: Protection against an escalating cycle of peer victimization. Developmental Psychology, 35, 94-101. doi:10.1037/0012-1649.35.1.94

29.

Holt

M. K.

Espelage

D. L.

(2007). Perceived social support among bullies, victims, and bully-victims. Journal of Youth and Adolescence, 36, 984-994. doi:10.1007/s10964-006-9153-3

30.

Huesmann

L. R.

Guerra

N. G.

(1997). Children’s normative beliefs about aggression and aggressive behavior. Journal of Personality and Social Psychology, 72, 408-419. doi:10.1037/0022-3514.72.2.408

31.

Huitsing

Veenstra

(2012). Bullying in classrooms: Participant roles from a social network perspective. Aggressive Behavior, 38, 494-509. doi:10.1002/ab.21438

32.

Hymel

Schonert-Reichl

K. A.

Bonanno

R. A.

Vaillancourt

Henderson

N. R.

(2010). Bullying and morality: Understanding how good kids can behave badly. In Jimerson

S. R.

Swearer

S. M.

Espelage

D. L.

(Eds.), Handbook of bullying in schools: An international perspective (pp. 101-118). New York, NY: Routledge.

33.

Juvonen

Nishina

Graham

(2001). Self-views versus peer perceptions of victim status among early adolescents. In Juvonen

Nishina

Graham

(Eds.), Peer harassment in school: The plight of the vulnerable and victimized (pp. 105-124). New York, NY: Guilford.

34.

Kärnä

Voeten

Little

T. D.

Poskiparta

Kaljonen

Salmivalli

(2011). A large-scale evaluation of the KiVa antibullying program: Grades 4-6. Child Development, 82, 311-330. doi:10.1111/j.1467-8624.2010.01557.x

35.

Kenny

D. A.

West

T. V.

Cillessen

A. H. N.

Coie

J. D.

Dodge

K. A.

Hubbard

J. A.

Schwartz

(2007). Accuracy in judgments of aggressiveness. Personality and Social Psychology Bulletin, 33, 1225-1236.

36.

Krantz

(1982). Sociometric awareness, social participation, and perceived popularity in preschool children. Child Development, 53, 376-379. doi:10.2307/1128979

37.

LaFontana

K. M.

Cillessen

A. H. N.

(1999). Children’s interpersonal perceptions as a function of sociometric and peer-perceived popularity. Journal of Genetic Psychology, 160, 225-242.

38.

Nansel

T. R.

Overpeck

Pilla

R. S.

Ruan

Simons-Morton

Scheidt

(2001). Bullying behaviors among US youth: Prevalence and association with psychosocial adjustment. Journal of the American Medical Association, 285, 2094-2100. doi:10.1001/jama.285.16.2094

39.

Newcomb

A. F.

Bagwell

(1995). Children’s friendship relations: A meta-analytic review. Psychological Bulletin, 117, 306-347. doi:10.1037/0033-2909.117.2.306

40.

Olweus

(1991). Bully/victim problems among schoolchildren: Basic facts and effects of a school based intervention program. In Pepler

D. J.

Rubin

K. H.

(Eds.), The development and treatment of childhood aggression (pp. 411-448). Hillsdale, NJ: Lawrence Erlbaum.

41.

Parault

S. J.

Davis

H. A.

Pellegrini

A. D.

(2007). The social contexts of bullying and victimization. Journal of Early Adolescence, 27, 145-174. doi:10.1177/0272431606294831

42.

Pellegrini

A. D.

Bartini

Brooks

(1999). School bullies, victims, and aggressive victims: Factors relating to group affiliation and victimization in early adolescence. Journal of Educational Psychology, 91, 216-224. doi:10.1037/0022-0663.91.2.216

43.

Pellegrini

A. D.

Long

J. D.

(2002). A longitudinal study of bullying, dominance, and victimization during the transition from primary school through secondary school. British Journal of Developmental Psychology, 20, 259-280. doi:10.1348/026151002166442

44.

Pepler

D. J.

Craig

W. M.

(1995). A peek behind the fence: Naturalistic observations of aggressive children with remote audiovisual recording. Developmental Psychology, 31, 548-553. doi:10.1037/0012-1649.31.4.548

45.

Polanin

Espelage

D. L.

Pigott

T. D.

(2012). A meta-analysis of school-based bullying prevention programs’ effects on bystander intervention behavior and empathy attitude. School Psychology Review, 41, 47-65.

46.

Pöyhönen

Juvonen

Salmivalli

(2010). What does it take to stand up for the victim of bullying? Merrill-Palmer Quarterly, 56, 143-163. doi:10.1353/mpq.0.0046

47.

Rivers

Poteat

V. P.

Noret

Ashurts

(2009). Observing bullying at school: The mental health implications of witness status. School Psychology Quarterly, 24, 211-223. doi:10.1037/a0018164

48.

Rock

P. F.

Cole

D. J.

Houshyar

Lythcott

Prinstein

M. J.

(2011). Peer status in an ethnic context: Associations with African American adolescents’ ethnic identity. Journal of Applied Developmental Psychology, 32, 163-169.

49.

Rodkin

P. C.

Berger

(2008). Who bullies whom? Social status asymmetries by victim gender. International Journal of Behavioral Development, 32, 473-485. doi:10.1177/0165025408093667

50.

Rodkin

P. C.

Espelage

D. L.

Hanish

L. D.

(2015). A relational framework for understanding bullying: Developmental antecedents and outcomes. American Psychologist, 70(4), 311-321.

51.

Rodkin

P. C.

Farmer

T. W.

Pearl

Van Acker

(2006). They’re cool: Social status and peer group supports for aggressive boys and girls. Social Development, 15, 175-204.

52.

Rodkin

P. C.

Hanish

L. D.

Wang

Logis

H. A.

(2014). Why the bully/victim relationship is so pernicious: A gendered perspective on power and animosity among bullies and their victims. Development and Psychopathology, 26, 689-704.

53.

Sainio

Veenstra

Huitsing

Salmivalli

(2010). Victims and their defenders: A dyadic approach. International Journal of Behavioral Development, 35, 144-151. doi:10.1177/0165025410378068

54.

Scholte

R. H. J.

Overbeek

ten Brink

Rommes

de Kemp

R. A. T.

Goossens

Engels

R. C. M. E.

(2009). The significance of reciprocal and unilateral friendships for peer victimization in adolescence. Journal of Youth and Adolescence, 38, 89-100. doi:10.1007/s10964-008-9287-6

55.

Sijtsema

J. J.

Rambaran

A. J.

Ojanen

T. J.

(2013). Overt and relational victimization and adolescent friendships: Selection, de-selection, and social influence. Social Influence, 8, 177-195.

56.

Slaby

R. G.

Guerra

N. G.

(1988). Cognitive mediators of aggression in adolescent offenders: I. Assessment. Developmental Psychology, 24, 580-588. doi:10.1037/0012-1649.24.4.580

57.

Swann

W. B.

Stephenson

Pittman

T. S.

(1981). Curiosity and control: On the determinants of the search for social knowledge. Journal of Personality and Social Psychology, 40, 635-642. doi:10.1037/0022-3514.40.4.635

58.

Terasahjo

Salmivalli

(2003). “She is not actually bullied.” The discourse of harassment in student groups. Aggressive Behavior, 29, 134-154. doi:10.1002/ab.10045

59.

Tomarken

A. J.

Serlin

R. C.

(1986). Comparison of ANOVA alternatives under variance heterogeneity and specific noncentrality structures. Psychological Bulletin, 99, 90-99. doi:10.1037/0033-2909.99.1.90

60.

Trope

Liberman

(2003). Temporal construal. Psychological Review, 110, 403-421. doi:10.1037/0033-295X.110.3.403

61.

Tversky

Kahneman

(1973). Availability: A heuristic for judging frequency and probability. Cognitive Psychology, 5, 207-232. doi:10.1016/0010-0285(73)90033-9

62.

Vaillancourt

McDougall

Hymel

Krygsman

Miller

Stiver

Davis

(2008). Bullying: Are researchers and children/youth talking about the same thing? International Journal of Behavioral Development, 32, 486-495. doi:10.1177/0165025408095553

63.

Vaughn

B. E.

Colvin

T. N.

Azria

M. R.

Caya

Krzysik

(2001). Dyadic analyses of friendship in a sample of preschool-age children attending Head Start: Correspondence between measures and implications for social competence. Child Development, 72, 862-878. doi:10.1111/1467-8624.00320

64.

Veenstra

Lindenberg

Zijlstra

B. J. H.

De Winter

A. F.

Verhulst

F. C.

Ormel

(2007). The dyadic nature of bullying and victimization: Testing a dual-perspective theory. Child Development, 78, 1843-1854. doi:10.1111/j.1467-8624.2007.01102.x

65.

Visconti

K. J.

Troop-Gordon

(2010). Prospective relations between children’s responses to peer victimization and their socioemotional adjustment. Journal of Applied Developmental Psychology, 31, 261-272. doi:10.1016/j.appdev.2010.05.003

66.

Wasserman

Faust

(1994). Social network analysis: Methods and applications. New York, NY: Cambridge University Press.

67.

Wyer

R. S.

(1980). The acquisition and use of social knowledge: Basic postulates and representative research. Personality and Social Psychology Bulletin, 6, 558-573. doi:10.1177/014616728064007

68.

Zimmerman

D. W.

Zumbo

B. D.

(1993). Rank transformations and the power of the student t test and Welch t’ test for non-normal populations with unequal variances. Canadian Journal of Experimental Psychology, 47, 523-539. doi:10.103/h0078850